1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.349 2006/04/27 08:53:24 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
10 other effects) broke the use of negated acl sub-conditions.
12 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
15 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
16 "Deny" causes Exim to reject the incoming connection with a 554 error.
17 Unfortunately, if there is a major crisis, such as a disk failure,
18 tcp-wrappers gives "deny", whereas what one would like would be some
19 kind of temporary error. A kludge has been added to help with this.
20 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
21 554 error is used if errno is still zero or contains ENOENT (which occurs
22 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
25 PH/03 Add -lutil to the default FreeBSD LIBS setting.
27 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
28 errors. Otherwise a message that provokes a temporary error (when other
29 messages do not) can cause a whole host to time out.
31 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
32 addresses were routed directly to files or pipes from a redirect router.
33 File deliveries just didn't batch; pipe deliveries might have suffered
36 PH/06 A failure to get a lock for a hints database would erroneously always say
37 "Failed to get write lock", even when it was really a read lock.
39 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
40 of 0600. This has been changed to use the value of the lockfile_mode
41 option (which defaults to 0600).
43 PH/08 Applied small patch from the Sieve maintainer.
45 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
46 folder from quota calculations, a direct delivery into this folder messed
47 up the contents of the maildirsize file. This was because the regex was
48 used only to exclude .Trash (or whatever) when the size of the mailbox
49 was calculated. There was no check that a delivery was happening into an
50 excluded directory. This bug has been fixed by ignoring all quota
51 processing for deliveries into excluded directories.
53 PH/10 Added the maildirfolder_create_regex option to appendfile.
59 PH/01 The code for finding all the local interface addresses on a FreeBSD
60 system running IPv6 was broken. This may well have applied to all BSD
61 systems, as well as to others that have similar system calls. The broken
62 code found IPv4 interfaces correctly, but gave incorrect values for the
63 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
64 that it would not match correctly against @[] and not recognize the IPv6
67 PH/02 The ipliteral router was not recognizing addresses of the form user@
68 [ipv6:....] because it didn't know about the "ipv6:" prefix.
70 PH/03 Added disable_ipv6.
72 PH/04 Changed $reply_address to use the raw form of the headers instead of the
73 decoded form, because it is most often used to construct To: headers
74 lines in autoreplies, and the decoded form may well be syntactically
75 invalid. However, $reply_address has leading white space removed, and all
76 newlines turned into spaces so that the autoreply transport does not
79 PH/05 If group was specified without a user on a router, and no group or user
80 was specified on a transport, the group from the router was ignored.
82 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
83 for visible compile-time settings that can be used to change these
84 numbers, for those that want even more. Backwards compatibility with old
85 spool files has been maintained. However, going back to a previous Exim
86 release will lost any variables that are in spool files.
88 PH/07 Two small changes when running in the test harness: increase delay when
89 passing a TCP/IP connection to a new process, in case the original
90 process has to generate a bounce, and remove special handling of
91 127.0.0.2 (sic), which is no longer necessary.
93 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
94 be the same on different OS.
96 PH/09 Moved a debug statement in filter processing to avoid a race problem when
99 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
100 whether --show-vars was specified or not
102 JJ/02 exipick: Added support for new ACL variable spool format introduced
105 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
106 syntactically invalid From: or Reply-to: line, and a filter used this to
107 generate an autoreply, and therefore failed to obtain an address for the
108 autoreply, Exim could try to deliver to a non-existent relative file
109 name, causing unrelated and misleading errors. What now happens is that
110 it logs this as a hard delivery error, but does not attempt to create a
113 PH/11 The exinext utility has a -C option for testing purposes, but although
114 the given file was scanned by exinext itself; it wasn't being passed on
115 when Exim was called.
117 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
118 an end-of-file indication when reading a command response.
120 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
121 compiled. In many other places in Exim, IPv6 addresses are always
122 recognized, so I have changed this. It also means that IPv4 domain
123 literals of the form [IPV4:n.n.n.n] are now always recognized.
125 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
126 used if the router is not running as root, for example, when verifying at
127 ACL time, or when using -bh. The debugging output from this situation was
128 non-existent - all you got was a failure to exec. I have made two
131 (a) Failures to set uid/gid, the current directory, or a process leader
132 in a subprocess such as that created by queryprogram now generate
133 suitable debugging ouput when -d is set.
135 (b) The queryprogram router detects when it is not running as root,
136 outputs suitable debugging information if -d is set, and then runs
137 the subprocess without attempting to change uid/gid.
139 PH/15 Minor change to Makefile for building test_host (undocumented testing
142 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
143 additional section of a DNS packet that returns MX or SRV records.
144 Instead, it always explicitly searches for A/AAAA records. This avoids
145 major problems that occur when a DNS server includes only records of one
146 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
147 fixed another bug: if SRV records were looked up and the corresponding
148 address records were *not* found in the additional section, the port
149 values from the SRV records were lost.
151 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
152 using the correct key (the original address) when searching the retry
153 rules in order to find which one to use for generating the retry hint.
155 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
156 adding the default one. Similarly, if it contains a Reply-To: header, the
157 errors_reply_to option, if set, is not used.
159 PH/19 When calculating a retry time, Exim used to measure the "time since
160 failure" by looking at the "first failed" field in the retry record. Now
161 it does not use this if it is later than than the arrival time of the
162 message. Instead it uses the arrival time. This makes for better
163 behaviour in cases where some deliveries succeed, thus re-setting the
164 "first failed" field. An example is a quota failure for a huge message
165 when small messages continue to be delivered. Without this change, the
166 "time since failure" will always be short, possible causing more frequent
167 delivery attempts for the huge message than are intended.
168 [Note: This change was subsequently modified - see PH/04 for 4.62.]
170 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
171 $1, $2, $3) because the numerical variables can be reset during some
172 expansion items (e.g. "match"), thereby losing the authentication data.
174 PH/21 Make -bV show the size of off_t variables so that the test suite can
175 decide whether to run tests for quotas > 2G.
177 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
178 mailbox_size, and mailbox_filecount in the appendfile transport. If a
179 filecount value is greater than 2G or if a quota value is greater than 2G
180 on a system where the size of off_t is not greater than 4, a panic error
183 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
184 never match. The debug and -bh output now contains an explicit error
185 message indicating a malformed IPv4 address or mask.
187 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
188 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
191 PH/25 Do not write to syslog when running in the test harness. The only
192 occasion when this arises is a failure to open the main or panic logs
193 (for which there is an explicit test).
195 PH/26 Added the /no_tell option to "control=freeze".
197 PH/27 If a host name lookup failed very early in a connection, for example, if
198 the IP address matched host_lookup and the reverse lookup yielded a name
199 that did not have a forward lookup, an error message of the form "no IP
200 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
201 could be logged. Now it outputs the IP address instead of "NULL".
203 PH/28 An enabling patch from MH: add new function child_open_exim2() which
204 allows the sender and the authenticated sender to be set when
205 submitting a message from within Exim. Since child_open_exim() is
206 documented for local_scan(), the new function should be too.
208 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
209 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
210 results in an empty string is now treated as unset.
212 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
214 PH/31 Added sender_verify_fail logging option.
216 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
217 needed by RFC 822 but not by RFC 2822 was commented out. I have now
218 tidied the source and removed it altogether.
220 PH/33 When a queue run was abandoned because the load average was too high, a
221 log line was always written; now it is written only if the queue_run log
222 selector is set. In addition, the log line for abandonment now contains
223 information about the queue run such as the pid. This is always present
224 in "start" and "stop" lines but was omitted from the "abandon" line.
226 PH/34 Omit spaces between a header name and the colon in the error message that
227 is given when verify = headers_syntax fails (if there are lots of them,
228 the message gets confusing).
230 PH/35 Change the default for dns_check_names_pattern to allow slashes within
231 names, as there are now some PTR records that contain slashes. This check
232 is only to protect against broken name servers that fall over on strange
233 characters, so the fact that it applies to all lookups doesn't matter.
235 PH/36 Now that the new test suite is complete, we can remove some of the
236 special code in Exim that was needed for the old test suite. For example,
237 sorting DNS records because real resolvers return them in an arbitrary
238 order. The new test suite's fake resolver always returns records in the
241 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
242 bounces) except when queue_only is set, to avoid logging races between
243 the different processes.
245 PH/38 Panic-die if .include specifies a non-absolute path.
247 PH/39 A tweak to the "H" retry rule from its user.
249 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
250 a label. They prevented compilation on older perls.
252 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
253 a warning to be raised on newish perls.
255 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
256 on queue. Changes to match documented behaviour of showing count of
257 messages matching specified criteria.
259 PH/40 Changed the default ident timeout from 30s to 5s.
261 PH/41 Added support for the use of login_cap features, on those BSD systems
262 that have them, for controlling the resources used by pipe deliveries.
264 PH/42 The content-scanning code uses fopen() to create files in which to put
265 message data. Previously it was not paying any attention to the mode of
266 the files. Exim runs with umask(0) because the rest of the code creates
267 files with open(), and sets the required mode explicitly. Thus, these
268 files were ending up world-writeable. This was not a big issue, because,
269 being within the spool directory, they were not world-accessible. I have
270 created a function called modefopen, which takes an additional mode
271 argument. It sets umask(777), creates the file, chmods it to the required
272 mode, then resets the umask. All the relevant calls to fopen() in the
273 content scanning code have been changed to use this function.
275 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
276 to 24 hours. This avoids potential overflow problems when processing G
277 and H retry rules. I suspect nobody ever tinkers with this value.
279 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
281 PH/45 When the plaintext authenticator is running as a client, the server's
282 challenges are checked to ensure they are valid base64 strings. By
283 default, the authentication attempt is cancelled if an invalid string is
284 received. Setting client_ignore_invalid_base64 true ignores these errors.
285 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
286 they are received. Thus, the responses can be made to depend on the
287 challenges. If an invalid string is ignored, an empty string is placed in
290 PH/46 Messages that are created by the autoreply transport now contains a
291 References: header, in accordance with RFCs 2822 and 3834.
293 PH/47 Added authenticated_sender_force to the smtp transport.
295 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
297 PH/49 Installed latest patch from the Sieve maintainer.
299 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
300 was also set, the appendfile transport was unnecessarily scanning a
301 directory of message files (e.g. for maildir delivery) to find the count
302 of files (along with the size), even though it did not need this
303 information. It now does the scan only if it needs to find either the
304 size of the count of files.
306 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
308 PH/52 Two bugs concerned with error handling when the smtp transport is
311 (i) Exim was not creating retry information for temporary errors given
312 for individual recipients after the DATA command when the smtp transport
313 was used in LMTP mode. This meant that they could be retried too
314 frequently, and not timed out correctly.
316 (ii) Exim was setting the flag that allows error details to be returned
317 for LMTP errors on RCPT commands, but not for LMTP errors for individual
318 recipients that were returned after the DATA command.
320 PH/53 This is related to PH/52, but is more general: for any failing address,
321 when detailed error information was permitted to be returned to the
322 sender, but the error was temporary, then after the final timeout, only
323 "retry timeout exceeded" was returned. Now it returns the full error as
324 well as "retry timeout exceeded".
326 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
327 do this, and (what is worse) MTAs that accept it.
329 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
330 will now be deprecated.
332 PH/56 New os.c-cygwin from the Cygwin maintainer.
334 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
335 formats (previously only available in exim formats via -bpr, -bpru,
336 and -bpra. Now also available in native and exiqgrep formats)
338 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
339 with very large, slow to parse queues
341 JJ/08 exipick: added ! as generic prefix to negate any criteria format
343 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
345 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
346 responses to authentication challenges, though it was showing the
347 challenges; (ii) I've removed the CR characters from the debug output for
350 PH/58 Allow for the insertion of a newline as well as a space when a string
351 is turned into more than one encoded-word during RFC 2047 encoding. The
352 Sieve code now uses this.
354 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
355 data_4xx, lost_connection, tls_required.
357 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
358 message was being sent as an SMTP response.
360 PH/61 Add -l and -k options to exicyclog.
362 PH/62 When verifying, if an address was redirected to one new address, so that
363 verification continued, and the new address failed or deferred after
364 having set something in $address_data, the value of $address_data was not
365 passed back to the ACL. This was different to the case when no
366 redirection occurred. The value is now passed back in both cases.
368 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
369 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
370 use login_cap.h, so on its own it isn't the distinguishing feature. The
371 new name refers directly to the setclassresources() function.
373 PH/65 Added configuration files for NetBSD3.
375 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
377 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
378 is preferred over IPv4.
380 PH/68 The bounce_return_message and bounce_return_body options were not being
381 honoured for bounces generated during the reception of non-SMTP messages.
382 In particular, this applied to messages rejected by the ACL. This bug has
383 been fixed. However, if bounce_return_message is true and bounce_return_
384 body is false, the headers that are returned for a non-SMTP message
385 include only those that have been read before the error was detected.
386 (In the case of an ACL rejection, they have all been read.)
388 PH/69 The HTML version of the specification is now built in a directory called
389 spec_html instead of spec.html, because the latter looks like a path with
390 a MIME-type, and this confuses some software.
392 PH/70 Catch two compiler warnings in sieve.c.
394 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
395 function verify_get_ident() calls ip_connect() to connect a socket, but
396 if the "connect()" function timed out, ip_connect() used to close the
397 socket. However, verify_get_ident() also closes the socket later, and in
398 between Exim writes to the log, which may get opened at this point. When
399 the socket was closed in ip_connect(), the log could get the same file
400 descriptor number as the socket. This naturally causes chaos. The fix is
401 not to close the socket in ip_connect(); the socket should be closed by
402 the function that creates it. There was only one place in the code where
403 this was missing, in the iplookup router, which I don't think anybody now
404 uses, but I've fixed it anyway.
406 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
407 well as to direct DNS lookups. Otherwise the handling of names in host
408 lists is inconsistent and therefore confusing.
414 PH/01 Two changes to the default runtime configuration:
416 (1) Move the checks for relay_from_hosts and authenticated clients from
417 after to before the (commented out) DNS black list checks.
419 (2) Add control=submission to the relay_from_hosts and authenticated
420 clients checks, on the grounds that messages accepted by these
421 statements are most likely to be submissions.
423 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
425 (1) Generate an error if the third argument for the ${prvs expansion is
428 (2) Treat a missing third argument of ${prvscheck as if it were an empty
431 (3) Reset the variables that are obtained from the first argument of
432 ${prvscheck and used in the second argument before leaving the code,
433 because their memory is reclaimed, so using them afterwards may do
436 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
437 one (it's much easier than Tom thought :-).
439 (5) Because of (4), we can now allow for the use of $prvscheck_result
440 inside the third argument.
442 PH/03 For some reason, the default setting of PATH when running a command from
443 a pipe transport was just "/usr/bin". I have changed it to
446 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
447 anything to be listed in the output from -bV.
449 PH/05 When a filter generated an autoreply, the entire To: header line was
450 quoted in the delivery log line, like this:
452 => >A.N.Other <ano@some.domain> <original@ddress> ...
454 This has been changed so that it extracts the operative address. There
455 may be more than one such address. If so, they are comma-separated, like
458 => >ano@some.domain,ona@other.domain <original@ddress> ...
460 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
461 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
462 not being looked up in the rDNS to get a host name, Exim was showing the
463 IP address twice in Received: lines, even though the IP addresses were
464 identical. For example:
466 Received: from [1.2.3.4] (helo=[1.2.3.4])
468 However, if the real host name was known, it was omitting the HELO data
469 if it matched the actual IP address. This has been tidied up so that it
470 doesn't show the same IP address twice.
472 PH/07 When both +timestamp and +memory debugging was on, the value given by
473 $tod_xxx expansions could be wrong, because the tod_stamp() function was
474 called by the debug printing, thereby overwriting the timestamp buffer.
475 Debugging no longer uses the tod_stamp() function when +timestamp is set.
477 PH/08 When the original message was included in an autoreply transport, it
478 always said "this is a copy of the message, including all the headers",
479 even if body_only or headers_only was set. It now gives an appropriate
482 PH/09 Applied a patch from the Sieve maintainer which:
484 o fixes some comments
485 o adds the (disabled) notify extension core
486 o adds some debug output for the result of if/elsif tests
487 o points to the current vacation draft in the documentation
488 and documents the missing references header update
492 o fixes a bug in processing the envelope test (when testing
493 multiple envelope elements, the last element determinted the
496 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
497 Electronic Mail") by including:
499 Auto-submitted: auto-generated
501 in the messages that it generates (bounce messages and others, such as
502 warnings). In the case of bounce messages for non-SMTP mesages, there was
503 also a typo: it was using "Auto_submitted" (underscore instead of
504 hyphen). Since every message generated by Exim is necessarily in response
505 to another message, thes have all been changed to:
507 Auto-Submitted: auto-replied
509 in accordance with these statements in the RFC:
511 The auto-replied keyword:
513 - SHOULD be used on messages sent in direct response to another
514 message by an automatic process,
516 - MUST NOT be used on manually-generated messages,
518 - MAY be used on Delivery Status Notifications (DSNs) and Message
519 Disposition Notifications (MDNs),
521 - MUST NOT be used on messages generated by automatic or periodic
522 processes, except for messages which are automatic responses to
525 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
526 to the default Received: header definition.
528 PH/12 Added log selector acl_warn_skipped (default on).
530 PH/13 After a successful wildlsearch lookup, discard the values of numeric
531 variables because (a) they are in the wrong storage pool and (b) even if
532 they were copied, it wouldn't work properly because of the caching.
534 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
535 checking when decoding. Apparently there are clients that generate
536 overlong encoded strings. Why am I not surprised?
538 PH/15 If the first argument of "${if match_address" was not empty, but did not
539 contain an "@" character, Exim crashed. Now it writes a panic log message
540 and treats the condition as false.
542 PH/16 In autoreply, treat an empty string for "once" the same as unset.
544 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
545 extension "envelope-auth". The code is finished and in agreement with
546 other implementations, but there is no documentation so far and in fact,
547 nobody wrote the draft yet. This extension is currently #undef'ed, thus
548 not changing the active code.
550 Print executed "if" and "elsif" statements when debugging is used. This
551 helps a great deal to understand what a filter does.
553 Document more things not specified clearly in RFC3028. I had all this
554 sorted out, when out of a sudden new issues came to my mind. Oops."
556 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
559 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
560 canonical form (as documented). However, after a host name lookup from
561 the IP address, check_host() was doing a simple string comparison with
562 addresses acquired from the DNS when checking that the found name did
563 have the original IP as one of its addresses. Since any found IPv6
564 addresses are likely to be in abbreviated form, the comparison could
565 fail. Luckily, there already exists a function for doing the comparison
566 by converting both addresses to binary, so now that is used instead of
569 PH/20 There was another similar case to PH/19, when a complete host name was
570 given in a host list; looking up its IP address could give an abbreviated
571 form, whereas the current host's name might or might not be abbreviated.
572 The same fix has been applied.
578 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
579 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
582 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
583 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
585 PH/03 Typo: missing ".o" in src/pcre/Makefile.
587 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
588 header line, restrict the check to what is listed in RFCs 2369 and 2929.
589 Also, for "Auto-Submitted", treat anything other than "no" as
590 non-personal, in accordance with RFC 3834. (Previously it treated
591 anything starting "auto-" as non-personal.)
593 TF/01 The control=submission/name=... option had a problem with syntax
594 errors if the name included a slash character. The /name= option
595 now slurps the rest of the string, so it can include any characters
596 but it must come last in the list of options (after /sender_retain
599 PH/05 Some modifications to the interface to the fake nameserver for the new
607 TK/01 Added the "success_on_redirect" address verification option. See
608 NewStuff for rationale and an example.
610 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
612 PH/02 Patch to exigrep to allow it to work on syslog lines.
614 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
615 fread() to skip over the body file's header line, because in Cygwin the
616 header line is locked and is inaccessible.
618 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
619 co-exist for some time) to make it clear that it is the Exim ID that is
620 referenced, not the Message-ID: header line.
622 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
623 string_format() function, because snprintf() does not exist on all
626 PH/06 The use of forbid_filter_existstest now also locks out the use of the
627 ${stat: expansion item.
629 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
630 protocol synchronization error", to keep the pedants happy.
632 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
633 well as for IRIX systems, when gcc is being used. See the host.c source
636 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
638 PH/10 Named domain lists were not working if used in a queue_smtp_domains
641 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
642 transport and to the smtp transport in LMTP mode.
644 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
646 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
647 run a filter in a subprocess. This could lead to confusion in subsequent
648 lookups in the parent process. There should also be a search_tidyup() at
649 the end of the subprocess.
651 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
652 only if the host matched helo_try_verify_hosts, which caused the
653 verification to occur when the EHLO/HELO command was issued. The ACL just
654 tested the remembered result. Now, if a previous verification attempt has
655 not happened, "verify = helo" does it there and then.
657 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
659 TK/03 Fix log output including CR from clamd.
661 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
662 exist provoked a memory error which could cause a segfault.
664 PH/15 Installed PCRE 6.2
666 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
668 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
669 of the problem. Specifically, suggested +O2 rather than +O1 for the
672 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
674 PH/20 If a delivery was routed to a non-standard port by means of an SRV
675 record, the port was not correctly logged when the outgoing_port log
676 selector was set (it logged the transort's default port).
678 PH/21 Added support for host-specific ports to manualroute, queryprogram,
679 fallback_hosts, and "hosts" in the smtp transport.
681 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
682 host errors such as "Connection refused".
684 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
685 authentication with radiusclient 0.4.9:
687 - Error returned from rc_read_config was caught wrongly
688 - Username/password not passed on to radius server due to wrong length.
690 The presumption is that some radiusclient API changes for 4.51/PH/17
691 were not taken care of correctly. The code is still untested by me (my
692 Linux distribution still has 0.3.2 of radiusclient), but it was
693 contributed by a Radius user.
695 PH/24 When doing a callout, the value of $domain wasn't set correctly when
696 expanding the "port" option of the smtp transport.
698 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
699 while reading a MIME header. Thanks to Tom Hughes for a patch.
701 PH/24 Include config.h inside local_scan.h so that configuration settings are
704 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
705 that in an ACL for RCPT (for example), you can examine exactly what was
708 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
709 commands, but it was not correctly comparing the address with the actual
710 client host address. Thus, it would show the EHLO address in Received:
711 header lines when this was not necessary.
713 PH/27 Added the % operator to ${eval:}.
715 PH/28 Exim tries to create and chdir to its spool directory when it starts;
716 it should be ignoring failures (because with -C, for example, it has lost
717 privilege). It wasn't ignoring creation failures other than "already
720 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
723 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
724 because an input line was too long, either on its own, or by virtue of
725 too many continuations, the temporary file was not being removed, and the
726 return code was incorrect.
728 PH/31 Missing "BOOL" in function definition in filtertest.c.
730 PH/32 Applied Sieve patches from the maintainer.
732 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
734 PH/33 Added "verify = not_blind".
736 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
737 Local/Makefile (with some defaults set). These are used in built scripts
738 such as exicyclog, but they have never been used in the exim_install
739 script (though there are many overriding facilities there). I have
740 arranged that the exim_install script now takes note of these two
743 PH/35 Installed configuration files for Dragonfly.
745 PH/36 When a locally submitted message by a trusted user did not contain a
746 From: header, and the sender address was obtained from -f or from an SMTP
747 MAIL command, and the trusted user did not use -F to supply a sender
748 name, $originator_name was incorrectly used when constructing a From:
749 header. Furthermore, $originator_name was used for submission mode
750 messages from external hosts without From: headers in a similar way,
751 which is clearly wrong.
753 PH/37 Added control=suppress_local_fixups.
755 PH/38 When log_selector = +received_sender was set, and the addition of the
756 sender made the log line's construction buffer exactly full, or one byte
757 less than full, an overflow happened when the terminating "\n" was
760 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
761 when the result of a list match is failure because a DNS lookup failed.
763 PH/40 RM_COMMAND is now used in the building process.
765 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
766 the "build-* directories that it finds.
768 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
769 address in a domain literal was a prefix of an interface address.
771 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
772 when verifying a sender address, unless rewrite_headers is false.
774 PH/44 Wrote a long comment about why errors_to addresses are verified as
775 recipients, not senders.
777 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
778 the ratelimit ACL was added.
780 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
782 PH/46 Added extra information about PostgreSQL errors to the error string.
784 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
785 avoiding the need to install special zones in a real server. This is
786 backwards compatible; if it can't find the fake resolver, it drops back.
787 Thus, both old and new test suites can be run.
789 TF/02 Added util/ratelimit.pl
791 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
792 clock is set back in time.
794 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
795 Candler <B.Candler@pobox.com>.
797 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
798 OK for addresses that are the result of redirections.
800 PH/48 A number of further additions for the benefit of the new test suite,
801 including a fake gethostbyname() that interfaces to the fake DNS resolver
804 TF/06 The fix for widen_domains has also been applied to qualify_single and
805 search_parents which are the other dnslookup options that can cause
808 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
811 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
813 TF/07 Exim produced the error message "an SRV record indicated no SMTP
814 service" if it encountered an MX record with an empty target hostname.
815 The message is now "an MX or SRV record indicated no SMTP service".
817 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
818 if the DNS of the sending site is misconfigured. This is quite a
819 common situation. This change restores the behaviour of treating a
820 helo verification defer as a failure.
822 PH/51 If self=fail was set on a router, the bounce message did not include the
823 actual error message.
829 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
831 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
832 command itself ended in error, the underlying message about the transport
833 filter timeout was being overwritten with the pipe command error. Now the
834 underlying error message should be appended to the second error message.
836 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
838 PH/02 Reduce the amount of output that "make" produces by default. Full output
839 can still be requested.
841 PH/03 The warning log line about a condition test deferring for a "warn" verb
842 was being output only once per connection, rather than after each
843 occurrence (because it was using the same function as for successful
844 "warn" verbs). This seems wrong, so I have changed it.
846 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
847 it should not have, which might have caused a crash in the right
848 circumstances, but probably never did.
850 PH/04 Installed a modified version of Tony Finch's patch to make submission
851 mode fix the return path as well as the Sender: header line, and to
852 add a /name= option so that you can make the user's friendly name appear
855 TF/03 Added the control = fakedefer ACL modifier.
857 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
858 Mark Lowes for thorough testing.
860 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
862 TK/03 Merged latest SRS patch from Miles Wilton.
864 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
865 with the definition in sysexits.h (which is #included earlier).
866 Fortunately, Exim does not actually use EX_OK. The code used to try to
867 preserve the sysexits.h value, by assumimg that macro definitions were
868 scanned for macro replacements. I have been disabused of this notion,
869 so now the code just undefines EX_OK before #including unistd.h.
871 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
872 in the smtp transport. When a block could not be written in a single
873 write() function, the timeout was being re-applied to each part-write.
874 This seems wrong - if the receiver was accepting one byte at a time it
875 would take for ever. The timeout is now adjusted when this happens. It
876 doesn't have to be particularly precise.
878 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
879 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
881 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
882 without a domain if the check to <postmaster@domain> fails.
884 SC/01 Eximstats: added -xls and the ability to specify output files
885 (patch written by Frank Heydlauf).
887 SC/02 Eximstats: use FileHandles for outputing results.
889 SC/03 Eximstats: allow any combination of xls, txt, and html output.
891 SC/04 Eximstats: fixed display of large numbers with -nvr option
893 SC/05 Eximstats: fixed merging of reports with empty tables.
895 SC/06 Eximstats: added the -include_original_destination flag
897 SC/07 Eximstats: removed tabs and trailing whitespace.
899 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
901 TK/06 MBOX spool code: Add real "From " MBOX separator line
902 so the .eml file is really in mbox format (even though
903 most programs do not really care). Patch from Alex Miller.
905 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
906 The latter is generated from $received_to and is only set if the
907 message has one envelope recipient. SA can use these headers,
908 obviously out-of-the-box. Patch from Alex Miller.
910 PH/08 The ${def test on a variable was returning false if the variable's
911 value was "0", contrary to what the specification has always said!
912 The result should be true unless the variable is empty.
914 PH/09 The syntax error of a character other than { following "${if
915 def:variable_name" (after optional whitespace) was not being diagnosed.
916 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
917 accidental colon was present, for example, could give incorrect results.
919 PH/10 Tidied the code in a number of places where the st_size field of a stat()
920 result is used (not including appendfile, where other changes are about
923 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
924 This involved changing a lot of size variables from int to off_t. It
925 should work with maildirs and everything.
927 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
928 spamd dying while we are connected to it.
930 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
933 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
934 function for BDB 4.3.
936 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
938 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
939 the sources for PCRE have been re-arranged and more files are now
942 PH/15 The code I had for printing potentially long long variables in PH/11
943 above was not the best (it lost precision). The length of off_t variables
944 is now inspected at build time, and an appropriate printing format (%ld
945 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
946 to be "long long int" or "long int". This is needed for the internal
947 formatting function string_vformat().
949 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
950 the configuration file to be ":syslog", then the script "guesses" where
951 the logs files are, rather than using the compiled in default. In our
952 case the guess is not the same as the compiled default, so the script
953 suddenly stopped working when I started to use syslog. The patch checks
954 to see if log_file_path is "". If so, it attempts to read it from exim
955 with no configuration file to get the compiled in version, before it
956 falls back to the previous guessing code."
958 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
959 implementing BATV in an Exim configuration. See NewStuff for the gory
962 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
963 Makefile that are specific to HP-UX.
965 PH/18 If the "use_postmaster" option was set for a recipient callout together
966 with the "random" option, the postmaster address was used as the MAIL
967 FROM address for the random test, but not for the subsequent recipient
968 test. It is now used for both.
970 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
971 patch removes a few documentation additions to RFC 3028, because the
972 latest draft now contains them. It adds the new en;ascii-case comparator
973 and a new error check for 8bit text in MIME parts. Comparator and
974 require names are now matched exactly. I enabled the subaddress
975 extension, but it is not well tested yet (read: it works for me)."
977 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
978 rework some of the code of TK/09 above to avoid the hardwired use of
979 "%lld" and "long long". Replaced the call to snprintf() with a call to
982 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
983 records point to non-existent hosts", "retry timeout exceeded", and
984 "retry time not reached for any host after a long failure period".
986 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
987 experimental DomainKeys support:
989 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
990 (2) On an error such as an illegally used "control", the wrong name for
991 the control was given.
993 These problems did NOT occur unless DomainKeys support was compiled.
995 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
997 PH/24 Added ${if match_ip condition.
999 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1000 databases so that it will be absolutely obvious if a crash occurs in the
1001 DB library. This is a regular occurrence (often caused by mis-matched
1004 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1005 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1006 picked up on a user's system that detects such things. There doesn't seem
1007 to be a gcc warning option for this - only an attribute that has to be
1008 put on the function's prototype. It seems that in Fedora Core 4 they have
1009 set this on a number of new functions. No doubt there will be more in due
1012 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1013 specify a transport. However, if an address that was verified by such a
1014 router was the subject of a callout, Exim crashed because it tried to
1015 read the rcpt_include_affixes from the non-existent transport. Now it
1016 just assumes that the setting of that option is false. This bug was
1017 introduced by 4.51/PH/31.
1019 PH/28 Changed -d+all to exclude +memory, because that information is very
1020 rarely of interest, but it makes the output a lot bigger. People tend to
1021 do -d+all out of habit.
1023 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1024 code in os-type was giving problems when libc.so lives in lib64, like on
1027 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1028 aren't the modern standard, and it seems that some systems' include files
1029 don't always have them. Exim was already checking for some of the newer
1030 ones like T_AAAA, and defining it itself. I've added checks for all the
1031 record types that Exim uses.
1033 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1034 not automatically generating a new one, as it is supposed to. This
1035 prevented TLS from working. If the file did exist, but contained invalid
1036 data, a new version was generated, as expected. It was only the case of a
1037 non-existent file that was broken.
1039 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1040 with a change in libdomainkeys > 0.64.
1042 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1043 from DNS. If the selector record carries the flag, it now has
1044 precedence over the domain-wide flag.
1046 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1048 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1049 the use of a port name that isn't defined in /etc/services) occurred, the
1050 message was deferred as in a normal delivery, and thus remained on the
1051 spool, instead of being failed because of the mua_wrapper setting. This
1052 is now fixed, and I tidied up some of the mua_wrapper messages at the
1055 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1056 the messages in a hash of arrays rather than using individual hashes.
1057 This is a bit cleaner and results in dramatic memory savings, albeit
1058 at a slight CPU cost.
1060 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1061 as requested by Marc Sherman.
1063 SC/10 Eximstats: added histograms for user specified patterns as requested
1066 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1068 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1069 fopen() in the content-scanning modules that did not already have it.
1075 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1076 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1078 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1080 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1082 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1083 to test Sieve filters that use "vacation".
1085 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1086 that changes the way the GnuTLS parameters are stored in the cache file.
1087 The new format can be generated externally. For backward compatibility,
1088 if the data in the cache doesn't make sense, Exim assumes it has read an
1089 old-format file, and it generates new data and writes a new file. This
1090 means that you can't go back to an older release without removing the
1093 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1094 work if there are any delivery delays because "one_time" forces the
1095 parent to be marked "delivered", so its unseen clone is never tried
1096 again. For this reason, Exim now forbids the simultaneous setting of
1099 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1100 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1101 entry if you want to know the details. The fix, however, neglected to
1102 consider the case when local delivery batching is involved. The test for
1103 "previously delivered" was not happening when checking to see if an
1104 address could be batched with a previous (undelivered) one; under
1105 certain circumstances this could lead to multiple deliveries to the same
1108 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1109 in its include files, and this causes problems building Exim.
1111 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1112 header_syntax) but Exim was just ignoring anything given after a slash.
1113 In particular, this caused confusion with an attempt to use "verify =
1114 reverse_host_lookup/defer_ok". An error is now given when options are
1115 supplied for verify items that do not have them. (Maybe reverse_host_
1116 lookup should have a defer_ok option, but that's a different point.)
1118 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1119 defined by RFC 821) to 2048, because there were problems with some AUTH
1120 commands, and RFC 1869 says the size should be increased for extended
1121 SMTP commands that take arguments.
1123 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1126 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1127 "unknown" error; now it says that the functionality isn't in the binary.
1129 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1130 an address' error message when a string expansion fails (syntax or
1131 whatever). Otherwise the password may appear in the log. Following change
1132 PH/42 below, there is no longer a chance of it appearing in a bounce
1135 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1137 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1138 of its addresses was ever tried. (Bugzilla bug #2.)
1140 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1141 the result incorrectly in the debug output. (It correctly added a newline
1142 to what was transported.)
1144 TF/01 Added $received_time.
1146 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1147 commented out examples of how to interface to a virus scanner and to
1148 SpamAssassin. Also added commented examples of av_scanner and
1149 spamd_address settings.
1151 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1152 and controls are allowed in which ACLs. There were a couple of minor
1153 errors. Some of the entries in the conditions table (which is a table of
1154 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1155 negation of where the condition IS allowed.
1157 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1159 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1160 header file does not have a version number, so I've had to invent a new
1161 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1162 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1163 radiusclient), but it was contributed by a Radius user.
1165 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1166 files or directories, for OpenSSL.
1168 PH/19 When an Exim process that is running as root has to create an Exim log
1169 file, it does so in a subprocess that runs as exim:exim so as to get the
1170 ownership right at creation (otherwise, other Exim processes might see
1171 the file with the wrong ownership). There was no test for failure of this
1172 fork() call, which would lead to the process getting stuck as it waited
1173 for a non-existent subprocess. Forks do occasionally fail when resources
1174 run out. I reviewed all the other calls to fork(); they all seem to check
1177 PH/20 When checking for unexpected SMTP input at connect time (before writing
1178 the banner), Exim was not dealing correctly with a non-positive return
1179 from the read() function. If the client had disconnected by this time,
1180 the result was a log entry for a synchronization error with an empty
1181 string after "input=" when read() returned zero. If read() returned -1
1182 (an event I could not check), uninitialized data bytes were printed.
1183 There were reports of junk text (parts of files, etc) appearing after
1186 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1188 PH/22 Added support for macro redefinition, and (re)definition in between
1189 driver and ACL definitions.
1191 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1192 forgetting to use the resulting value; it was using the unexpanded value.
1194 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1195 hadn't been configured. The fix is from Juergen Kreileder, who
1196 understands it better than I do:
1198 "Here's what I see happening with three configured cyrus_sasl
1199 authenticators configured (plain, login, cram-md5):
1201 On startup auth_cyrus_sasl_init() gets called for each of these.
1202 This means three calls to sasl_listmech() without a specified mech_list.
1203 => SASL tests which mechs of all available mechs actually work
1204 => three warnings about OTP not working
1205 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1207 With the patch, sasl_listmech() also gets called three times. But now
1208 SASL's mech_list option is set to the server_mech specified in the the
1209 authenticator. Or in other words, the answer from sasl_listmech()
1210 gets limited to just the mech you're testing for (which is different
1212 => the return list contains just 'plain' or 'login', 'cram-md5' or
1213 nothing depending on the value of ob->server_mech.
1215 I've just tested the patch: Authentication still works fine,
1216 unavailable mechs specified in the exim configuration are still
1217 caught, and the auth.log warnings about OTP are gone."
1219 PH/25 When debugging is enabled, the contents of the command line are added
1220 to the debugging output, even when log_selector=+arguments is not
1223 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1224 answer is "GNU", and only if the return is "GNU/something" is the answer
1227 PH/27 $acl_verify_message is now set immediately after the failure of a
1228 verification in an ACL, and so is available in subsequent modifiers. In
1229 particular, the message can be preserved by coding like this:
1231 warn !verify = sender
1232 set acl_m0 = $acl_verify_message
1234 Previously, $acl_verify_message was set only while expanding "message"
1235 and "log_message" when a very denied access.
1237 PH/28 Modified OS/os.c-Linux with
1239 -#ifndef OS_LOAD_AVERAGE
1240 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1242 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1243 nomenclature these days.)
1245 PH/29 Installed patch from the Sieve maintainer that adds the options
1246 sieve_useraddress and sieve_subaddress to the redirect router.
1248 PH/30 In these circumstances:
1249 . Two addresses routed to the same list of hosts;
1250 . First host does not offer TLS;
1251 . First host accepts first address;
1252 . First host gives temporary error to second address;
1253 . Second host offers TLS and a TLS session is established;
1254 . Second host accepts second address.
1255 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1256 and peerdn, if requested) that were in fact used only for the second
1259 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1260 attention to any local part prefix or suffix that was matched by the
1261 router that accepted the address. It now behaves in the same way as it
1262 does for delivery: the affixes are removed from the local part unless
1263 rcpt_include_affixes is set on the transport.
1265 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1266 timeout during the DATA phase of an incoming message.
1268 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1269 applied a patch sanctioned by the Sieve maintainer.
1271 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1272 the uid or gid is negative. A case of a negative gid caused this to be
1273 noticed. The fix allows for either to be negative.
1275 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1276 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1279 PH/36 The change PH/12 above was broken. Fixed it.
1281 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1282 the grounds that routing the same address twice would always produce the
1283 same answer. This might have been true once, but it is certainly no
1284 longer true now. Routing a child address may depend on the previous
1285 routing that produced that child. Some complicated redirection strategies
1286 went wrong when messages had multiple recipients, and made Exim's
1287 behaviour dependent on the order in which the addresses were given.
1289 I have moved the duplicate checking until after the routing is complete.
1290 Exim scans the addresses that are assigned to local and remote
1291 transports, and removes any duplicates. This means that more work will be
1292 done, as duplicates will always all be routed, but duplicates are
1293 presumably rare, so I don't expect this is of any significance.
1295 For deliveries to pipes, files, and autoreplies, the duplicate checking
1296 still happens during the routing process, since they are not going to be
1299 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1300 It corrects a timeout issue with spamd. This is Ian's comment: "The
1301 background is that sometimes spamd either never reads data from a
1302 connection it has accepted, or it never writes response data. The exiscan
1303 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1304 blindly assumes that writes won't block so it may never time out."
1306 PH/39 Allow G after quota size as well as K and M.
1308 PH/40 The value set for $authenticated_id in an authenticator may not contain
1309 binary zeroes or newlines because the value is written to log lines and
1310 to spool files. There was no check on this. Now the value is run through
1311 the string_printing() function so that such characters are converted to
1312 printable escape sequences.
1314 PH/41 $message_linecount is a new variable that contains the total number of
1315 lines in the message. Compare $body_linecount, which is the count for the
1318 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1319 bounce and delay warning messages, except in certain special cases, which
1322 (a) An SMTP error message from a remote host;
1323 (b) A message specified in a :fail: redirection;
1324 (c) A message specified in a "fail" command in a system filter;
1325 (d) A message specified in a FAIL return from the queryprogram router;
1326 (e) A message specified by the cannot_route_message router option.
1328 In these cases only, Exim does include the error details in bounce and
1329 warning messages. There are also a few cases where bland messages such
1330 as "unrouteable address" or "local delivery error" are given.
1332 PH/43 $value is now also set for the "else" part of a ${run expansion.
1334 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1335 being worked on, but at least Exim now implements the latest version to
1338 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1339 process to complete was treated as a delivery failure, a timeout while
1340 writing the message to the pipe was logged, but erroneously treated as a
1341 successful delivery. Such timeouts include transport filter timeouts. For
1342 consistency with the overall process timeout, these timeouts are now
1343 treated as errors, giving rise to delivery failures by default. However,
1344 there is now a new Boolean option for the pipe transport called
1345 timeout_defer, which, if set TRUE, converts the failures into defers for
1346 both kinds of timeout. A transport filter timeout is now identified in
1349 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1350 systems where "make" and "gmake" are different, calling "gmake" at top
1351 level broke things. I've arranged for the value of $(MAKE) to be passed
1352 from the Makefile to this script so that it can call the same version of
1356 A note about Exim versions 4.44 and 4.50
1357 ----------------------------------------
1359 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1360 changes of various kinds. As a consequence, a big documentation update was
1361 needed. This delayed the release for rather longer than seemed good, especially
1362 in the light of a couple of (minor) security issues. Therefore, the changes
1363 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1364 release. So 4.44 and 4.50 are in effect two different branches that both start
1367 I have left the 4.50 change log unchanged; it contains all the changes since
1368 4.43. The change log for 4.44 is below; many of its items are identical to
1369 those for 4.50. This seems to be the most sensible way to preserve the
1370 historical information.
1376 1. Minor wording change to the doc/README.SIEVE file.
1378 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1379 computation of the current number of files was incorrect.
1381 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1382 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1385 4. Give more explanation in the error message when the command for a transport
1386 filter fails to execute.
1388 5. There are several places where Exim runs a non-Exim command in a
1389 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1390 was being done only for the command run by the queryprogram router. It is
1391 now done for all such subprocesses. The other cases are: ${run, transport
1392 filters, and the commands run by the lmtp and pipe transports.
1394 6. Added CONFIGURE_GROUP build-time option.
1396 7. Some older OS have a limit of 256 on the maximum number of file
1397 descriptors. Exim was using setrlimit() to set 1000 as a large value
1398 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1399 systems. I've change it so that if it can't get 1000, it tries for 256.
1401 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1402 was an oversight, and furthermore, ever since the addition of extra
1403 controls (e.g. 4.43/32), the checks on when to allow different forms of
1404 "control" were broken. There should now be diagnostics for all cases when a
1405 control that does not make sense is encountered.
1407 9. Added the /retain_sender option to "control=submission".
1409 10. $recipients is now available in the predata ACL (oversight).
1411 11. Tidy the search cache before the fork to do a delivery from a message
1412 received from the command line. Otherwise the child will trigger a lookup
1413 failure and thereby defer the delivery if it tries to use (for example) a
1414 cached ldap connection that the parent has called unbind on.
1416 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1417 of $address_data from the recipient verification was clobbered by the
1418 sender verification.
1420 13. The value of address_data from a sender verification is now available in
1421 $sender_address_data in subsequent conditions in the ACL statement.
1423 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1425 15. Added a new option "connect=<time>" to callout options, to set a different
1428 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1429 was its contents. (It was OK if the option was not defined at all.)
1431 17. A "Completed" log line is now written for messages that are removed from
1432 the spool by the -Mrm option.
1434 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1435 information about exactly what failed.
1437 19. Added -dd to debug only the daemon process.
1439 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1440 handles timeouts, both on the server side and network timeouts. Renamed the
1441 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1443 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1444 It is now set to "smtps".
1446 22. $host_address is now set to the target address during the checking of
1447 ignore_target_hosts.
1449 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1450 being passed; this would have caused $sender_host_name to have been used if
1451 matching the list had actually called for a host name (not very likely,
1452 since this list is usually IP addresses). A host name is now passed as
1455 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1456 code that specifies a non-restarting handler (typically sigaction() in
1457 modern systems) in an attempt to fix a rare and obscure crash bug.
1459 25. Narrowed the window for a race in the daemon that could cause it to ignore
1460 SIGCHLD signals. This is not a major problem, because they are used only to
1461 wake it up if nothing else does.
1463 26. A malformed maildirsize file could cause Exim to calculate negative values
1464 for the mailbox size or file count. Odd effects could occur as a result.
1465 The maildirsize information is now recalculated if the size or filecount
1468 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1469 support for a long time. Removed HAVE_SYS_VFS_H.
1471 28. Installed the latest version of exipick from John Jetmore.
1473 29. In an address list, if the pattern was not a regular expression, an empty
1474 subject address (from a bounce message) matched only if the pattern was an
1475 empty string. Non-empty patterns were not even tested. This was the wrong
1476 because it is perfectly reasonable to use an empty address as part of a
1477 database query. An empty address is now tested by patterns that are
1478 lookups. However, all the other forms of pattern expect the subject to
1479 contain a local part and a domain, and therefore, for them, an empty
1480 address still always fails if the pattern is not itself empty.
1482 30. Exim went into a mad DNS loop when attempting to do a callout where the
1483 host was specified on an smtp transport, and looking it up yielded more
1484 than one IP address.
1486 31. Re-factored the code for checking spool and log partition space into a
1487 function that finds that data and another that does the check. The former
1488 is then used to implement four new variables: $spool_space, $log_space,
1489 $spool_inodes, and $log_inodes.
1491 32. The RFC2047 encoding function was originally intended for short strings
1492 such as real names; it was not keeping to the 75-character limit for
1493 encoded words that the RFC imposes. It now respects the limit, and
1494 generates multiple encoded words if necessary. To be on the safe side, I
1495 have increased the buffer size for the ${rfc2047: expansion operator from
1498 33. It is now permitted to omit both strings after an "if" condition; if the
1499 condition is true, the result is "true". As before, when the second string
1500 is omitted, a false condition yields an empty string. This makes it less
1501 cumbersome to write custom ACL and router conditions.
1503 34. Failure to deliver a bounce message always caused it to be frozen, even if
1504 there was an errors_to setting on the router. The errors_to setting is now
1507 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1508 canonical form (fully expanded) before being placed in
1509 $sender_host_address.
1511 36. The table in the code that translates DNS record types into text (T_A to
1512 "A" for instance) was missing entries for NS and CNAME. It is just possible
1513 that this could have caused confusion if both these types were looked up
1514 for the same domain, because the text type is used as part of Exim's
1515 per-process caching. But the chance of anyone hitting this buglet seems
1518 37. The dnsdb lookup has been extended in a number of ways.
1520 (1) There is a new type, "zns", which walks up the domain tree until it
1521 finds some nameserver records. It should be used with care.
1523 (2) There is a new type, "mxh", which is like "mx" except that it returns
1524 just the host names, not the priorities.
1526 (3) It is now possible to give a list of domains (or IP addresses) to be
1527 looked up. The behaviour when one of the lookups defers can be
1528 controlled by a keyword.
1530 (4) It is now possible to specify the separator character for use when
1531 multiple records are returned.
1533 38. The dnslists ACL condition has been extended: it is now possible to supply
1534 a list of IP addresses and/or domains to be looked up in a particular DNS
1537 39. Added log_selector=+queue_time_overall.
1539 40. When running the queue in the test harness, wait just a tad after forking a
1540 delivery process, to get repeatability of debugging output.
1542 41. Include certificate and key file names in error message when GnuTLS fails
1543 to set them up, because the GnuTLS error message doesn't include the name
1544 of the failing file when there is a problem reading it.
1546 42. Allow both -bf and -bF in the same test run.
1548 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1550 44. The "Exiscan patch" is now merged into the mainline Exim source.
1552 45. Sometimes the final signoff response after QUIT could fail to get
1553 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1554 < 0 before doing a fflush(). This bug looks as though it goes back to the
1555 introduction of TLS in release 3.20, but "sometimes" must have been rare
1556 because the tests only now provoked it.
1558 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1559 (this can affect the format of dates).
1561 47. exim_tidydb, when checking for the continued existence of a message for
1562 which it has found a message-specific retry record, was not finding
1563 messages that were in split spool directories. Consequently, it was
1564 deleting retry records that should have stayed in existence.
1566 48. Steve fixed some bugs in eximstats.
1568 49. The SPA authentication driver was not abandoning authentication and moving
1569 on to the next authenticator when an expansion was forced to fail,
1570 contradicting the general specification for all authenticators. Instead it
1571 was generating a temporary error. It now behaves as specified.
1573 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1574 (the order specifies the preference for clients). The order is now AES256,
1575 AES128, 3DES, ARCFOUR128.
1577 51. Small patch to Sieve code - explicitly set From: when generating an
1580 52. Exim crashed if a remote delivery caused a very long error message to be
1581 recorded - for instance if somebody sent an entire SpamAssassin report back
1582 as a large number of 550 error lines. This bug was coincidentally fixed by
1583 increasing the size of one of Exim's internal buffers (big_buffer) that
1584 happened as part of the Exiscan merge. However, to be on the safe side, I
1585 have made the code more robust (and fixed the comments that describe what
1588 53. Now that there can be additional text after "Completed" in log lines (if
1589 the queue_time_overall log selector is set), a one-byte patch to exigrep
1590 was needed to allow it to recognize "Completed" as not the last thing in
1593 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1594 patch that reportedly fixes this has been added. I am not expert enough to
1595 create a test for it. This is what the patch creator wrote:
1597 "I found a little strange behaviour of ldap code when working with
1598 Windows 2003 AD Domain, where users was placed in more than one
1599 Organization Units. When I tried to give exim partial DN, the exit code
1600 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1601 But simultaneously result of request was absolutely normal ldap result,
1602 so I produce this patch..."
1604 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1605 REFERENCE, so I have modified the code to exclude the patch when that macro
1608 55. Some experimental protocols are using DNS PTR records for new purposes. The
1609 keys for these records are domain names, not reversed IP addresses. The
1610 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1611 leaves it alone. Component reversal etc. now happens only for IP addresses.
1614 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1616 57. Double the size of the debug message buffer (to 2048) so that more of very
1617 long debug lines gets shown.
1619 58. The exicyclog utility now does better if the number of log files to keep
1620 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1622 59. Two changes related to the smtp_active_hostname option:
1624 (1) $smtp_active_hostname is now available as a variable.
1625 (2) The default for smtp_banner uses $smtp_active_hostname instead
1626 of $primary_hostname.
1628 60. The host_aton() function is supposed to be passed a string that is known
1629 to be a valid IP address. However, in the case of IPv6 addresses, it was
1630 not checking this. This is a hostage to fortune. Exim now panics and dies
1631 if the condition is not met. A case was found where this could be provoked
1632 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1633 components; fortuitously, this particular loophole had already been fixed
1634 by change 4.50/55 above.
1636 If there are any other similar loopholes, the new check in host_aton()
1637 itself should stop them being exploited. The report I received stated that
1638 data on the command line could provoke the exploit when Exim was running as
1639 exim, but did not say which command line option was involved. All I could
1640 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1641 running as the user.
1644 61. There was a buffer overflow vulnerability in the SPA authentication code
1645 (which came originally from the Samba project). I have added a test to the
1646 spa_base64_to_bits() function which I hope fixes it.
1649 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1650 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1652 63. The daemon start-up calls getloadavg() while still root for those OS that
1653 need the first call to be done as root, but it missed one case: when
1654 deliver_queue_load_max is set with deliver_drop_privilege. This is
1655 necessary for the benefit of the queue runner, because there is no re-exec
1656 when deliver_drop_privilege is set.
1658 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1659 This has been fixed.
1661 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1662 was in use, was not putting the data itself into the right store pool;
1663 consequently, it could be overwritten for a subsequent message in the same
1664 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1667 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1669 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1670 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1671 the function were treating the return as a boolean value, which happened to
1672 work because 0=false and not-0=true, but is not correct code.
1674 68. The host_aton() function was not handling scoped IPv6 addresses (those
1675 with, for example, "%eth0" on the end) correctly.
1677 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1678 negated items (that is, ~something) in unsigned ints. Some compilers
1679 apparently mutter when there is no cast.
1681 70. If an address verification called from an ACL failed, and did not produce a
1682 user-specific message (i.e. there was only a "system" message), nothing was
1683 put in $acl_verify_message. In this situation, it now puts the system
1686 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1687 session; change 4.31/43 added the unwanted input to the log line - except
1688 that it did not do this in the start of session case. It now does.
1690 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1691 This is wrong and can cause the other end to generate a synchronization
1692 error if it is another Exim or anything else that does the synchronization
1693 check. A QUIT command is no longer sent after a timeout.
1695 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1696 during host lookups.
1698 74. The defer_ok option of callout verification was not working if it was used
1699 when verifying addresses in header lines, that is, for this case:
1701 verify = header_sender/callout=defer_ok
1703 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1704 those file descriptors could be used for SMTP connections. If anything
1705 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1706 could be sent to the SMTP client, causing chaos. The daemon now opens
1707 stdin, stdout, and stderr to /dev/null when it puts itself into the
1710 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1711 log by default. The user can override this with suitable Perl magic.
1713 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1714 the log message when discard triggers, was not working for the DATA ACL or
1715 for the non-SMTP ACL.
1717 78. Error message wording change in sieve.c.
1719 79. If smtp_accept_max_per_host was set, the number of connections could be
1720 restricted to fewer than expected, because the daemon was trying to set up
1721 a new connection before checking whether the processes handling previous
1722 connections had finished. The check for completed processes is now done
1723 earlier. On busy systems, this bug wouldn't be noticed because something
1724 else would have woken the daemon, and it would have reaped the completed
1727 80. If a message was submitted locally by a user whose login name contained one
1728 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1729 It caused a spool format error. I have fixed the spool reading code. A
1730 related problem was that the "from" clause in the Received: line became
1731 illegal because of the space(s). It is now covered by ${quote_local_part.
1733 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1736 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1738 83. Patch from Sieve maintainer to fix unterminated string problem in
1739 "vacation" handling.
1741 84. Some minor changes to the Linux configuration files to help with other
1742 OS variants using glibc.
1744 85. One more patch for Sieve to update vacation handling to latest spec.
1747 ----------------------------------------------------
1748 See the note above about the 4.44 and 4.50 releases.
1749 ----------------------------------------------------
1755 1. Change 4.43/35 introduced a bug that caused file counts to be
1756 incorrectly computed when quota_filecount was set in an appendfile
1759 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1760 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1763 3. Give more explanation in the error message when the command for a transport
1764 filter fails to execute.
1766 4. There are several places where Exim runs a non-Exim command in a
1767 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1768 was being done only for the command run by the queryprogram router. It is
1769 now done for all such subprocesses. The other cases are: ${run, transport
1770 filters, and the commands run by the lmtp and pipe transports.
1772 5. Some older OS have a limit of 256 on the maximum number of file
1773 descriptors. Exim was using setrlimit() to set 1000 as a large value
1774 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1775 systems. I've change it so that if it can't get 1000, it tries for 256.
1777 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1778 was an oversight, and furthermore, ever since the addition of extra
1779 controls (e.g. 4.43/32), the checks on when to allow different forms of
1780 "control" were broken. There should now be diagnostics for all cases when a
1781 control that does not make sense is encountered.
1783 7. $recipients is now available in the predata ACL (oversight).
1785 8. Tidy the search cache before the fork to do a delivery from a message
1786 received from the command line. Otherwise the child will trigger a lookup
1787 failure and thereby defer the delivery if it tries to use (for example) a
1788 cached ldap connection that the parent has called unbind on.
1790 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1791 of $address_data from the recipient verification was clobbered by the
1792 sender verification.
1794 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1795 was its contents. (It was OK if the option was not defined at all.)
1797 11. A "Completed" log line is now written for messages that are removed from
1798 the spool by the -Mrm option.
1800 12. $host_address is now set to the target address during the checking of
1801 ignore_target_hosts.
1803 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1804 being passed; this would have caused $sender_host_name to have been used if
1805 matching the list had actually called for a host name (not very likely,
1806 since this list is usually IP addresses). A host name is now passed as
1809 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1810 code that specifies a non-restarting handler (typically sigaction() in
1811 modern systems) in an attempt to fix a rare and obscure crash bug.
1813 15. Narrowed the window for a race in the daemon that could cause it to ignore
1814 SIGCHLD signals. This is not a major problem, because they are used only to
1815 wake it up if nothing else does.
1817 16. A malformed maildirsize file could cause Exim to calculate negative values
1818 for the mailbox size or file count. Odd effects could occur as a result.
1819 The maildirsize information is now recalculated if the size or filecount
1822 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1823 support for a long time. Removed HAVE_SYS_VFS_H.
1825 18. Updated exipick to current release from John Jetmore.
1827 19. Allow an empty sender to be matched against a lookup in an address list.
1828 Previously the only cases considered were a regular expression, or an
1831 20. Exim went into a mad DNS lookup loop when doing a callout where the
1832 host was specified on the transport, if the DNS lookup yielded more than
1835 21. The RFC2047 encoding function was originally intended for short strings
1836 such as real names; it was not keeping to the 75-character limit for
1837 encoded words that the RFC imposes. It now respects the limit, and
1838 generates multiple encoded words if necessary. To be on the safe side, I
1839 have increased the buffer size for the ${rfc2047: expansion operator from
1842 22. Failure to deliver a bounce message always caused it to be frozen, even if
1843 there was an errors_to setting on the router. The errors_to setting is now
1846 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1847 canonical form (fully expanded) before being placed in
1848 $sender_host_address.
1850 24. Updated eximstats to version 1.33
1852 25. Include certificate and key file names in error message when GnuTLS fails
1853 to set them up, because the GnuTLS error message doesn't include the name
1854 of the failing file when there is a problem reading it.
1856 26. Expand error message when OpenSSL has problems setting up cert/key files.
1859 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1860 (this can affect the format of dates).
1862 28. exim_tidydb, when checking for the continued existence of a message for
1863 which it has found a message-specific retry record, was not finding
1864 messages that were in split spool directories. Consequently, it was
1865 deleting retry records that should have stayed in existence.
1867 29. eximstats updated to version 1.35
1868 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1869 1.35 - bugfix such that pie charts by volume are generated correctly
1871 30. The SPA authentication driver was not abandoning authentication and moving
1872 on to the next authenticator when an expansion was forced to fail,
1873 contradicting the general specification for all authenticators. Instead it
1874 was generating a temporary error. It now behaves as specified.
1876 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1877 (the order specifies the preference for clients). The order is now AES256,
1878 AES128, 3DES, ARCFOUR128.
1880 31. Small patch to Sieve code - explicitly set From: when generating an
1883 32. Exim crashed if a remote delivery caused a very long error message to be
1884 recorded - for instance if somebody sent an entire SpamAssassin report back
1885 as a large number of 550 error lines. This bug was coincidentally fixed by
1886 increasing the size of one of Exim's internal buffers (big_buffer) that
1887 happened as part of the Exiscan merge. However, to be on the safe side, I
1888 have made the code more robust (and fixed the comments that describe what
1891 33. Some experimental protocols are using DNS PTR records for new purposes. The
1892 keys for these records are domain names, not reversed IP addresses. The
1893 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1894 leaves it alone. Component reversal etc. now happens only for IP addresses.
1897 34. The host_aton() function is supposed to be passed a string that is known
1898 to be a valid IP address. However, in the case of IPv6 addresses, it was
1899 not checking this. This is a hostage to fortune. Exim now panics and dies
1900 if the condition is not met. A case was found where this could be provoked
1901 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1902 components; fortuitously, this particular loophole had already been fixed
1903 by change 4.50/55 or 4.44/33 above.
1905 If there are any other similar loopholes, the new check in host_aton()
1906 itself should stop them being exploited. The report I received stated that
1907 data on the command line could provoke the exploit when Exim was running as
1908 exim, but did not say which command line option was involved. All I could
1909 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1910 running as the user.
1913 35. There was a buffer overflow vulnerability in the SPA authentication code
1914 (which came originally from the Samba project). I have added a test to the
1915 spa_base64_to_bits() function which I hope fixes it.
1918 36. The daemon start-up calls getloadavg() while still root for those OS that
1919 need the first call to be done as root, but it missed one case: when
1920 deliver_queue_load_max is set with deliver_drop_privilege. This is
1921 necessary for the benefit of the queue runner, because there is no re-exec
1922 when deliver_drop_privilege is set.
1924 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1925 was in use, was not putting the data itself into the right store pool;
1926 consequently, it could be overwritten for a subsequent message in the same
1927 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1930 38. Sometimes the final signoff response after QUIT could fail to get
1931 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1932 < 0 before doing a fflush(). This bug looks as though it goes back to the
1933 introduction of TLS in release 3.20, but "sometimes" must have been rare
1934 because the tests only now provoked it.
1940 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1941 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1942 Now it returns an integer. A number of calls were still expecting a T/F
1943 return. Fortuitously, in all cases, the tests worked in OK situations,
1944 which is the norm. However, things would have gone wrong on any write
1945 failures on the smtp file descriptor. This function is used when sending
1946 messages over SMTP and also when doing verify callouts.
1948 2. When Exim is called to do synchronous delivery of a locally submitted
1949 message (the -odf or -odi options), it no longer closes stderr before doing
1952 3. Implemented the mua_wrapper option.
1954 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1956 5. Implemented the functions header_remove(), header_testname(),
1957 header_add_at_position(), and receive_remove_recipient(), and exported them
1960 6. If an ACL "warn" statement specified the addition of headers, Exim already
1961 inserted X-ACL-Warn: at the start if there was no header name. However, it
1962 was not making this test for the second and subsequent header lines if
1963 there were newlines in the string. This meant that an invalid header could
1964 be inserted if Exim was badly configured.
1966 7. Allow an ACL "warn" statement to add header lines at the start or after all
1967 the Received: headers, as well as at the end.
1969 8. Added the rcpt_4xx retry error code.
1971 9. Added postmaster_mailfrom=xxx to callout verification option.
1973 10. Added mailfrom=xxxx to the callout verification option, for verify=
1976 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1977 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1979 12. Inserted some casts to stop certain compilers warning when using pointer
1980 differences as field lengths or precisions in printf-type calls (mostly
1981 affecting debugging statements).
1983 13. Added optional readline() support for -be (dynamically loaded).
1985 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1986 same clock tick as a message's arrival, so that its received time was the
1987 same as the "first fail" time on the retry record, and that message
1988 remained on the queue past the ultimate address timeout, every queue runner
1989 would try a delivery (because it was past the ultimate address timeout) but
1990 after another failure, the ultimate address timeout, which should have then
1991 bounced the address, did not kick in. This was a "< instead of <=" error;
1992 in most cases the first failure would have been in the next clock tick
1993 after the received time, and all would be well.
1995 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1996 being recognized when the domain list was tested by the match_domain
1997 condition in an expansion string.
1999 16. Added the ${str2b64: operator.
2001 17. Exim was always calling setrlimit() to set a large limit for the number of
2002 processes, without checking whether the existing limit was already
2003 adequate. (It did check for the limit on file descriptors.) Furthermore,
2004 errors from getrlimit() and setrlimit() were being ignored. Now they are
2005 logged to the main and panic logs, but Exim does carry on, to try to do its
2006 job under whatever limits there are.
2008 18. Imported PCRE 5.0.
2010 19. Trivial typo in log message " temporarily refused connection" (the leading
2013 20. If the log selector return_path_on_delivery was set and an address was
2014 redirected to /dev/null, the delivery process crashed because it assumed
2015 that a return path would always be set for a "successful" delivery. In this
2016 case, the whole delivery is bypassed as an optimization, and therefore no
2019 21. Internal re-arrangement: the function for sending a challenge and reading
2020 a response while authentication was assuming a zero-terminated challenge
2021 string. It's now changed to take a pointer and a length, to allow for
2022 binary data in such strings.
2024 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2026 23. Exim was not respecting finduser_retries when seeking the login of the
2027 uid under which it was called; it was always trying 10 times. (The default
2028 setting of finduser_retries is zero.) Also, it was sleeping after the final
2029 failure, which is pointless.
2031 24. Implemented tls_on_connect_ports.
2033 25. Implemented acl_smtp_predata.
2035 26. If the domain in control=submission is set empty, Exim assumes that the
2036 authenticated id is a complete email address when it generates From: or
2037 Sender: header lines.
2039 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2040 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2041 chown and chgrp in /bin and hostname in /usr/bin.
2043 28. Exim was keeping the "process log" file open after each use, just as it
2044 does for the main log. This opens the possibility of it remaining open for
2045 long periods when the USR1 signal hits a daemon. Occasional processlog
2046 errors were reported, that could have been caused by this. Anyway, it seems
2047 much more sensible not to leave this file open at all, so that is what now
2050 29. The long-running daemon process does not normally write to the log once it
2051 has entered its main loop, and it closes the log before doing so. This is
2052 so that log files can straightforwardly be renamed and moved. However,
2053 there are a couple of unusual error situations where the daemon does write
2054 log entries, and I had neglected to close the log afterwards.
2056 30. The text of an SMTP error response that was received during a remote
2057 delivery was being truncated at 512 bytes. This is too short for some of
2058 the long messages that one sometimes sees. I've increased the limit to
2061 31. It is now possible to make retry rules that apply only when a message has a
2062 specific sender, in particular, an empty sender.
2064 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2065 it possible to be selective about when SMTP synchronization is enforced.
2067 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2069 32. Implemented hosts_connection_nolog.
2071 33. Added an ACL for QUIT.
2073 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2076 35. Added mailbox_size and mailbox_filecount to appendfile.
2078 36. Added control = no_multiline_responses to ACLs.
2080 37. There was a bug in the logic of the code that waits for the clock to tick
2081 in the case where the clock went backwards by a substantial amount such
2082 that the microsecond fraction of "now" was more than the microsecond
2083 fraction of "then" (but the whole seconds number was less).
2085 38. Added support for the libradius Radius client library this is found on
2086 FreeBSD (previously only the radiusclient library was supported).
2092 1. When certain lookups returned multiple values in the form name=value, the
2093 quoting of the values was not always being done properly. Specifically:
2094 (a) If the value started with a double quote, but contained no whitespace,
2096 (b) If the value contained whitespace other than a space character (i.e.
2097 tabs or newlines or carriage returns) it was not quoted.
2098 This fix has been applied to the mysql and pgsql lookups by writing a
2099 separate quoting function and calling it from the lookup code. The fix
2100 should probably also be applied to nisplus, ibase and oracle lookups, but
2101 since I cannot test any of those, I have not disturbed their existing code.
2103 2. A hit in the callout cache for a specific address caused a log line with no
2104 reason for rejecting RCPT. Now it says "Previous (cached) callout
2105 verification failure".
2107 3. There was an off-by-one bug in the queryprogram router. An over-long
2108 return line was truncated at 256 instead of 255 characters, thereby
2109 overflowing its buffer with the terminating zero. As well as fixing this, I
2110 have increased the buffer size to 1024 (and made a note to document this).
2112 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2113 when Exim is waiting for an SMTP response from a remote server, Exim
2114 restarts its select() call on the socket, thereby resetting its timeout.
2115 This is not a problem when such interrupts are rare. Somebody set up a cron
2116 job to run exiwhat every 2 minutes, which is less than the normal select()
2117 timeout (5 or 10 minutes). This meant that the select() timeout never
2118 kicked in because it was always reset. I have fixed this by comparing the
2119 time when an interrupt arrives with the time at the start of the first call
2120 to select(). If more time than the timeout has elapsed, the interrupt is
2121 treated as a timeout.
2123 5. Some internal re-factoring in preparation for the addition of Sieve
2124 extensions (by MH). In particular, the "personal" test is moved to a
2125 separate function, and given an option for scanning Cc: and Bcc: (which is
2126 not set for Exim filters).
2128 6. When Exim created an email address using the login of the caller as the
2129 local part (e.g. when creating a From: or Sender: header line), it was not
2130 quoting the local part when it contained special characters such as @.
2132 7. Installed new OpenBSD configuration files.
2134 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2135 try to make them clearer.
2137 9. Callout options, other than the timeout value, were being ignored when
2138 verifying sender addresses in header lines. For example, when using
2140 verify = header_sender/callout=no_cache
2142 the cache was (incorrectly) being used.
2144 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2145 only the Cygwin environment.
2147 11. When return_path_on_delivery was set as a log selector, if different remote
2148 addresses in the same message used different return paths and parallel
2149 remote delivery occurred, the wrong values would sometimes be logged.
2150 (Whenever a remote delivery process finished, the return path value from
2151 the most recently started remote delivery process was logged.)
2153 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2154 lines when AUTH and/or TLS are in use. This is the "received protocol"
2155 field. Exim used to use "asmtp" for authenticated SMTP, without any
2156 indication (in the protocol name) for TLS use. Now it follows the RFC and
2157 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2158 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2159 names appear in log lines as well as in Received: header lines.
2161 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2162 extensions, and comparison tests, and to fix some bugs.
2164 14. Changes to the "personal" filter test:
2166 (1) The test was buggy in that it was just doing the equivalent of
2167 "contains" tests on header lines. For example, if a user's address was
2168 anne@some.where, the "personal" test would incorrectly be true for
2170 To: susanne@some.where
2172 This test is now done by extracting each address from the header in turn,
2173 and checking the entire address. Other tests that are part of "personal"
2174 are now done using regular expressions (for example, to check local parts
2175 of addresses in From: header lines).
2177 (2) The list of non-personal local parts in From: addresses has been
2178 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2179 taken from the Sieve specification recommendations.
2181 (3) If the message contains any header line starting with "List-" it is
2182 treated as non-personal.
2184 (4) The test for "circular" in the Subject: header line has been removed
2185 because it now seems ill-conceived.
2187 15. Minor typos in src/EDITME comments corrected.
2189 16. Installed latest exipick from John Jetmore.
2191 17. If headers_add on a router specified a text string that was too long for
2192 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2193 of string_sprintf() is now avoided.
2195 18. $message_body_size was not set (it was always zero) when running the DATA
2196 ACL and the local_scan() function.
2198 19. For the "mail" command in an Exim filter, no default was being set for
2199 the once_repeat time, causing a random time value to be used if "once" was
2200 specified. (If the value happened to be <= 0, no repeat happened.) The
2201 default is now 0s, meaning "never repeat". The "vacation" command was OK
2202 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2203 (I found it when inspecting the code).
2205 20. There is now an overall timeout for performing a callout verification. It
2206 defaults to 4 times the callout timeout, which applies to individual SMTP
2207 commands during the callout. The overall timeout applies when there is more
2208 than one host that can be tried. The timeout is checked before trying the
2209 next host. This prevents very long delays if there are a large number of
2210 hosts and all are timing out (e.g. when the network connections are timing
2211 out). The value of the overall timeout can be changed by specifying an
2212 additional sub-option for "callout", called "maxwait". For example:
2214 verify = sender/callout=5s,maxwait=20s
2216 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2217 to the end before writing, but this should make it even safer).
2219 22. Exim was forgetting that it had advertised PIPELINING for the second and
2220 subsequent messages on an SMTP connection. It was also not resetting its
2221 memory on STARTTLS and an internal HELO.
2223 23. When Exim logs an SMTP synchronization error within a session, it now
2224 records whether PIPELINING has been advertised or not.
2226 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2227 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2228 rather than long int.
2230 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2232 26. Added the never_mail option to autoreply.
2238 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2239 crash if the getsockname() call failed; this can happen if a connection is
2240 closed very soon after it is established. The problem was simply in the
2241 order in which certain operations were done, causing Exim to try to write
2242 to the SMTP stream before it had set up the file descriptor. The bug has
2243 been fixed by making things happen in the correct order.
2249 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2250 before the connection was closed, thus losing the rejection response.
2252 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2253 some early Solaris releases, but causes trouble in current releases where
2254 socklen_t is defined.
2256 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2259 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2262 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2263 that releasing the top of it at the end releases what was used for sub-
2264 expansions (unless the block got too big). However, discard this block if
2265 the first thing is a variable or header, so that we can use its block when
2266 it is dynamic (useful for very large $message_headers, for example).
2268 6. Lookups now cache *every* query, not just the most recent. A new, separate
2269 store pool is used for this. It can be recovered when all lookup caches are
2270 flushed. Lookups now release memory at the end of their result strings.
2271 This has involved some general refactoring of the lookup sources.
2273 7. Some code has been added to the store_xxx() functions to reduce the amount
2274 of flapping under certain conditions.
2276 8. log_incoming_interface used to affect only the <= reception log lines. Now
2277 it causes the local interface and port to be added to several more SMTP log
2278 lines, for example "SMTP connection from", and rejection lines.
2280 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2282 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2284 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2285 could be overwritten at the end of the current message (or the start of a
2286 new message if it was set in a HELO ACL). The value is now preserved for
2287 the duration of the SMTP connection.
2289 12. If a transport had a headers_rewrite setting, and a matching header line
2290 contained an unqualified address, that address was qualified, even if it
2291 did not match any rewriting rules. The underlying bug was that the values
2292 of the flags that permit the existence of unqualified sender and recipient
2293 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2294 non-local messages, and by -bnq for local messages) were not being
2295 preserved with the message after it was received.
2297 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2298 "next input=" as part of the text comprising the host identity instead of
2299 the correct text. The code was using the same buffer for two different
2300 strings. However, depending on which order the printing function evaluated
2301 its arguments, the bug did not always show up. Under Linux, for example, my
2302 test suite worked just fine.
2304 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2305 doesn't work with some older versions of Perl. It has been changed to "my",
2306 which in any case is probably the better facility to use.
2308 15. A really picky compiler found some instances of statements for creating
2309 error messages that either had too many or two few arguments for the format
2312 16. The size of the buffer for calls to the DNS resolver has been increased
2313 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2314 for addresses that have a lot of PTR records. This alleviates a problem; it
2315 does not fully solve it.
2317 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2318 buffer now truncates the list and logs the incident, which is the same
2319 action as happens when Exim is looking up a host name and its aliases.
2320 Previously in this situation something unpredictable would happen;
2321 sometimes it was "internal error: store_reset failed".
2323 18. If a server dropped the connection unexpectedly when an Exim client was
2324 using GnuTLS and trying to read a response, the client delivery process
2325 crashed while trying to generate an error log message.
2327 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2328 string, for example:
2330 warn message = H1: something\nH2: something
2332 the text was added as a single header line from Exim's point of view
2333 though it ended up OK in the delivered message. However, searching for the
2334 second and subsequent header lines using $h_h2: did not work. This has been
2335 fixed. Similarly, if a system filter added multiple headers in this way,
2336 the routers could not see them.
2338 20. Expanded the error message when iplsearch is called with an invalid key to
2339 suggest using net-iplsearch in a host list.
2341 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2342 ACLs are no longer actually imposed (and a message to that effect is
2345 22. If a "gecos" field in a passwd entry contained escaped characters, in
2346 particular, if it contained a \" sequence, Exim got it wrong when building
2347 a From: or a Sender: header from that name. A second bug also caused
2348 incorrect handling when an unquoted " was present following a character
2349 that needed quoting.
2351 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2352 was not being matched caselessly.
2354 24. Arranged for all hyphens in the exim.8 source to be escaped with
2357 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2358 properly. Recipient callout cache records were still being keyed to include
2359 the sender, even when use_sender was set false. This led to far more
2360 callouts that were necessary. The sender is no longer included in the key
2361 when use_sender is false.
2363 26. Added "control = submission" modifier to ACLs.
2365 27. Added the ${base62d: operator to decode base 62 numbers.
2367 28. dnsdb lookups can now access SRV records.
2369 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2370 the configuration file.
2372 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2373 (-v) mode. This makes the output for a verbose queue run more intelligible.
2375 31. Added a use_postmaster feature to recipient callouts.
2377 32. Added the $body_zerocount variable, containing the number of binary zero
2378 bytes in the message body.
2380 33. The time of last modification of the "new" subdirectory is now used as the
2381 "mailbox time last read" when there is a quota error for a maildir
2384 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2386 35. Added +ignore_unknown as a special item in host lists.
2388 36. Code for decoding IPv6 addresses in host lists is now included, even if
2389 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2390 address was recognized as an IP address, but was then not correctly decoded
2391 into binary, causing unexpected and incorrect effects when compared with
2398 1. Very minor rewording of debugging text in manualroute to say "list of
2399 hosts" instead of "hostlist".
2401 2. If verify=header_syntax was set, and a header line with an unqualified
2402 address (no domain) and a large number of spaces between the end of the
2403 name and the colon was received, the reception process suffered a buffer
2404 overflow, and (when I tested it) crashed. This was caused by some obsolete
2405 code that should have been removed. The fix is to remove it!
2407 3. When running in the test harness, delay a bit after writing a bounce
2408 message to get a bit more predictability in the log output.
2410 4. Added a call to search_tidyup() just before forking a reception process. In
2411 theory, someone could use a lookup in the expansion of smtp_accept_max_
2412 per_host which, without the tidyup, could leave open a database connection.
2414 5. Added the variables $recipient_data and $sender_data which get set from a
2415 lookup success in an ACL "recipients" or "senders" condition, or a router
2416 "senders" option, similar to $domain_data and $local_part_data.
2418 6. Moved the writing of debug_print from before to after the "senders" test
2421 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2422 problems for message scanning, either using a data ACL, or using
2423 local_scan() because the Received: header was not generated till after they
2424 were called (in order to set the time as the time of reception completion).
2425 I have revised the way this works. The header is now generated after the
2426 body is received, but before the ACL or local_scan() are called. After they
2427 are run, the timestamp in the header is updated.
2433 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2434 before starting a queue runner without re-exec. This happened only when
2435 deliver_drop_privilege was set or when the Exim user was set to root. The
2436 effect of the bug was that timeouts during subsequent deliveries caused
2437 crashes instead of being properly handled. The handler is now left at its
2438 default (and expected) setting.
2440 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2441 message, again when deliver_drop_privilege is set or Exim is run as root.
2442 The bug described in (1) was not present in this case, but the tidying up
2443 of the other signals was missing. I have made the two cases consistent.
2445 3. The ignore_target_hosts setting on a manualroute router was being ignored
2446 for hosts that were looked up using the /MX notation.
2448 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2451 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2452 operated on the sender address. After changing the $sender_address to <>
2453 for the sender address verify, Exim was re-instated it as the original
2454 (before rewriting) address, but remembering that it had rewritten it, so it
2455 wasn't rewriting it again. This bug also had the effect of breaking the
2456 sender address verification caching when the sender address was rewritten.
2458 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2459 This has been changed so that if the ip literal address matches
2460 ignore_target_hosts, the router declines.
2462 7. Added expansion conditions match_domain, match_address, and match_local_
2463 part (NOT match_host).
2465 8. The placeholder for the Received: header didn't have a length field set.
2467 9. Added code to Exim itself and to exim_lock to test for a specific race
2468 condition that could lead to file corruption when using MBX delivery. The
2469 issue is with the lockfile that is created in /tmp. If this file is removed
2470 after a process has opened it but before that process has acquired a lock,
2471 there is the potential for a second process to recreate the file and also
2472 acquire a lock. This could lead to two Exim processes writing to the file
2473 at the same time. The added code performs the same test as UW imapd; it
2474 checks after acquiring the lock that its file descriptor still refers to
2475 the same named file.
2477 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2478 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2481 11. Added the smtp_active_hostname option. If used, this will typically be made
2482 to depend on the incoming interface address. Because $interface_address is
2483 not set up until the daemon has forked a reception process, error responses
2484 that can happen earlier (such as "too many connections") no longer contain
2487 12. If an expansion in a condition on a "warn" statement fails because a lookup
2488 defers, the "warn" statement is abandoned, and the next ACL statement is
2489 processed. Previously this caused the whole ACL to be aborted.
2491 13. Added the iplsearch lookup type.
2493 14. Added ident_timeout as a log selector.
2495 15. Added tls_certificate_verified as a log selector.
2497 16. Added a global option tls_require_ciphers (compare the smtp transport
2498 option of the same name). This controls incoming TLS connections.
2500 17. I finally figured out how to make tls_require_ciphers do a similar thing
2501 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2502 before starting the TLS session.
2504 18. Tabs are now shown as \t in -bP output.
2506 19. If the log selector return_path_on_delivery was set, Exim crashed when
2507 bouncing a message because it had too many Received: header lines.
2509 20. If two routers both had headers_remove settings, and the first one included
2510 a superfluous trailing colon, the final name in the first list and the
2511 first name in the second list were incorrectly joined into one item (with a
2512 colon in the middle).
2518 1. Added -C and -D options to the exinext utility, mainly to make it easier
2519 to include in the automated testing, but these could be helpful when
2520 multiple configurations are in use.
2522 2. The exinext utility was not formatting the output nicely when there was
2523 an alternate port involved in the retry record key, nor when there was a
2524 message id as well (for retries that were specific to a specific message
2525 and a specific host). It was also confused by IPv6 addresses, because of
2526 the additional colons they contain. I have fixed the IPv4 problem, and
2527 patched it up to do a reasonable job for IPv6.
2529 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2530 delivery, the log line now contains "pipelined" if PIPELINING was used.
2532 4. An SMTP transport process used to panic and die if the bind() call to set
2533 an explicit outgoing interface failed. This has been changed; it is now
2534 treated in the same way as a connect() failure.
2536 5. A reference to $sender_host_name in the part of a conditional expansion
2537 that was being skipped was still causing a DNS lookup. This no longer
2540 6. The def: expansion condition was not recognizing references to header lines
2541 that used bh_ and bheader_.
2543 7. Added the _cache feature to named lists.
2545 8. The code for checking quota_filecount in the appendfile transport was
2546 allowing one more file than it should have been.
2548 9. For compatibility with Sendmail, the command line option
2556 and sets the incoming protocol and host name (for trusted callers). The
2557 host name and its colon can be omitted when only the protocol is to be set.
2558 Note the Exim already has two private options, -pd and -ps, that refer to
2559 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2560 "s", but I don't think that's a major issue.
2562 10. A number of refactoring changes to the code, none of which should affect
2565 (a) The number of logging options was getting close to filling up the
2566 32-bit word that was used as a bit map. I have split them into two classes:
2567 those that are passed in the argument to log_write(), and those that are
2568 only ever tested independently outside of that function. These are now in
2569 separate 32-bit words, so there is plenty of room for expansion again.
2570 There is no change in the user interface or the logging behaviour.
2572 (b) When building, for example, log lines, the code previously used a
2573 macro that called string_cat() twice, in order to add two strings. This is
2574 not really sufficiently general. Furthermore, there was one instance where
2575 it was actually wrong because one of the argument was used twice, and in
2576 one call a function was used. (As it happened, calling the function twice
2577 did not affect the overall behaviour.) The macro has been replaced by a
2578 function that can join an arbitrary number of extra strings onto a growing
2581 (c) The code for expansion conditions now uses a table and a binary chop
2582 instead of a serial search (which was left over from when there were very
2583 few conditions). Also, it now recognizes conditions like "pam" even when
2584 the relevant support is not compiled in: a suitably worded error message is
2585 given if an attempt is made to use such a condition.
2587 11. Added ${time_interval:xxxxx}.
2589 12. A bug was causing one of the ddress fields not to be passed back correctly
2590 from remote delivery subprocesses. The field in question was not being
2591 subsequently used, so this caused to problems in practice.
2593 13. Added new log selectors queue_time and deliver_time.
2595 14. Might have fixed a bug in maildirsizefile handling that threw up
2596 "unexpected character" debug warnings, and recalculated the data
2597 unnecessarily. In any case, I expanded the warning message to give more
2600 15. Added the message "Restricted characters in address" to the statements in
2601 the default ACL that block characters like @ and % in local parts.
2603 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2604 Three changes have been made:
2606 (a) There was a serious bug; a negative response to MAIL caused the whole
2607 recipient domain to be cached as invalid, thereby blocking all messages
2608 to all local parts at the same domain, from all senders. This bug has
2609 been fixed. The domain is no longer cached after a negative response to
2610 MAIL if the sender used is not empty.
2612 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2615 (c) A new callout option, "use_sender" has been added for people who want
2616 the modified behaviour.
2622 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2625 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2626 indeed breaks things for older releases.
2628 3. Added additional logging to the case where there is a problem reading data
2629 from a filter that is running in a subprocess using a pipe, in order to
2630 try to track down a specific problem.
2632 4. Testing facility fudge: when running in the test harness and attempting
2633 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2634 getting "No route to host". Convert this to a timeout.
2636 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2639 6. Some OS don't have socklen_t but use size_t instead. This affects the
2640 fifth argument of getsockopt() amongst other things. This is now
2641 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2642 can be set for individual OS. I have set it for SunOS5, OSF1, and
2643 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2644 some earlier ones do not.
2646 7. Change 4.30/15 was not doing the test caselessly.
2648 8. The standard form for an IPv6 address literal was being rejected by address
2649 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2650 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2651 this, as well as the form without the "IPv6" on the front (but only when
2652 address literals are enabled, of course).
2654 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2656 10. Exim crashed if a message with an empty sender address specified by -f
2657 encountered a router with an errors_to setting. This could be provoked only
2658 by a command such as
2662 where an empty string was supplied; "<>" did not hit this bug.
2664 11. Installed PCRE release 4.5.
2666 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2667 remained set. It is now erased.
2669 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2670 times from message ids (which are base 36 rather than the normal 62).
2672 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2673 were being counted as actual protocol errors, and logged if the log
2674 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2675 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2676 and DATA following a set of rejected RCPTs do not count as protocol errors.
2677 In other words, Exim assumes they were pipelined, though this may not
2678 actually be the case. Of course, in all cases the client gets an
2679 appropriate error code.
2681 15. If a lookup fails in an ACL condition, a message about the failure may
2682 be available; it is used if testing the ACL cannot continue, because most
2683 such messages specify what the cause of the deferral is. However, some
2684 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2685 that caused an old message to be retained and used if a later statement
2686 caused a defer, replacing the real cause of the deferral.
2688 16. If an IP address had so many PTR records that the DNS lookup buffer
2689 was not large enough to hold them, Exim could crash while trying to process
2690 the truncated data. It now detects and logs this case.
2692 17. Further to 4.21/58, another change has been made: if (and only if) the
2693 first line of a message (the first header line) ends with CRLF, a bare LF
2694 in a subsequent header line has a space inserted after it, so as not to
2695 terminate the header.
2697 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2698 unnecessarily, used atoi() instead of strtol(), and didn't check the
2699 termination when getting file sizes from file names by regex.
2701 19. Completely re-implemented the support for maildirsize files, in the light
2702 of a number of problems with the previous contributed implementation
2703 (4.30/29). In particular:
2705 . If the quota is zero, the maildirsize file is maintained, but no quota is
2708 . If the maildir directory does not exist, it is created before any attempt
2709 to write a maildirsize file.
2711 . The quota value in the file is just a cache; if the quota is changed in
2712 the transport, the new value overrides.
2714 . A regular expression is available for excluding directories from the
2717 20. The autoreply transport checks the characters in options that define the
2718 message's headers; it allows continued headers, but it was checking with
2719 isspace() after an embedded newline instead of explicitly looking for a
2722 21. If all the "regular" hosts to which an address was routed had passed their
2723 expiry times, and had not reached their retry times, the address was
2724 bounced, even if fallback hosts were defined. Now Exim should go on to try
2727 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2728 equivalent code in the SMTP transport. Some hosts send humungous responses
2729 to HELO/EHLO, more than 1024 it seems.
2731 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2732 gives compiler warnings in some environments. I've now done it "properly",
2735 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2736 (because of problems with the built-in one) was declared to return uschar *
2737 instead of char *, causing compiler failure.
2739 25. Fixed a file descriptor leak when processing alias/forward files.
2741 26. Fixed a minor format string issue in dbfn.c.
2743 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2745 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2746 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2749 29. When a sender address is verified, it is cached, to save repeating the test
2750 when there is more than one recipient in a message. However, when the
2751 verification involves a callout, it is possible for different callout
2752 options to be set for different recipients. It is too complicated to keep
2753 track of this in the cache, so now Exim always runs a verification when a
2754 callout is required, relying on the callout cache for the optimization.
2755 The overhead is duplication of the address routing, but this should not be
2758 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2759 to be verified with callout=postmaster, and the main callout worked but the
2760 postmaster check failed, the verification correctly failed. However, if a
2761 subsequent RCPT command asked for sender verification *without* the
2762 postmaster check, incorrect caching caused this verification also to fail,
2765 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2766 it was not caching the DNS options (qualify_single, search_parents) that
2767 were used when the lookup failed. A subsequent lookup with different
2768 options therefore always gave the same answer, though there were cases
2769 where it should not have. (Example: a "domains = !$mx_any" option on a
2770 dnslookup router: the "domains" option is always processed without any
2771 widening, but the router might have qualify_single set.) Now Exim uses the
2772 cached value only when the same options are set.
2774 32. Added John Jetmore's "exipick" utility to the distribution.
2776 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2777 than a timeout (e.g. a certificate is required, and is not provided), an
2778 Exim server now closes the connection immediately. Previously it waited for
2779 the client to close - but if the client is SSL, it seems that they each
2780 wait for each other, leading to a delay before one of them times out.
2782 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2783 maintained 0.8.x compatibility because I don't think many are using it, and
2784 it is clearly obsolete.
2786 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2789 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2790 client certificate was expired. A simple patch fixes this, though I don't
2791 understand the full logic of why the verify callback is called multiple
2794 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2795 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2796 which causes problems with some clients (such as the Certicom SSL Plus
2797 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2798 disables the coutermeasure allowing Eudora to connect."
2800 38. Exim was not checking that a write() to a log file succeeded. This could
2801 lead to Bad Things if a log got too big, in particular if it hit a file
2802 size limit. Exim now panics and dies if it cannot write to a log file, just
2803 as it does if it cannot open a log file.
2805 39. Modified OS/Makefile-Linux so that it now contains
2807 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2809 The two -D definitions ensure that Exim is compiled with large file
2810 support, which makes it possible to handle log files that are bigger than
2813 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2814 instance) a domain was checked against a named list that involved a lookup,
2815 causing $domain_data to be set, then another domain was checked against the
2816 same list, then the first domain was re-checked, the value of $domain_data
2817 after the final check could be wrong. In particular, if the second check
2818 failed, it could be set empty. This bug probably also applied to
2821 41. The strip_trailing_dot option was not being applied to the address given
2822 with the -f command-line option.
2824 42. The code for reading a message's header from the spool was incrementing
2825 $received_count, but never initializing it. This meant that the value was
2826 incorrect (doubled) while delivering a message in the same process in which
2827 it was received. In the most common configuration of Exim, this never
2828 happens - a fresh exec is done - but it can happen when
2829 deliver_drop_privilege is set.
2831 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2832 it now includes up to 150 characters of the unexpected data in the log
2835 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2836 and building data strings. The size of both of these buffers was 10 000
2837 bytes - far larger than anybody would *ever* want, thought I. Needless to
2838 say, somebody hit the limit. I have increased the maximum line length to
2839 20 000 and the maximum data length of concatenated lines to 100 000. I have
2840 also fixed two bugs, because there was no checking on these buffers. Tsk,
2841 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2842 buffer is too small.
2844 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2845 lsearch lookups. Now it does.
2847 46. When parsing a route_list item in a manualroute router, a fixed-length
2848 buffer was used for the list of hosts. I made this 1024 bytes long,
2849 thinking that nobody would ever have a list of hosts that long. Wrong.
2850 Somebody had a whole pile of complicated expansion conditions, and the
2851 string was silently truncated, leading to an expansion error. It turns out
2852 that it is easier to change to an unlimited length (owing to other changes
2853 that have happened since this code was originally written) than to build
2854 structure for giving a limitation error. The length of the item that
2855 expands into the list of hosts is now unlimited.
2857 47. The lsearch lookup could not handle data where the length of text line was
2858 more than 4095 characters. Such lines were truncated, leading to shortened
2859 data being returned. It should now handle lines of any length.
2861 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2862 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2865 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2866 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2867 no longer happens in comment lines. A list of replacements is now placed
2868 at the head of all of the source files, except those whose only change is
2869 to replace PERL_COMMAND in the very first #! line.
2871 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2872 messages on the queue, with a bottom-up merge sort, using code contributed
2873 by Michael Haardt. This should make operations like -bp somewhat faster on
2874 large queues. It won't affect queue runners, except when queue_run_in_order
2877 51. Installed eximstats 1.31 in the distribution.
2879 52. Added support for SRV lookups to the dnslookup router.
2881 53. If an ACL referred to $message_body or $message_body_end, the value was not
2882 reset for any messages that followed in the same SMTP session.
2884 54. The store-handling optimization for building very long strings was not
2885 differentiating between the different store pools. I don't think this
2886 actually made any difference in practice, but I've tidied it.
2888 55. While running the routers to verify a sender address, $sender_address
2889 was still set to the sender address. This is wrong, because when routing to
2890 send a bounce to the sender, it would be empty. Therefore, I have changed
2891 it so that, while verifying a sender address, $sender_address is set to <>.
2892 (There is no change to what happens when verifying a recipient address.)
2894 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2895 target A or AAAA records (if not already returned) without resetting the
2896 qualify_single or search_parents options of the DNS resolver. These are
2897 inappropriate in this case because the targets of MX and SRV records must
2898 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2899 target that, when qualified, matched something in the local domain. These
2900 two options are now turned off when doing these lookups.
2902 57. It seems that at least some releases of Reiserfs (which does not have the
2903 concept of a fixed number of inodes) returns zero and not -1 for the
2904 number of available inodes. This interacted badly with check_spool_inodes,
2905 which assumed that -1 was the "no such thing" setting. What I have done is
2906 to check that the total number of inodes is greater than zero before doing
2907 the test of how many are available.
2909 58. When a "warn" ACL statement has a log_message modifier, the message is
2910 remembered, and not repeated. This is to avoid a lot of repetition when a
2911 message has many recipients that cause the same warning to be written.
2912 Howewer, Exim was preserving the list of already written lines for an
2913 entire SMTP session, which doesn't seem right. The memory is now reset if a
2914 new message is started.
2916 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2917 debugging output unless log_rewrite was also set.
2919 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2920 of (int)(handle) when we know that handle contains (void *)(-1).
2922 61. The Exim daemon panic-logs an error return when it closes the incoming
2923 connection. However "connection reset by peer" seems to be common, and
2924 isn't really an error worthy of noting specially, so that particular error
2927 62. When Exim is trying to find all the local interfaces, it used to panic and
2928 die if the ioctl to get the interface flags failed. However, it seems that
2929 on at least one OS (Solaris 9) it is possible to have an interface that is
2930 included in the list of interfaces, but for which you get a failure error
2931 for this call. This happens when the interface is not "plumbed" into a
2932 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2933 failure of the "get flags" call assumes that the interface is down.
2935 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2936 makes life easier for people who are doing arithmetic on fields extracted
2937 from dates, where you often get leading zeros that should not be
2938 interpreted as octal.
2940 64. Added qualify_domain to the redirect router, to override the global
2943 65. If a pathologically long header line contained very many addresses (the
2944 report of this problem mentioned 10 000) and each of them was rewritten,
2945 Exim could use up a very large amount of memory. (It kept on making new
2946 copies of the header line as it rewrote, and never released the old ones.)
2947 At the expense of a bit more processing, the header rewriting function has
2948 been changed so that it no longer eats memory in this way.
2950 66. The generation of the Received: header has been moved from the time that a
2951 message starts to be received, to the time that it finishes. The timestamp
2952 in the Received: header should now be very close to that of the <= log
2953 line. There are two side-effects of this change:
2955 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2956 logged header lines no longer include the local Received: line, because
2957 it has not yet been created. The same applies to a copy of the message
2958 that is returned to a non-SMTP sender when a message is rejected.
2960 (b) When a filter file is tested using -bf, no additional Received: header
2961 is added to the test message. After some thought, I decided that this
2964 This change does not affect the value of $received_for. It is still set
2965 after address rewriting, but before local_scan() is called.
2967 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2969 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2970 gave an unhelpful panic error message, and a defer error. I have managed to
2971 change this behaviour so that it now rejects any supplied certificate,
2972 which seems right, as the list of acceptable certificates is empty.
2974 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2975 gave an unhelpful defer error. I have not managed to make this reject any
2976 supplied certificates, but the error message it gives is "no certificate
2977 supplied", which is not helpful.
2979 70. exigrep's output now also includes lines that are not associated with any
2980 message, but which match the given pattern. Implemented by a patch from
2981 Martin Sluka, which also tidied up the Perl a bit.
2983 71. Recipient callout verification, like sender verification, was using <> in
2984 the MAIL FROM command. This isn't really the right thing, since the actual
2985 sender may affect whether the remote host accepts the recipient or not. I
2986 have changed it to use the actual sender in the callout; this means that
2987 the cache record is now keyed on a recipient/sender pair, not just the
2988 recipient address. There doesn't seem to be a real danger of callout loops,
2989 since a callout by the remote host to check the sender would use <>.
2990 [SEE ABOVE: changed after hitting problems.]
2992 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2993 temporary errors. However, in the case of such a code being given after
2994 the end of a data transmission (i.e. after ".") Exim was failing to write
2995 a retry record for the message. (Yes, there was some broken host that was
2996 actually sending 8xx at this point.)
2998 73. An unknown lookup type in a host list could cause Exim to panic-die when
2999 the list was checked. (An example that provoked this was putting <; in the
3000 middle of a list instead of at the start.) If this happened during a DATA
3001 ACL check, a -D file could be left lying around. This kind of configuration
3002 error no longer causes Exim to die; instead it causes a defer errror. The
3003 incident is still logged to the main and panic logs.
3005 74. Buglet left over from Exim 3 conversion. The message "too many messages
3006 in one connection" was written to the rejectlog but not the mainlog, except
3007 when address rewriting (yes!) was being logged.
3009 75. Added write_rejectlog option.
3011 76. When a system filter was run not as root (that is, when system_filter_user
3012 was set), the values of the $n variables were not being returned to the
3013 main process; thus, they were not subsequently available in the $sn
3016 77. Added +return_path_on_delivery log selector.
3018 78. A connection timeout was being treated differently from recipients deferred
3019 when testing hosts_max_try with a message that was older than the host's
3020 retry timeout. (The host should not be counted, thus allowing all hosts to
3021 be tried at least once before bouncing.) This may have been the cause of an
3022 occasionally reported bug whereby a message would remain on the queue
3023 longer than the retry timeout, but would be bounced if a delivery was
3024 forced. I say "may" because I never totally pinned down the problem;
3025 setting up timeout/retry tests is difficult. See also the next item.
3027 79. The ultimate address timeout was not being applied to errors that involved
3028 a combination of host plus message (for example, a timeout on a MAIL
3029 command). When an address resolved to a number of possible hosts, and they
3030 were not all tried for each delivery (e.g. because of hosts_max_try), a
3031 message could remain on the queue longer than the retry timeout.
3033 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3036 81. Fixed an obscure SMTP outgoing bug which required at least the following
3037 conditions: (a) there was another message waiting for the same server;
3038 (b) the server returned 5xx to all RCPT commands in the first message so
3039 that the message was not completed; (c) the server dropped the connection
3040 or gave a negative response to the RSET that Exim sends to abort the
3041 transaction. The observed case was a dropped connection after DATA that had
3042 been sent in pipelining mode. That is, the server had advertised PIPELINING
3043 but was not implementing it correctly. The effect of the bug was incorrect
3044 behaviour, such as trying another host, and this could lead to a crash.
3050 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3051 and daemon.c were passed as pointers to ints; they should have been
3052 pointers to socklen_t variables (which are typically unsigned ints).
3054 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3057 3. Fixed a really odd bug that affected only the testing scheme; patching a
3058 certain fixed string in the binary changed the value of another string that
3059 happened to be identical to the end of the original first string.
3061 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3062 name", it returns that address as the IP address. On some operating
3063 systems (e.g. Solaris), it also passes back the IP address string as the
3064 "host name". However, on others (e.g. Linux), it passes back an empty
3065 string. Exim wasn't checking for this, and was changing the host name to an
3066 empty string, assuming it had been canonicized.
3068 5. Although rare, it is permitted to have more than one PTR record for a given
3069 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3070 all the names associated with an address, because they do in Solaris.
3071 However, it seems that they do not in Linux for data that comes from the
3072 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3073 I found this out when I moved to a new Linux workstation and tried to run
3074 the Exim test suite.
3076 To get round this problem I have changed the code so that it now does its
3077 own call to the DNS to look up PTR records when searching for a host name.
3078 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3079 addresses that are only in /etc/hosts are still found.
3081 This behaviour is, however, controlled by an option called host_lookup_
3082 order, which defaults to "bydns:byaddr". If people want to use the other
3083 order, or indeed, just use one or the other means of lookup, they can
3084 specify it in this variable.
3086 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3087 some operating systems, this comes back from gethostbyaddr() as an empty
3088 string, and this is what Exim used to test for. However, it seems that in
3089 other systems, "." is yielded. Exim now tests for this case too.
3091 7. The values of check_spool_space and check_log_space are now held internally
3092 as a number of kilobytes instead of an absolute number of bytes. If a
3093 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3094 kilobyte. This means that much larger values can be stored.
3096 8. Exim monitor: an attempt to get the action menu when not actually pointing
3097 at a message produces an empty menu entitled "No message selected". This
3098 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3099 no entries in it ("Shell widget menu has zero width and/or height"). So I
3100 have added a single, blank menu entry in this case.
3102 9. Added ${quote_local_part.
3104 10. MIME decoding is now applied to the contents of Subject: header lines when
3107 11. Now that a reference to $sender_host_address automatically causes a reverse
3108 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3109 host lookup before query-style lookups in lists that might use this
3110 variable. This has therefore been abolished, and the "net-" prefix is no
3111 longer necessary for query-style lookups.
3113 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3114 have been a typo for LFLAGS, so it has been changed.
3116 13. The install script calls Exim with "-C /dev/null" in order to find the
3117 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3118 to be output. Howeve, since Exim outputs its version number before the
3119 error, it didn't break the script. It just looked ugly. I fixed this by
3120 always allowing "-C /dev/null" if the caller is root.
3122 14. Ignore overlarge ACL variable number when reading spool file - insurance
3123 against a later release with more variables having written the file.
3125 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3126 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3127 this, as well as the form without the "IPv6" on the front.
3129 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3130 OS/Makefile-Darwin file.
3132 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3133 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3135 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3136 variable. However, some OS use ints for inodes. I've added cast to long int
3137 to get rid of the compiler warning.
3139 19. I had forgotten to lock out "/../" in configuration file names when
3140 ALT_CONFIG_PREFIX was set.
3142 20. Routers used for verification do not need to specify transports. However,
3143 if such a router generated a host list, and callout was configured, Exim
3144 crashed, because it could not find a port number from the (non-existent)
3145 transport. It now assumes port 25 in this circumstance.
3147 21. Added the -t option to exigrep.
3149 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3150 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3151 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3152 them is set, LOOKUP_LSEARCH is forced.
3154 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3156 24. Added sender and host information to the "rejected by local_scan()" log
3157 line; previously there was no indication of these.
3159 25. Added .include_if_exists.
3161 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3162 Linux. It turns out that not all file systems support this. Apparently some
3163 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3164 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3165 not supported on the file descriptor, is now ignored when Exim is trying to
3166 sync a directory. This applies only to Linux.
3168 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3170 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3171 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3174 29. Added support for maildirsize files from supplied patch (modified a bit).
3176 30. The use of :fail: followed by an empty string could lead Exim to respond to
3177 sender verification failures with (e.g.):
3179 550 Verification failed for <xxx>
3180 550 Sender verify failed
3182 where the first response line was missing the '-' that indicates it is not
3183 the final line of the response.
3185 31. The loop for finding the name of the user that called Exim had a hardwired
3186 limit of 10; it now uses the value of finduser_retries, which is used for
3187 all other user lookups.
3189 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3192 33. Exim was neglecting to zero errno before one call of strtol() when
3193 expanding a string and expecting an integer value. On some systems this
3194 resulted in spurious "integer overflow" errors. Also, it was casting the
3195 result into an int without checking.
3197 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3198 did not work. The code looks as if it has *never* worked, though it appears
3199 to have been documented since at least releast 1.62. I have made it work.
3201 35. The "timeout_DNS" error in retry rules, also documented since at least
3202 1.62, also never worked. As it isn't clear exactly what this means, and
3203 clearly it isn't a major issue, I have abolished the feature by treating it
3204 as "timeout", and writing a warning to the main and panic logs.
3206 36. The display of retry rules for -brt wasn't always showing the error code
3209 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3210 timeout_connect_A, timeout_connect_MX.
3212 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3213 to the empty sender.
3215 39. The daemon was not analysing the content of -oX till after it had closed
3216 stderr and disconnected from the controlling terminal. This meant that any
3217 syntax errors were only noted on the panic log, and the return code from
3218 the command was 0. By re-arranging the code a little, I've made the
3219 decoding happen first, so such errors now appear on stderr, and the return
3220 code is 1. However, the actual setting up of the sockets still happens in
3221 the disconnected process, so errors there are still only recorded on the
3224 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3225 connections (as happens on some IP stacks) was logged at start up time as
3226 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3227 it from "IPv6 and IPv4", which means that two separate sockets are being
3230 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3231 says whether AF_INET or AF_INET6 was passed as an argument.
3233 42. Exiwhat output was messed up when time zones were included in log
3236 43. Exiwhat now gives more information about the daemon's listening ports,
3237 and whether -tls-on-connect was used.
3239 44. The "port" option of the smtp transport is now expanded.
3241 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3242 silently ignored. Now an error message is written to the main and panic
3245 46. There's a new ACL modifier called "logwrite" which writes to a log file
3246 as soon as it is encountered.
3248 47. Added $local_user_uid and $local_user_gid at routing time.
3250 48. Exim crashed when trying to verify a sender address that was being
3253 49. Exim was recognizing only a space character after ".include". It now also
3254 recognizes a tab character.
3256 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3257 extracting the relevant information from the specification. The man page no
3258 longer contains scrambled data for the -d option, and I've added a section
3259 at the front about calling Exim under different names.
3261 51. Added "extra_headers" argument to the "mail" command in filter files.
3263 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3266 53. Installed eximstats 1.29.
3268 54. Added transport_filter_timeout as a generic transport option.
3270 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3271 Cc: header lines. This was required by RFC 822, but it not required by RFC
3274 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3275 incoming messages that did not have them. Now it does so only if the
3276 message originates locally, that is, if there is no associated remote host
3277 address. When Resent- header lines are present, this applies to the Resent-
3278 lines rather than the non-Resent- lines.
3280 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3281 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3283 58. Messages for configuration errors now include the name of the main
3284 configuration file - useful now that there may be more than one file in a
3285 list (.included file names were always shown).
3287 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3288 for those rare installations that do not start the daemon as root or run it
3289 setuid root. I've cut out the call to initgroups() if the daemon is not
3292 60. The Exim user and group can now be bound into the binary as text strings
3293 that are looked up at the start of Exim's processing.
3295 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3297 62. Added $mailstore_basename variable.
3299 63. Installed patch to sieve.c from Michael Haardt.
3301 64. When Exim failed to open the panic log after failing to open the main log,
3302 the original message it was trying to log was written to stderr and debug
3303 output, but if they were not available (the usual case in production), it
3304 was lost. Now it is written to syslog before the two lines that record the
3305 failures to open the logs.
3307 65. Users' Exim filters run in subprocesses under the user's uid. It is
3308 possible for a "deliver" command or an alias in a "personal" command to
3309 provoke an address rewrite. If logging of address rewriting is configured,
3310 this fails because the process is not running as root or exim. There may be
3311 a better way of dealing with this, but for the moment (because 4.30 needs
3312 to be released), I have disabled address rewrite logging when running a
3313 filter in a non-root, non-exim process.
3319 1. The buildconfig auxiliary program wasn't quoting the value set for
3320 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3321 not defined. This bug was masked in 4.22 by the effect that was fixed in
3324 2. Some messages that were rejected after a message id was allocated were
3325 shown as "incomplete" by exigrep. It no longer does this for messages that
3326 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3328 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3329 have allow_domain_literals set, the ID did not get logged in the <= line.
3330 Domain literals are now always recognized in Message-ID: header lines.
3332 4. The first argument for a ${extract expansion item is the key name or field
3333 number. Leading and trailing spaces in this item were not being ignored,
3334 causing some misleading effects.
3336 5. When deliver_drop_privilege was set, single queue runner processes started
3337 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3338 same command in the process it spins off) were not dropping privilege.
3340 6. When the daemon running as "exim" started a queue runner, it always
3341 re-executed Exim in the spun-off process. This is a waste of effort when
3342 deliver_drop_privilege is set. The new process now just calls the
3343 queue-runner function directly.
3349 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3352 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3353 ignored. Though the use of -oP was forcing the writing of a pid file, it
3354 was always written to the default place.
3356 3. If the message "no IP address found for host xxxx" is generated during
3357 incoming verification, it is now followed by identification of the incoming
3358 connection (so you can more easily find what provoked it).
3360 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3362 5. Added some features to "harden" Exim a bit more against certain attacks:
3364 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3365 be put in Local/Makefile. This is like the never_users runtime option,
3366 but it cannot be overridden. The default setting is "root".
3368 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3369 prefix string with which any file named in a -C command line option
3372 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3373 is retained for -C and -D only if the caller of Exim is root. Without
3374 it, the exim user may also use -C and -D and retain privilege.
3376 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3377 command line option is disabled.
3379 6. Macro names set by the -D option must start with an upper case letter, just
3380 like macro names defined in the configuration file.
3382 7. Added "dereference=" facility to LDAP.
3384 8. Two instances of the typo "uknown" in the source files are fixed.
3386 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3387 the Configure-Makefile script screwed up while processing it.
3389 10. Incorporated PCRE 4.4.
3391 11. The SMTP synchronization check was not operating right at the start of an
3392 SMTP session. For example, it could not catch a HELO sent before the client
3393 waited for the greeting. There is now a check for outstanding input at the
3394 point when the greeting is written. Because of the duplex, asynchronous
3395 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3396 way, but not yet received, when the check is performed.
3398 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3399 on TCP/IP sockets, because this apparently causes some broken clients to
3402 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3403 unchanged) from the Cygwin maintainer.
3405 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3406 was supported instead of testing for mbx. Effectively a typo.
3408 15. The spa authenticator server code was not checking that the input it
3409 received was valid base64.
3411 16. The debug output line for the "set" modifier in ACLs was not showing the
3412 name of the variable that was being set.
3414 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3416 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3417 lookup on its own. Something else was needed to trigger the lookup. For
3418 example, a match in host_lookup or the need for a host name in a host list.
3419 Now, if $sender_host_name is referenced and the host name has not yet been
3420 looked up, a lookup is performed. If the lookup fails, the variable remains
3421 empty, and $host_lookup_failed is set to "1".
3423 19. Added "eqi" as a case-independent comparison operator.
3425 20. The saslauthd authentication condition could segfault if neither service
3426 nor realm was specified.
3428 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3429 error message that was logged was misleading, and incoming SMTP
3430 connections were dropped. The message is now more accurate, and temporary
3431 errors are given to SMTP connections.
3433 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3434 (except RSET) with a 421 error, until QUIT is received. However, it was
3435 failing to send a response to QUIT.
3437 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3438 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3439 ACL is now run after the helo_try_verify_hosts code.
3441 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3442 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3443 case-independent, so other case variants are also recognized). Apparently
3444 some systems use these upper case variants.
3446 25. If more than two messages were waiting for the same host, and a transport
3447 filter was specified for the transport, Exim sent two messages over the
3448 same TCP/IP connection, and then failed with "socket operation on non-
3449 socket" when it tried to send the third.
3451 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3453 27. The extern definition of crypt16() in expand.c was not being excluded when
3454 the OS had its own crypt16() function.
3456 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3457 as a preferred synonym for return_size_limit, both as an option and as an
3460 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3462 30. Changed the default configuration ACL to relax the local part checking rule
3463 for addresses that are not in any local domains. For these addresses,
3464 slashes and pipe symbols are allowed within local parts, but the sequence
3465 /../ is explicitly forbidden.
3467 31. SPA server authentication was not clearing the challenge buffer before
3470 32. log_message in a "warn" ACL statement was writing to the reject log as
3471 well as to the main log, which contradicts the documentation and doesn't
3472 seem right (because no rejection is happening). So I have stopped it.
3474 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3475 However, I am unable to do any testing of this.
3477 34. Fixed an infelicity in the appendfile transport. When checking directories
3478 for a mailbox, to see if any needed to be created, it was accidentally
3479 using path names with one or more superfluous leading slashes; tracing
3480 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3482 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3483 contents are added to the log line that is written for every discarded
3484 recipient. (Previously a log_message setting was ignored.)
3486 36. The ${quote: operator now quotes the string if it is empty.
3488 37. The install script runs exim in order to find its version number. If for
3489 some reason other than non-existence or emptiness, which it checks, it
3490 could not run './exim', it was installing it with an empty version number,
3491 i.e. as "exim-". This error state is now caught, and the installation is
3494 38. An argument was missing from the function that creates an error message
3495 when Exim fails to connect to the socket for saslauthd authentication.
3496 This could cause Exim to crash, or give a corrupted message.
3498 39. Added isip, isip4, and isip6 to ${if conditions.
3500 40. The ACL variables $acl_xx are now saved with the message, and can be
3501 accessed later in routers, transports, and filters.
3503 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3504 strings in the file are not string-expanded.
3506 42. If a MAIL command specified a SIZE value that was too large to fit into an
3507 int variable, the check against message_size_limit failed. Such values are
3508 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3509 day this will have to be increased, but I don't think I want to be around
3510 when emails are that large.
3517 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3518 iconv() is not standard in FreeBSD.
3520 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3521 IPv6 enabled. The observed symptom was a segmentation fault on return from
3522 the function os_common_find_running_interfaces() in src/os.c.
3524 3. In the check_special_case() function in daemon.c I had used "errno" as an
3525 argument name, which causes warnings on some systems. This was basically a
3526 typo, since it was named "eno" in the comments!
3528 4. The code that waits for the clock to tick (at a resolution of some fraction
3529 of a second) so as to ensure message-id uniqueness was always waiting for
3530 at least one whole tick, when it could have waited for less. [This is
3531 almost certainly not relevant at current processor speeds, where it is
3532 unlikely to ever wait at all. But we try to future-proof.]
3534 5. The function that sleeps for a time interval that includes fractions of a
3535 second contained a race. It did not block SIGALRM between setting the
3536 timer, and suspending (a couple of lines later). If the interval was short
3537 and the sigsuspend() was delayed until after it had expired, the suspension
3538 never ended. On busy systems this could lead to processes getting stuck for
3541 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3542 process, before it forks any delivery processes. The open lookup caching
3543 mechanism meant that the open file or database connection was passed into
3544 the delivery process. The problem was that delivery processes always tidy
3545 up cached lookup data. This could cause a problem for the next delivery
3546 process started by the queue runner, because the external queue runner
3547 process does not know about the closure. So the next delivery process
3548 still has data in the lookup cache. In the case of a file lookup, there was
3549 no problem because closing a file descriptor in a subprocess doesn't affect
3550 the parent. However, if the lookup was caching a connection to a database,
3551 the connection was closed, and the second delivery process was likely to
3552 see errors such as "PGSQL: query failed: server closed the connection
3553 unexpectedly". The problem has been fixed by closing all cached lookups
3554 in a queue runner before running a delivery process.
3556 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3557 seem to have the "const" qualifier which it has on other OS. I've
3560 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3561 *of the same type* (client or server) with the same public name that an
3562 error should be diagnosed.
3564 9. When Exim looked up a host name for an IP address, but failed to find the
3565 original IP address when looking up the host name (a safety check), it
3566 output the message "<ip address> does not match any IP for NULL", which was
3567 confusing, to say the least. The bug was that the host name should have
3568 appeared instead of "NULL".
3570 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3571 user that is built into the binary, and the -C or -D options is used, root
3572 privilege is dropped before the configuration file is read. In addition,
3573 logging is switched to stderr instead of the normal log files. If the
3574 configuration then re-defines the Exim user, the unprivileged environment
3575 is probably not what is expected, so Exim logs a panic warning message (but
3578 However, if deliver_drop_privilege is set, the unprivileged state may well
3579 be exactly what is intended, so the warning has been cut out in that case,
3580 and Exim is allowed to try to write to its normal log files.
3586 1. smtp_return_error_details was not giving details for temporary sender
3587 or receiver verification errors.
3589 2. Diagnose a configuration error if two authenticators have the same public
3592 3. Exim used not to create the message log file for a message until the first
3593 delivery attempt. This could be confusing when incoming messages were held
3594 for policy or load reasons. The message log file is now created at the time
3595 the message is received, and an initial "Received" line is written to it.
3597 4. The automatically generated man page for command line options had a minor
3598 bug that caused no ill effects; however, a more serious problem was that
3599 the procedure for building the man page automatically didn't always
3600 operate. Consequently, release 4.20 contains an out-of-date version. This
3601 shouldn't happen again.
3603 5. When building Exim with embedded Perl support, the script that builds the
3604 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3605 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3607 6. The freeze_tell option was not being used for messages that were frozen on
3608 arrival, either by an ACL or by local_scan().
3610 7. Added the smtp_incomplete_transaction log selector.
3612 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3613 was accepting AUTH without a new EHLO.
3615 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3616 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3619 10. Logging of TCP/IP connections (when configured) now happens in the main
3620 daemon process instead of the child process, so that the TCP/IP connection
3621 count is more accurate (but it can never be perfect).
3623 11. The use of "drop" in a nested ACL was not being handled correctly in the
3624 outer ACL. Now, if condition failure induced by the nested "drop" causes
3625 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3626 or "require"), the connection is dropped.
3628 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3629 that yield "discard" can now be used with an "accept" or a "discard" verb,
3630 but an error is generated for any others (because I can't see a useful way
3631 to define what should happen).
3633 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3634 are now processed in the same way as lines in the Exim configuration file.
3635 In particular, continuation lines are supported.
3637 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3639 15. Added -ti meaning -t -i.
3641 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3642 domains, and warn by logging if others are found.
3644 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3645 returned from GIFCONF when Exim is trying to find the list of interfaces on
3646 a host. The code in os.c has been modified to copy each ifreq to an aligned
3647 structure in all cases.
3649 Also, in some cases, the returned ifreq's were being copied to a 'struct
3650 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3651 means the last couple of bytes of an IPv6 address could be chopped if the
3652 ifreq contained only a normal sockaddr (14 bytes storage).
3654 18. Named domain lists were not supported in the hosts_treat_as_local option.
3655 An entry such as +xxxx was not recognized, and was treated as a literal
3658 19. Ensure that header lines added by a DATA ACL are included in the reject log
3659 if the ACL subsequently rejects the message.
3661 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3662 MD5 (which is deprecated).
3664 21. When testing a filter file using -bf, Exim was writing a message when it
3665 took the sender from a "From " line in the message, but it was not doing so
3666 when it took $return_path from a Return-Path: header line. It now does.
3668 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3669 with a valid header line field name (a series of printing characters
3670 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3672 23. Changed "disc" in the source to "disk" to conform to the documentation and
3673 the book and for uniformity.
3675 24. Ignore Sendmail's -Ooption=value command line item.
3677 25. When execve() failed while trying to run a command in a pipe transport,
3678 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3679 could be confused with a return value of 69 from the command itself. This
3680 has been changed to 127, the value the shell returns if it is asked to run
3681 a non-existent command. The wording for the related log line suggests a
3682 non-existent command as the problem.
3684 26. If received_header_text expands to an empty string, do not add a Received:
3685 header line to the message. (Well, it adds a token one on the spool, but
3686 marks it "old" so that it doesn't get used or transmitted.)
3688 27. Installed eximstats 1.28 (addition of -nt option).
3690 28. There was no check for failure on the call to getsockname() in the daemon
3691 code. This can fail if there is a shortage of resources on the system, with
3692 ENOMEM, for example. A temporary error is now given on failure.
3694 29. Contrary to the C standard, it seems that in some environments, the
3695 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3696 program. Exim now does this explicitly; it affects the formatting of
3697 timestamps using strftime().
3699 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3700 complaints. I've now initialized all the variables, to avoid this.
3702 32. Header lines added by a system filter were not being "seen" during
3703 transport-time rewrites.
3705 33. The info_callback() function passed to OpenSSL is set up with type void
3706 (*)(SSL *, int, int), as described somewhere. However, when calling the
3707 function (actually a macro) that sets it up, the type void(*)() is
3708 expected. I've put in a cast to prevent warnings from picky compilers.
3710 34. If a DNS black list lookup found a CNAME record, but there were no A
3711 records associated with the domain it pointed at, Exim crashed.
3713 35. If a DNS black list lookup returned more than one A record, Exim ignored
3714 all but the first. It now scans all returned addresses if a particular IP
3715 value is being sought. In this situation, the contents of the
3716 $dnslist_value variable are a list of all the addresses, separated by a
3719 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3720 to accept a host name and all its aliases if the forward lookup for any of
3721 them yielded the IP address of the incoming connection. Now it accepts only
3722 those names whose forward lookup yields the correct IP address. Any other
3723 names are discarded. This closes a loophole whereby a rogue DNS
3724 administrator could create reverse DNS records to break through a
3725 wildcarded host restriction in an ACL.
3727 37. If a user filter or a system filter that ran in a subprocess used any of
3728 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3729 the wrong values were passed to the pipe command ($thisaddress had the
3730 value of $0, $0 had the value of $1, etc). This bug was introduced by
3731 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3733 38. Improved the line breaking for long SMTP error messages from ACLs.
3734 Previously, if there was no break point between 40 and 75 characters, Exim
3735 left the rest of the message alone. Two changes have been made: (a) I've
3736 reduced the minimum length to 35 characters; (b) if it can't find a break
3737 point between 35 and 75 characters, it looks ahead and uses the first one
3738 that it finds. This may give the occasional overlong line, but at least the
3739 remaining text gets split now.
3741 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3742 file descriptors might be low, and that setting 1000 would always raise it.
3743 It turns out that in some environments, the limit is already over 1000 and
3744 that lowering it causes trouble. So now Exim takes care not to decrease it.
3746 40. When delivering a message, the value of $return_path is set to $sender_
3747 address at the start of routing (routers may change the value). By an
3748 oversight, this default was not being set up when an address was tested by
3749 -bt or -bv, which affected the outcome if any router or filter referred to
3752 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3753 log only when "message" or "log_message" are set. However, if one of the
3754 conditions was an address verification, or a call to a nested ACL, the
3755 messages generated by the underlying test were being passed through. This
3756 no longer happens. The underlying message is available in $acl_verify_
3757 message for both "message" and "log_message" expansions, so it can be
3758 passed through if needed.
3760 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3761 new expansion $bh_ to give the encoded byte string without charset
3762 translation. Translation happens only if iconv() is available; HAVE_ICONV
3763 indicates this at build time. HEADERS_CHARSET gives the charset to
3764 translate to; headers_charset can change it in the configuration, and
3765 "headers charset" can change it in an individual filter file.
3767 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3768 that creates RFC 2047 encoded "words" labels them as that charset instead
3769 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3770 expansion operator; (ii) when Exim creates a From: line for a local
3771 message; (iii) when a header line is rewritten to include a "phrase" part.
3773 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3774 buggy, causing it to skip the first lines of messages whose message ID
3775 ended in 'D'. This would not have bitten before Exim release 4.14, because
3776 message IDs were unlikely to end in 'D' before then. The effect was to have
3777 incorrect size information for certain domains.
3779 45. #include "config.h" was missing at the start of the crypt16.c module. This
3780 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3783 46. If there was a timeout during a "random" callout check, Exim treated it as
3784 a failure of the random address, and carried on sending RSET and the real
3785 address. If the delay was just some slowness somewhere, the response to the
3786 original RCPT would be taken as a response to RSET and so on, causing
3787 mayhem of various kinds.
3789 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3790 when I implemented it. It didn't allow for the fact that some option values
3791 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3792 the right test for positive values.
3794 48. Domain names in DNS records are case-independent. Exim always looks them up
3795 in lower case. Some resolvers return domain names in exactly the case they
3796 appear in the zone file, that is, they may contain uppercase letters. Not
3797 all resolvers do this - some return always lower case. Exim was treating a
3798 change of case by a resolver as a change of domain, similar to a widening
3799 of a domain abbreviation. This triggered its re-routing code and so it was
3800 trying to route what was effectively the same domain again. This normally
3801 caused routing to fail (because the router wouldn't handle the domain
3802 twice). Now Exim checks for this case specially, and just changes the
3803 casing of the domain that it ultimately uses when it transmits the message
3806 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3809 50. If a filter generated a file delivery with a non-absolute name (possible if
3810 no home directory exists for the router), the forbid_file option was not
3813 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3814 the existing equality matching.
3816 52. Exim was using ints instead of ino_t variables in some places where it was
3817 dealing with inode numbers.
3819 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3820 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3821 called TMPDIR, and if it finds it is different, it changes its value.
3823 54. The smtp_printf() function is now made available to local_scan() so
3824 additional output lines can be written before returning. There is also an
3825 smtp_fflush() function to enable the detection of a dropped connection.
3826 The variables smtp_input and smtp_batched_input are exported to
3829 55. Changed the default runtime configuration: the message "Unknown user"
3830 has been removed from the ACL, and instead placed on the localuser router,
3831 using the cannot_route_message feature. This means that any verification
3832 failures that generate their own messages won't get overridden. Similarly,
3833 the "Unrouteable address" message that was in the ACL for unverifiable
3834 relay addresses has also been removed.
3836 56. Added hosts_avoid_esmtp to the smtp transport.
3838 57. The exicyclog script was not checking for the esoteric option
3839 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3840 will work only if exicyclog is run under the appropriate euid.
3842 58. Following a discussion on the list, the rules by which Exim recognises line
3843 endings on incoming messages have been changed. The -dropcr and drop_cr
3844 options are now no-ops, retained only for backwards compatibility. The
3845 following line terminators are recognized: LF CRLF CR. However, special
3846 processing applies to CR:
3848 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3849 nor a local message in the state where . is a terminator.
3851 (ii) If a bare CR is encountered in a header line, an extra space is added
3852 after the line terminator so as not to end the header. The reasoning
3853 behind this is that bare CRs in header lines are most likely either
3854 to be mistakes, or people trying to play silly games.
3856 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3857 was being incorrectly given as 18 bytes larger than it should have been.
3858 This is a VOB (very old bug).
3860 60. This may never have affected anything current, but just in case it has:
3861 When the local host is found other than at the start of a list of hosts,
3862 the local host, those with the same MX, and any that follow, are discarded.
3863 When the list in question was part of a longer list of hosts, the following
3864 hosts (not currently being processed) were also being discarded. This no
3865 longer happens. I'm not sure if this situation could ever has previously
3868 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3871 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3872 header. This is something that is recommended in a new Internet Draft, and
3873 is something that is documented as being done by Sendmail. There are two
3874 possible values. For messages generated by the autoreply transport, Exim
3877 Auto-Submitted: auto-replied
3879 whereas for all other generated messages (e.g. bounces) it adds
3881 Auto-Submitted: auto-generated
3883 63. The "personal" condition in filters now includes a test for the
3884 Auto-Submitted: header. If it contains the string "auto-" the message it
3885 not considered personal.
3887 64. Added rcpt_include_affixes as a generic transport option.
3889 65. Added queue_only_override (default true).
3891 66. Added the syslog_duplication option.
3893 67. If what should have been the first header line of a message consisted of
3894 a space followed by a colon, Exim was mis-interpreting it as a header line.
3895 It isn't of course - it is syntactically invalid and should therefore be
3896 treated as the start of the message body. The misbehaviour could have
3897 caused a number of strange effects, including loss of data in subsequent
3898 header lines, and spool format errors.
3900 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3901 client host had authenticated. This control can now be exercised by an ACL
3902 for more flexibility.
3904 69. By default, callouts do not happen when testing with -bh. There is now a
3905 variant, -bhc, which does actually run the callout code, including
3906 consulting and updating the callout cache.
3908 70. Added support for saslauthd authentication, courtesy of Alexander
3911 71. If statvfs() failed on the spool or log directories while checking their
3912 size for availability, Exim confusingly gave the error "space shortage".
3913 Furthermore, in debugging mode it crashed with a floating point exception.
3914 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3915 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3916 serious problem, Exim now writes to the main and panic logs when this
3917 happens, with details of the failure. It then refuses to accept the
3918 incoming message, giving the message "spool directory problem" or "log
3919 directory problem" with a 421 code for SMTP messages.
3921 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3922 0, 1, and 2 exist, because some OS complain for execs without them (see
3923 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3924 descriptors. However, the code omitted to check that the open succeeded,
3925 causing mysterious errors if for some reason the permissions on /dev/null
3926 got screwed. Now Exim writes a message to the main and panic logs, and
3927 bombs out if it can't open /dev/null.
3929 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3930 interact so that it is all more flexible. It is supposed to remain
3931 backwards compatible. Also added extra_local_interfaces.
3933 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3934 to bomb out with an assertion failure - to the client this appears as a
3935 connection drop. This problem occurs in the part of the code that was taken
3936 from the Samba project. Fortunately, the assertion is in a very simple
3937 function, so I have fixed this by reproducing the function inline in the
3938 one place where it is called, and arranging for authentication to fail
3939 instead of killing the process with assert().
3941 75. The SPA client code was not working when the server requested OEM rather
3942 than Unicode encoding.
3944 76. Added code to make require_files with a specific uid setting more usable in
3945 the case where statting the file as root fails - usually a non-root-mounted
3946 NFS file system. When this happens and the failure is EACCES, Exim now
3947 forks a subprocess and does the per-uid checking as the relevant uid.
3949 77. Added process_log_path.
3951 78. If log_file_path was not explicitly set, a setting of check_log_space or
3952 check_log_inodes was ignored.
3954 79. If a space check for the spool or log partitions fails, the incident is now
3955 logged. Of course, in the latter case the data may get lost...
3957 80. Added the %p formatting code to string_format() so that it can be used to
3958 print addresses in debug_print(). Adjusted all the address printing in the
3959 debugging in store.c to use %p rather than %d.
3961 81. There was a concern that a line of code in smtp_in.c could overflow a
3962 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3963 initially expressed, the concern was not well-founded, because trailing
3964 spaces are removed early. However, if the trailing spaces were followed by
3965 a NULL, they did not get removed, so the overflow was possible. Two fixes
3968 (a) I re-wrote the offending code in a cleaner fashion.
3969 (b) If an incoming SMTP command contains a NULL character, it is rejected
3972 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3973 runs initgroups(), so that if the Exim user is in any additional groups,
3974 they will be used during message reception.
3980 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob