1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.340 2006/04/18 11:13:19 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
19 PH/01 The code for finding all the local interface addresses on a FreeBSD
20 system running IPv6 was broken. This may well have applied to all BSD
21 systems, as well as to others that have similar system calls. The broken
22 code found IPv4 interfaces correctly, but gave incorrect values for the
23 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
24 that it would not match correctly against @[] and not recognize the IPv6
27 PH/02 The ipliteral router was not recognizing addresses of the form user@
28 [ipv6:....] because it didn't know about the "ipv6:" prefix.
30 PH/03 Added disable_ipv6.
32 PH/04 Changed $reply_address to use the raw form of the headers instead of the
33 decoded form, because it is most often used to construct To: headers
34 lines in autoreplies, and the decoded form may well be syntactically
35 invalid. However, $reply_address has leading white space removed, and all
36 newlines turned into spaces so that the autoreply transport does not
39 PH/05 If group was specified without a user on a router, and no group or user
40 was specified on a transport, the group from the router was ignored.
42 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
43 for visible compile-time settings that can be used to change these
44 numbers, for those that want even more. Backwards compatibility with old
45 spool files has been maintained. However, going back to a previous Exim
46 release will lost any variables that are in spool files.
48 PH/07 Two small changes when running in the test harness: increase delay when
49 passing a TCP/IP connection to a new process, in case the original
50 process has to generate a bounce, and remove special handling of
51 127.0.0.2 (sic), which is no longer necessary.
53 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
54 be the same on different OS.
56 PH/09 Moved a debug statement in filter processing to avoid a race problem when
59 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
60 whether --show-vars was specified or not
62 JJ/02 exipick: Added support for new ACL variable spool format introduced
65 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
66 syntactically invalid From: or Reply-to: line, and a filter used this to
67 generate an autoreply, and therefore failed to obtain an address for the
68 autoreply, Exim could try to deliver to a non-existent relative file
69 name, causing unrelated and misleading errors. What now happens is that
70 it logs this as a hard delivery error, but does not attempt to create a
73 PH/11 The exinext utility has a -C option for testing purposes, but although
74 the given file was scanned by exinext itself; it wasn't being passed on
77 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
78 an end-of-file indication when reading a command response.
80 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
81 compiled. In many other places in Exim, IPv6 addresses are always
82 recognized, so I have changed this. It also means that IPv4 domain
83 literals of the form [IPV4:n.n.n.n] are now always recognized.
85 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
86 used if the router is not running as root, for example, when verifying at
87 ACL time, or when using -bh. The debugging output from this situation was
88 non-existent - all you got was a failure to exec. I have made two
91 (a) Failures to set uid/gid, the current directory, or a process leader
92 in a subprocess such as that created by queryprogram now generate
93 suitable debugging ouput when -d is set.
95 (b) The queryprogram router detects when it is not running as root,
96 outputs suitable debugging information if -d is set, and then runs
97 the subprocess without attempting to change uid/gid.
99 PH/15 Minor change to Makefile for building test_host (undocumented testing
102 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
103 additional section of a DNS packet that returns MX or SRV records.
104 Instead, it always explicitly searches for A/AAAA records. This avoids
105 major problems that occur when a DNS server includes only records of one
106 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
107 fixed another bug: if SRV records were looked up and the corresponding
108 address records were *not* found in the additional section, the port
109 values from the SRV records were lost.
111 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
112 using the correct key (the original address) when searching the retry
113 rules in order to find which one to use for generating the retry hint.
115 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
116 adding the default one. Similarly, if it contains a Reply-To: header, the
117 errors_reply_to option, if set, is not used.
119 PH/19 When calculating a retry time, Exim used to measure the "time since
120 failure" by looking at the "first failed" field in the retry record. Now
121 it does not use this if it is later than than the arrival time of the
122 message. Instead it uses the arrival time. This makes for better
123 behaviour in cases where some deliveries succeed, thus re-setting the
124 "first failed" field. An example is a quota failure for a huge message
125 when small messages continue to be delivered. Without this change, the
126 "time since failure" will always be short, possible causing more frequent
127 delivery attempts for the huge message than are intended.
129 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
130 $1, $2, $3) because the numerical variables can be reset during some
131 expansion items (e.g. "match"), thereby losing the authentication data.
133 PH/21 Make -bV show the size of off_t variables so that the test suite can
134 decide whether to run tests for quotas > 2G.
136 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
137 mailbox_size, and mailbox_filecount in the appendfile transport. If a
138 filecount value is greater than 2G or if a quota value is greater than 2G
139 on a system where the size of off_t is not greater than 4, a panic error
142 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
143 never match. The debug and -bh output now contains an explicit error
144 message indicating a malformed IPv4 address or mask.
146 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
147 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
150 PH/25 Do not write to syslog when running in the test harness. The only
151 occasion when this arises is a failure to open the main or panic logs
152 (for which there is an explicit test).
154 PH/26 Added the /no_tell option to "control=freeze".
156 PH/27 If a host name lookup failed very early in a connection, for example, if
157 the IP address matched host_lookup and the reverse lookup yielded a name
158 that did not have a forward lookup, an error message of the form "no IP
159 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
160 could be logged. Now it outputs the IP address instead of "NULL".
162 PH/28 An enabling patch from MH: add new function child_open_exim2() which
163 allows the sender and the authenticated sender to be set when
164 submitting a message from within Exim. Since child_open_exim() is
165 documented for local_scan(), the new function should be too.
167 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
168 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
169 results in an empty string is now treated as unset.
171 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
173 PH/31 Added sender_verify_fail logging option.
175 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
176 needed by RFC 822 but not by RFC 2822 was commented out. I have now
177 tidied the source and removed it altogether.
179 PH/33 When a queue run was abandoned because the load average was too high, a
180 log line was always written; now it is written only if the queue_run log
181 selector is set. In addition, the log line for abandonment now contains
182 information about the queue run such as the pid. This is always present
183 in "start" and "stop" lines but was omitted from the "abandon" line.
185 PH/34 Omit spaces between a header name and the colon in the error message that
186 is given when verify = headers_syntax fails (if there are lots of them,
187 the message gets confusing).
189 PH/35 Change the default for dns_check_names_pattern to allow slashes within
190 names, as there are now some PTR records that contain slashes. This check
191 is only to protect against broken name servers that fall over on strange
192 characters, so the fact that it applies to all lookups doesn't matter.
194 PH/36 Now that the new test suite is complete, we can remove some of the
195 special code in Exim that was needed for the old test suite. For example,
196 sorting DNS records because real resolvers return them in an arbitrary
197 order. The new test suite's fake resolver always returns records in the
200 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
201 bounces) except when queue_only is set, to avoid logging races between
202 the different processes.
204 PH/38 Panic-die if .include specifies a non-absolute path.
206 PH/39 A tweak to the "H" retry rule from its user.
208 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
209 a label. They prevented compilation on older perls.
211 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
212 a warning to be raised on newish perls.
214 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
215 on queue. Changes to match documented behaviour of showing count of
216 messages matching specified criteria.
218 PH/40 Changed the default ident timeout from 30s to 5s.
220 PH/41 Added support for the use of login_cap features, on those BSD systems
221 that have them, for controlling the resources used by pipe deliveries.
223 PH/42 The content-scanning code uses fopen() to create files in which to put
224 message data. Previously it was not paying any attention to the mode of
225 the files. Exim runs with umask(0) because the rest of the code creates
226 files with open(), and sets the required mode explicitly. Thus, these
227 files were ending up world-writeable. This was not a big issue, because,
228 being within the spool directory, they were not world-accessible. I have
229 created a function called modefopen, which takes an additional mode
230 argument. It sets umask(777), creates the file, chmods it to the required
231 mode, then resets the umask. All the relevant calls to fopen() in the
232 content scanning code have been changed to use this function.
234 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
235 to 24 hours. This avoids potential overflow problems when processing G
236 and H retry rules. I suspect nobody ever tinkers with this value.
238 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
240 PH/45 When the plaintext authenticator is running as a client, the server's
241 challenges are checked to ensure they are valid base64 strings. By
242 default, the authentication attempt is cancelled if an invalid string is
243 received. Setting client_ignore_invalid_base64 true ignores these errors.
244 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
245 they are received. Thus, the responses can be made to depend on the
246 challenges. If an invalid string is ignored, an empty string is placed in
249 PH/46 Messages that are created by the autoreply transport now contains a
250 References: header, in accordance with RFCs 2822 and 3834.
252 PH/47 Added authenticated_sender_force to the smtp transport.
254 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
256 PH/49 Installed latest patch from the Sieve maintainer.
258 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
259 was also set, the appendfile transport was unnecessarily scanning a
260 directory of message files (e.g. for maildir delivery) to find the count
261 of files (along with the size), even though it did not need this
262 information. It now does the scan only if it needs to find either the
263 size of the count of files.
265 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
267 PH/52 Two bugs concerned with error handling when the smtp transport is
270 (i) Exim was not creating retry information for temporary errors given
271 for individual recipients after the DATA command when the smtp transport
272 was used in LMTP mode. This meant that they could be retried too
273 frequently, and not timed out correctly.
275 (ii) Exim was setting the flag that allows error details to be returned
276 for LMTP errors on RCPT commands, but not for LMTP errors for individual
277 recipients that were returned after the DATA command.
279 PH/53 This is related to PH/52, but is more general: for any failing address,
280 when detailed error information was permitted to be returned to the
281 sender, but the error was temporary, then after the final timeout, only
282 "retry timeout exceeded" was returned. Now it returns the full error as
283 well as "retry timeout exceeded".
285 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
286 do this, and (what is worse) MTAs that accept it.
288 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
289 will now be deprecated.
291 PH/56 New os.c-cygwin from the Cygwin maintainer.
293 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
294 formats (previously only available in exim formats via -bpr, -bpru,
295 and -bpra. Now also available in native and exiqgrep formats)
297 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
298 with very large, slow to parse queues
300 JJ/08 exipick: added ! as generic prefix to negate any criteria format
302 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
304 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
305 responses to authentication challenges, though it was showing the
306 challenges; (ii) I've removed the CR characters from the debug output for
309 PH/58 Allow for the insertion of a newline as well as a space when a string
310 is turned into more than one encoded-word during RFC 2047 encoding. The
311 Sieve code now uses this.
313 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
314 data_4xx, lost_connection, tls_required.
316 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
317 message was being sent as an SMTP response.
319 PH/61 Add -l and -k options to exicyclog.
321 PH/62 When verifying, if an address was redirected to one new address, so that
322 verification continued, and the new address failed or deferred after
323 having set something in $address_data, the value of $address_data was not
324 passed back to the ACL. This was different to the case when no
325 redirection occurred. The value is now passed back in both cases.
327 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
328 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
329 use login_cap.h, so on its own it isn't the distinguishing feature. The
330 new name refers directly to the setclassresources() function.
332 PH/65 Added configuration files for NetBSD3.
334 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
336 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
337 is preferred over IPv4.
339 PH/68 The bounce_return_message and bounce_return_body options were not being
340 honoured for bounces generated during the reception of non-SMTP messages.
341 In particular, this applied to messages rejected by the ACL. This bug has
342 been fixed. However, if bounce_return_message is true and bounce_return_
343 body is false, the headers that are returned for a non-SMTP message
344 include only those that have been read before the error was detected.
345 (In the case of an ACL rejection, they have all been read.)
347 PH/69 The HTML version of the specification is now built in a directory called
348 spec_html instead of spec.html, because the latter looks like a path with
349 a MIME-type, and this confuses some software.
351 PH/70 Catch two compiler warnings in sieve.c.
353 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
354 function verify_get_ident() calls ip_connect() to connect a socket, but
355 if the "connect()" function timed out, ip_connect() used to close the
356 socket. However, verify_get_ident() also closes the socket later, and in
357 between Exim writes to the log, which may get opened at this point. When
358 the socket was closed in ip_connect(), the log could get the same file
359 descriptor number as the socket. This naturally causes chaos. The fix is
360 not to close the socket in ip_connect(); the socket should be closed by
361 the function that creates it. There was only one place in the code where
362 this was missing, in the iplookup router, which I don't think anybody now
363 uses, but I've fixed it anyway.
365 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
366 well as to direct DNS lookups. Otherwise the handling of names in host
367 lists is inconsistent and therefore confusing.
373 PH/01 Two changes to the default runtime configuration:
375 (1) Move the checks for relay_from_hosts and authenticated clients from
376 after to before the (commented out) DNS black list checks.
378 (2) Add control=submission to the relay_from_hosts and authenticated
379 clients checks, on the grounds that messages accepted by these
380 statements are most likely to be submissions.
382 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
384 (1) Generate an error if the third argument for the ${prvs expansion is
387 (2) Treat a missing third argument of ${prvscheck as if it were an empty
390 (3) Reset the variables that are obtained from the first argument of
391 ${prvscheck and used in the second argument before leaving the code,
392 because their memory is reclaimed, so using them afterwards may do
395 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
396 one (it's much easier than Tom thought :-).
398 (5) Because of (4), we can now allow for the use of $prvscheck_result
399 inside the third argument.
401 PH/03 For some reason, the default setting of PATH when running a command from
402 a pipe transport was just "/usr/bin". I have changed it to
405 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
406 anything to be listed in the output from -bV.
408 PH/05 When a filter generated an autoreply, the entire To: header line was
409 quoted in the delivery log line, like this:
411 => >A.N.Other <ano@some.domain> <original@ddress> ...
413 This has been changed so that it extracts the operative address. There
414 may be more than one such address. If so, they are comma-separated, like
417 => >ano@some.domain,ona@other.domain <original@ddress> ...
419 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
420 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
421 not being looked up in the rDNS to get a host name, Exim was showing the
422 IP address twice in Received: lines, even though the IP addresses were
423 identical. For example:
425 Received: from [1.2.3.4] (helo=[1.2.3.4])
427 However, if the real host name was known, it was omitting the HELO data
428 if it matched the actual IP address. This has been tidied up so that it
429 doesn't show the same IP address twice.
431 PH/07 When both +timestamp and +memory debugging was on, the value given by
432 $tod_xxx expansions could be wrong, because the tod_stamp() function was
433 called by the debug printing, thereby overwriting the timestamp buffer.
434 Debugging no longer uses the tod_stamp() function when +timestamp is set.
436 PH/08 When the original message was included in an autoreply transport, it
437 always said "this is a copy of the message, including all the headers",
438 even if body_only or headers_only was set. It now gives an appropriate
441 PH/09 Applied a patch from the Sieve maintainer which:
443 o fixes some comments
444 o adds the (disabled) notify extension core
445 o adds some debug output for the result of if/elsif tests
446 o points to the current vacation draft in the documentation
447 and documents the missing references header update
451 o fixes a bug in processing the envelope test (when testing
452 multiple envelope elements, the last element determinted the
455 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
456 Electronic Mail") by including:
458 Auto-submitted: auto-generated
460 in the messages that it generates (bounce messages and others, such as
461 warnings). In the case of bounce messages for non-SMTP mesages, there was
462 also a typo: it was using "Auto_submitted" (underscore instead of
463 hyphen). Since every message generated by Exim is necessarily in response
464 to another message, thes have all been changed to:
466 Auto-Submitted: auto-replied
468 in accordance with these statements in the RFC:
470 The auto-replied keyword:
472 - SHOULD be used on messages sent in direct response to another
473 message by an automatic process,
475 - MUST NOT be used on manually-generated messages,
477 - MAY be used on Delivery Status Notifications (DSNs) and Message
478 Disposition Notifications (MDNs),
480 - MUST NOT be used on messages generated by automatic or periodic
481 processes, except for messages which are automatic responses to
484 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
485 to the default Received: header definition.
487 PH/12 Added log selector acl_warn_skipped (default on).
489 PH/13 After a successful wildlsearch lookup, discard the values of numeric
490 variables because (a) they are in the wrong storage pool and (b) even if
491 they were copied, it wouldn't work properly because of the caching.
493 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
494 checking when decoding. Apparently there are clients that generate
495 overlong encoded strings. Why am I not surprised?
497 PH/15 If the first argument of "${if match_address" was not empty, but did not
498 contain an "@" character, Exim crashed. Now it writes a panic log message
499 and treats the condition as false.
501 PH/16 In autoreply, treat an empty string for "once" the same as unset.
503 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
504 extension "envelope-auth". The code is finished and in agreement with
505 other implementations, but there is no documentation so far and in fact,
506 nobody wrote the draft yet. This extension is currently #undef'ed, thus
507 not changing the active code.
509 Print executed "if" and "elsif" statements when debugging is used. This
510 helps a great deal to understand what a filter does.
512 Document more things not specified clearly in RFC3028. I had all this
513 sorted out, when out of a sudden new issues came to my mind. Oops."
515 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
518 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
519 canonical form (as documented). However, after a host name lookup from
520 the IP address, check_host() was doing a simple string comparison with
521 addresses acquired from the DNS when checking that the found name did
522 have the original IP as one of its addresses. Since any found IPv6
523 addresses are likely to be in abbreviated form, the comparison could
524 fail. Luckily, there already exists a function for doing the comparison
525 by converting both addresses to binary, so now that is used instead of
528 PH/20 There was another similar case to PH/19, when a complete host name was
529 given in a host list; looking up its IP address could give an abbreviated
530 form, whereas the current host's name might or might not be abbreviated.
531 The same fix has been applied.
537 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
538 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
541 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
542 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
544 PH/03 Typo: missing ".o" in src/pcre/Makefile.
546 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
547 header line, restrict the check to what is listed in RFCs 2369 and 2929.
548 Also, for "Auto-Submitted", treat anything other than "no" as
549 non-personal, in accordance with RFC 3834. (Previously it treated
550 anything starting "auto-" as non-personal.)
552 TF/01 The control=submission/name=... option had a problem with syntax
553 errors if the name included a slash character. The /name= option
554 now slurps the rest of the string, so it can include any characters
555 but it must come last in the list of options (after /sender_retain
558 PH/05 Some modifications to the interface to the fake nameserver for the new
566 TK/01 Added the "success_on_redirect" address verification option. See
567 NewStuff for rationale and an example.
569 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
571 PH/02 Patch to exigrep to allow it to work on syslog lines.
573 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
574 fread() to skip over the body file's header line, because in Cygwin the
575 header line is locked and is inaccessible.
577 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
578 co-exist for some time) to make it clear that it is the Exim ID that is
579 referenced, not the Message-ID: header line.
581 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
582 string_format() function, because snprintf() does not exist on all
585 PH/06 The use of forbid_filter_existstest now also locks out the use of the
586 ${stat: expansion item.
588 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
589 protocol synchronization error", to keep the pedants happy.
591 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
592 well as for IRIX systems, when gcc is being used. See the host.c source
595 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
597 PH/10 Named domain lists were not working if used in a queue_smtp_domains
600 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
601 transport and to the smtp transport in LMTP mode.
603 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
605 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
606 run a filter in a subprocess. This could lead to confusion in subsequent
607 lookups in the parent process. There should also be a search_tidyup() at
608 the end of the subprocess.
610 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
611 only if the host matched helo_try_verify_hosts, which caused the
612 verification to occur when the EHLO/HELO command was issued. The ACL just
613 tested the remembered result. Now, if a previous verification attempt has
614 not happened, "verify = helo" does it there and then.
616 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
618 TK/03 Fix log output including CR from clamd.
620 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
621 exist provoked a memory error which could cause a segfault.
623 PH/15 Installed PCRE 6.2
625 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
627 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
628 of the problem. Specifically, suggested +O2 rather than +O1 for the
631 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
633 PH/20 If a delivery was routed to a non-standard port by means of an SRV
634 record, the port was not correctly logged when the outgoing_port log
635 selector was set (it logged the transort's default port).
637 PH/21 Added support for host-specific ports to manualroute, queryprogram,
638 fallback_hosts, and "hosts" in the smtp transport.
640 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
641 host errors such as "Connection refused".
643 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
644 authentication with radiusclient 0.4.9:
646 - Error returned from rc_read_config was caught wrongly
647 - Username/password not passed on to radius server due to wrong length.
649 The presumption is that some radiusclient API changes for 4.51/PH/17
650 were not taken care of correctly. The code is still untested by me (my
651 Linux distribution still has 0.3.2 of radiusclient), but it was
652 contributed by a Radius user.
654 PH/24 When doing a callout, the value of $domain wasn't set correctly when
655 expanding the "port" option of the smtp transport.
657 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
658 while reading a MIME header. Thanks to Tom Hughes for a patch.
660 PH/24 Include config.h inside local_scan.h so that configuration settings are
663 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
664 that in an ACL for RCPT (for example), you can examine exactly what was
667 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
668 commands, but it was not correctly comparing the address with the actual
669 client host address. Thus, it would show the EHLO address in Received:
670 header lines when this was not necessary.
672 PH/27 Added the % operator to ${eval:}.
674 PH/28 Exim tries to create and chdir to its spool directory when it starts;
675 it should be ignoring failures (because with -C, for example, it has lost
676 privilege). It wasn't ignoring creation failures other than "already
679 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
682 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
683 because an input line was too long, either on its own, or by virtue of
684 too many continuations, the temporary file was not being removed, and the
685 return code was incorrect.
687 PH/31 Missing "BOOL" in function definition in filtertest.c.
689 PH/32 Applied Sieve patches from the maintainer.
691 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
693 PH/33 Added "verify = not_blind".
695 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
696 Local/Makefile (with some defaults set). These are used in built scripts
697 such as exicyclog, but they have never been used in the exim_install
698 script (though there are many overriding facilities there). I have
699 arranged that the exim_install script now takes note of these two
702 PH/35 Installed configuration files for Dragonfly.
704 PH/36 When a locally submitted message by a trusted user did not contain a
705 From: header, and the sender address was obtained from -f or from an SMTP
706 MAIL command, and the trusted user did not use -F to supply a sender
707 name, $originator_name was incorrectly used when constructing a From:
708 header. Furthermore, $originator_name was used for submission mode
709 messages from external hosts without From: headers in a similar way,
710 which is clearly wrong.
712 PH/37 Added control=suppress_local_fixups.
714 PH/38 When log_selector = +received_sender was set, and the addition of the
715 sender made the log line's construction buffer exactly full, or one byte
716 less than full, an overflow happened when the terminating "\n" was
719 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
720 when the result of a list match is failure because a DNS lookup failed.
722 PH/40 RM_COMMAND is now used in the building process.
724 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
725 the "build-* directories that it finds.
727 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
728 address in a domain literal was a prefix of an interface address.
730 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
731 when verifying a sender address, unless rewrite_headers is false.
733 PH/44 Wrote a long comment about why errors_to addresses are verified as
734 recipients, not senders.
736 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
737 the ratelimit ACL was added.
739 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
741 PH/46 Added extra information about PostgreSQL errors to the error string.
743 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
744 avoiding the need to install special zones in a real server. This is
745 backwards compatible; if it can't find the fake resolver, it drops back.
746 Thus, both old and new test suites can be run.
748 TF/02 Added util/ratelimit.pl
750 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
751 clock is set back in time.
753 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
754 Candler <B.Candler@pobox.com>.
756 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
757 OK for addresses that are the result of redirections.
759 PH/48 A number of further additions for the benefit of the new test suite,
760 including a fake gethostbyname() that interfaces to the fake DNS resolver
763 TF/06 The fix for widen_domains has also been applied to qualify_single and
764 search_parents which are the other dnslookup options that can cause
767 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
770 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
772 TF/07 Exim produced the error message "an SRV record indicated no SMTP
773 service" if it encountered an MX record with an empty target hostname.
774 The message is now "an MX or SRV record indicated no SMTP service".
776 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
777 if the DNS of the sending site is misconfigured. This is quite a
778 common situation. This change restores the behaviour of treating a
779 helo verification defer as a failure.
781 PH/51 If self=fail was set on a router, the bounce message did not include the
782 actual error message.
788 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
790 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
791 command itself ended in error, the underlying message about the transport
792 filter timeout was being overwritten with the pipe command error. Now the
793 underlying error message should be appended to the second error message.
795 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
797 PH/02 Reduce the amount of output that "make" produces by default. Full output
798 can still be requested.
800 PH/03 The warning log line about a condition test deferring for a "warn" verb
801 was being output only once per connection, rather than after each
802 occurrence (because it was using the same function as for successful
803 "warn" verbs). This seems wrong, so I have changed it.
805 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
806 it should not have, which might have caused a crash in the right
807 circumstances, but probably never did.
809 PH/04 Installed a modified version of Tony Finch's patch to make submission
810 mode fix the return path as well as the Sender: header line, and to
811 add a /name= option so that you can make the user's friendly name appear
814 TF/03 Added the control = fakedefer ACL modifier.
816 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
817 Mark Lowes for thorough testing.
819 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
821 TK/03 Merged latest SRS patch from Miles Wilton.
823 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
824 with the definition in sysexits.h (which is #included earlier).
825 Fortunately, Exim does not actually use EX_OK. The code used to try to
826 preserve the sysexits.h value, by assumimg that macro definitions were
827 scanned for macro replacements. I have been disabused of this notion,
828 so now the code just undefines EX_OK before #including unistd.h.
830 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
831 in the smtp transport. When a block could not be written in a single
832 write() function, the timeout was being re-applied to each part-write.
833 This seems wrong - if the receiver was accepting one byte at a time it
834 would take for ever. The timeout is now adjusted when this happens. It
835 doesn't have to be particularly precise.
837 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
838 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
840 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
841 without a domain if the check to <postmaster@domain> fails.
843 SC/01 Eximstats: added -xls and the ability to specify output files
844 (patch written by Frank Heydlauf).
846 SC/02 Eximstats: use FileHandles for outputing results.
848 SC/03 Eximstats: allow any combination of xls, txt, and html output.
850 SC/04 Eximstats: fixed display of large numbers with -nvr option
852 SC/05 Eximstats: fixed merging of reports with empty tables.
854 SC/06 Eximstats: added the -include_original_destination flag
856 SC/07 Eximstats: removed tabs and trailing whitespace.
858 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
860 TK/06 MBOX spool code: Add real "From " MBOX separator line
861 so the .eml file is really in mbox format (even though
862 most programs do not really care). Patch from Alex Miller.
864 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
865 The latter is generated from $received_to and is only set if the
866 message has one envelope recipient. SA can use these headers,
867 obviously out-of-the-box. Patch from Alex Miller.
869 PH/08 The ${def test on a variable was returning false if the variable's
870 value was "0", contrary to what the specification has always said!
871 The result should be true unless the variable is empty.
873 PH/09 The syntax error of a character other than { following "${if
874 def:variable_name" (after optional whitespace) was not being diagnosed.
875 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
876 accidental colon was present, for example, could give incorrect results.
878 PH/10 Tidied the code in a number of places where the st_size field of a stat()
879 result is used (not including appendfile, where other changes are about
882 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
883 This involved changing a lot of size variables from int to off_t. It
884 should work with maildirs and everything.
886 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
887 spamd dying while we are connected to it.
889 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
892 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
893 function for BDB 4.3.
895 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
897 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
898 the sources for PCRE have been re-arranged and more files are now
901 PH/15 The code I had for printing potentially long long variables in PH/11
902 above was not the best (it lost precision). The length of off_t variables
903 is now inspected at build time, and an appropriate printing format (%ld
904 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
905 to be "long long int" or "long int". This is needed for the internal
906 formatting function string_vformat().
908 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
909 the configuration file to be ":syslog", then the script "guesses" where
910 the logs files are, rather than using the compiled in default. In our
911 case the guess is not the same as the compiled default, so the script
912 suddenly stopped working when I started to use syslog. The patch checks
913 to see if log_file_path is "". If so, it attempts to read it from exim
914 with no configuration file to get the compiled in version, before it
915 falls back to the previous guessing code."
917 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
918 implementing BATV in an Exim configuration. See NewStuff for the gory
921 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
922 Makefile that are specific to HP-UX.
924 PH/18 If the "use_postmaster" option was set for a recipient callout together
925 with the "random" option, the postmaster address was used as the MAIL
926 FROM address for the random test, but not for the subsequent recipient
927 test. It is now used for both.
929 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
930 patch removes a few documentation additions to RFC 3028, because the
931 latest draft now contains them. It adds the new en;ascii-case comparator
932 and a new error check for 8bit text in MIME parts. Comparator and
933 require names are now matched exactly. I enabled the subaddress
934 extension, but it is not well tested yet (read: it works for me)."
936 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
937 rework some of the code of TK/09 above to avoid the hardwired use of
938 "%lld" and "long long". Replaced the call to snprintf() with a call to
941 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
942 records point to non-existent hosts", "retry timeout exceeded", and
943 "retry time not reached for any host after a long failure period".
945 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
946 experimental DomainKeys support:
948 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
949 (2) On an error such as an illegally used "control", the wrong name for
950 the control was given.
952 These problems did NOT occur unless DomainKeys support was compiled.
954 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
956 PH/24 Added ${if match_ip condition.
958 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
959 databases so that it will be absolutely obvious if a crash occurs in the
960 DB library. This is a regular occurrence (often caused by mis-matched
963 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
964 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
965 picked up on a user's system that detects such things. There doesn't seem
966 to be a gcc warning option for this - only an attribute that has to be
967 put on the function's prototype. It seems that in Fedora Core 4 they have
968 set this on a number of new functions. No doubt there will be more in due
971 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
972 specify a transport. However, if an address that was verified by such a
973 router was the subject of a callout, Exim crashed because it tried to
974 read the rcpt_include_affixes from the non-existent transport. Now it
975 just assumes that the setting of that option is false. This bug was
976 introduced by 4.51/PH/31.
978 PH/28 Changed -d+all to exclude +memory, because that information is very
979 rarely of interest, but it makes the output a lot bigger. People tend to
980 do -d+all out of habit.
982 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
983 code in os-type was giving problems when libc.so lives in lib64, like on
986 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
987 aren't the modern standard, and it seems that some systems' include files
988 don't always have them. Exim was already checking for some of the newer
989 ones like T_AAAA, and defining it itself. I've added checks for all the
990 record types that Exim uses.
992 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
993 not automatically generating a new one, as it is supposed to. This
994 prevented TLS from working. If the file did exist, but contained invalid
995 data, a new version was generated, as expected. It was only the case of a
996 non-existent file that was broken.
998 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
999 with a change in libdomainkeys > 0.64.
1001 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1002 from DNS. If the selector record carries the flag, it now has
1003 precedence over the domain-wide flag.
1005 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1007 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1008 the use of a port name that isn't defined in /etc/services) occurred, the
1009 message was deferred as in a normal delivery, and thus remained on the
1010 spool, instead of being failed because of the mua_wrapper setting. This
1011 is now fixed, and I tidied up some of the mua_wrapper messages at the
1014 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1015 the messages in a hash of arrays rather than using individual hashes.
1016 This is a bit cleaner and results in dramatic memory savings, albeit
1017 at a slight CPU cost.
1019 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1020 as requested by Marc Sherman.
1022 SC/10 Eximstats: added histograms for user specified patterns as requested
1025 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1027 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1028 fopen() in the content-scanning modules that did not already have it.
1034 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1035 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1037 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1039 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1041 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1042 to test Sieve filters that use "vacation".
1044 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1045 that changes the way the GnuTLS parameters are stored in the cache file.
1046 The new format can be generated externally. For backward compatibility,
1047 if the data in the cache doesn't make sense, Exim assumes it has read an
1048 old-format file, and it generates new data and writes a new file. This
1049 means that you can't go back to an older release without removing the
1052 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1053 work if there are any delivery delays because "one_time" forces the
1054 parent to be marked "delivered", so its unseen clone is never tried
1055 again. For this reason, Exim now forbids the simultaneous setting of
1058 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1059 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1060 entry if you want to know the details. The fix, however, neglected to
1061 consider the case when local delivery batching is involved. The test for
1062 "previously delivered" was not happening when checking to see if an
1063 address could be batched with a previous (undelivered) one; under
1064 certain circumstances this could lead to multiple deliveries to the same
1067 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1068 in its include files, and this causes problems building Exim.
1070 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1071 header_syntax) but Exim was just ignoring anything given after a slash.
1072 In particular, this caused confusion with an attempt to use "verify =
1073 reverse_host_lookup/defer_ok". An error is now given when options are
1074 supplied for verify items that do not have them. (Maybe reverse_host_
1075 lookup should have a defer_ok option, but that's a different point.)
1077 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1078 defined by RFC 821) to 2048, because there were problems with some AUTH
1079 commands, and RFC 1869 says the size should be increased for extended
1080 SMTP commands that take arguments.
1082 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1085 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1086 "unknown" error; now it says that the functionality isn't in the binary.
1088 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1089 an address' error message when a string expansion fails (syntax or
1090 whatever). Otherwise the password may appear in the log. Following change
1091 PH/42 below, there is no longer a chance of it appearing in a bounce
1094 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1096 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1097 of its addresses was ever tried. (Bugzilla bug #2.)
1099 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1100 the result incorrectly in the debug output. (It correctly added a newline
1101 to what was transported.)
1103 TF/01 Added $received_time.
1105 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1106 commented out examples of how to interface to a virus scanner and to
1107 SpamAssassin. Also added commented examples of av_scanner and
1108 spamd_address settings.
1110 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1111 and controls are allowed in which ACLs. There were a couple of minor
1112 errors. Some of the entries in the conditions table (which is a table of
1113 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1114 negation of where the condition IS allowed.
1116 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1118 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1119 header file does not have a version number, so I've had to invent a new
1120 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1121 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1122 radiusclient), but it was contributed by a Radius user.
1124 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1125 files or directories, for OpenSSL.
1127 PH/19 When an Exim process that is running as root has to create an Exim log
1128 file, it does so in a subprocess that runs as exim:exim so as to get the
1129 ownership right at creation (otherwise, other Exim processes might see
1130 the file with the wrong ownership). There was no test for failure of this
1131 fork() call, which would lead to the process getting stuck as it waited
1132 for a non-existent subprocess. Forks do occasionally fail when resources
1133 run out. I reviewed all the other calls to fork(); they all seem to check
1136 PH/20 When checking for unexpected SMTP input at connect time (before writing
1137 the banner), Exim was not dealing correctly with a non-positive return
1138 from the read() function. If the client had disconnected by this time,
1139 the result was a log entry for a synchronization error with an empty
1140 string after "input=" when read() returned zero. If read() returned -1
1141 (an event I could not check), uninitialized data bytes were printed.
1142 There were reports of junk text (parts of files, etc) appearing after
1145 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1147 PH/22 Added support for macro redefinition, and (re)definition in between
1148 driver and ACL definitions.
1150 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1151 forgetting to use the resulting value; it was using the unexpanded value.
1153 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1154 hadn't been configured. The fix is from Juergen Kreileder, who
1155 understands it better than I do:
1157 "Here's what I see happening with three configured cyrus_sasl
1158 authenticators configured (plain, login, cram-md5):
1160 On startup auth_cyrus_sasl_init() gets called for each of these.
1161 This means three calls to sasl_listmech() without a specified mech_list.
1162 => SASL tests which mechs of all available mechs actually work
1163 => three warnings about OTP not working
1164 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1166 With the patch, sasl_listmech() also gets called three times. But now
1167 SASL's mech_list option is set to the server_mech specified in the the
1168 authenticator. Or in other words, the answer from sasl_listmech()
1169 gets limited to just the mech you're testing for (which is different
1171 => the return list contains just 'plain' or 'login', 'cram-md5' or
1172 nothing depending on the value of ob->server_mech.
1174 I've just tested the patch: Authentication still works fine,
1175 unavailable mechs specified in the exim configuration are still
1176 caught, and the auth.log warnings about OTP are gone."
1178 PH/25 When debugging is enabled, the contents of the command line are added
1179 to the debugging output, even when log_selector=+arguments is not
1182 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1183 answer is "GNU", and only if the return is "GNU/something" is the answer
1186 PH/27 $acl_verify_message is now set immediately after the failure of a
1187 verification in an ACL, and so is available in subsequent modifiers. In
1188 particular, the message can be preserved by coding like this:
1190 warn !verify = sender
1191 set acl_m0 = $acl_verify_message
1193 Previously, $acl_verify_message was set only while expanding "message"
1194 and "log_message" when a very denied access.
1196 PH/28 Modified OS/os.c-Linux with
1198 -#ifndef OS_LOAD_AVERAGE
1199 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1201 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1202 nomenclature these days.)
1204 PH/29 Installed patch from the Sieve maintainer that adds the options
1205 sieve_useraddress and sieve_subaddress to the redirect router.
1207 PH/30 In these circumstances:
1208 . Two addresses routed to the same list of hosts;
1209 . First host does not offer TLS;
1210 . First host accepts first address;
1211 . First host gives temporary error to second address;
1212 . Second host offers TLS and a TLS session is established;
1213 . Second host accepts second address.
1214 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1215 and peerdn, if requested) that were in fact used only for the second
1218 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1219 attention to any local part prefix or suffix that was matched by the
1220 router that accepted the address. It now behaves in the same way as it
1221 does for delivery: the affixes are removed from the local part unless
1222 rcpt_include_affixes is set on the transport.
1224 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1225 timeout during the DATA phase of an incoming message.
1227 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1228 applied a patch sanctioned by the Sieve maintainer.
1230 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1231 the uid or gid is negative. A case of a negative gid caused this to be
1232 noticed. The fix allows for either to be negative.
1234 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1235 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1238 PH/36 The change PH/12 above was broken. Fixed it.
1240 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1241 the grounds that routing the same address twice would always produce the
1242 same answer. This might have been true once, but it is certainly no
1243 longer true now. Routing a child address may depend on the previous
1244 routing that produced that child. Some complicated redirection strategies
1245 went wrong when messages had multiple recipients, and made Exim's
1246 behaviour dependent on the order in which the addresses were given.
1248 I have moved the duplicate checking until after the routing is complete.
1249 Exim scans the addresses that are assigned to local and remote
1250 transports, and removes any duplicates. This means that more work will be
1251 done, as duplicates will always all be routed, but duplicates are
1252 presumably rare, so I don't expect this is of any significance.
1254 For deliveries to pipes, files, and autoreplies, the duplicate checking
1255 still happens during the routing process, since they are not going to be
1258 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1259 It corrects a timeout issue with spamd. This is Ian's comment: "The
1260 background is that sometimes spamd either never reads data from a
1261 connection it has accepted, or it never writes response data. The exiscan
1262 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1263 blindly assumes that writes won't block so it may never time out."
1265 PH/39 Allow G after quota size as well as K and M.
1267 PH/40 The value set for $authenticated_id in an authenticator may not contain
1268 binary zeroes or newlines because the value is written to log lines and
1269 to spool files. There was no check on this. Now the value is run through
1270 the string_printing() function so that such characters are converted to
1271 printable escape sequences.
1273 PH/41 $message_linecount is a new variable that contains the total number of
1274 lines in the message. Compare $body_linecount, which is the count for the
1277 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1278 bounce and delay warning messages, except in certain special cases, which
1281 (a) An SMTP error message from a remote host;
1282 (b) A message specified in a :fail: redirection;
1283 (c) A message specified in a "fail" command in a system filter;
1284 (d) A message specified in a FAIL return from the queryprogram router;
1285 (e) A message specified by the cannot_route_message router option.
1287 In these cases only, Exim does include the error details in bounce and
1288 warning messages. There are also a few cases where bland messages such
1289 as "unrouteable address" or "local delivery error" are given.
1291 PH/43 $value is now also set for the "else" part of a ${run expansion.
1293 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1294 being worked on, but at least Exim now implements the latest version to
1297 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1298 process to complete was treated as a delivery failure, a timeout while
1299 writing the message to the pipe was logged, but erroneously treated as a
1300 successful delivery. Such timeouts include transport filter timeouts. For
1301 consistency with the overall process timeout, these timeouts are now
1302 treated as errors, giving rise to delivery failures by default. However,
1303 there is now a new Boolean option for the pipe transport called
1304 timeout_defer, which, if set TRUE, converts the failures into defers for
1305 both kinds of timeout. A transport filter timeout is now identified in
1308 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1309 systems where "make" and "gmake" are different, calling "gmake" at top
1310 level broke things. I've arranged for the value of $(MAKE) to be passed
1311 from the Makefile to this script so that it can call the same version of
1315 A note about Exim versions 4.44 and 4.50
1316 ----------------------------------------
1318 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1319 changes of various kinds. As a consequence, a big documentation update was
1320 needed. This delayed the release for rather longer than seemed good, especially
1321 in the light of a couple of (minor) security issues. Therefore, the changes
1322 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1323 release. So 4.44 and 4.50 are in effect two different branches that both start
1326 I have left the 4.50 change log unchanged; it contains all the changes since
1327 4.43. The change log for 4.44 is below; many of its items are identical to
1328 those for 4.50. This seems to be the most sensible way to preserve the
1329 historical information.
1335 1. Minor wording change to the doc/README.SIEVE file.
1337 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1338 computation of the current number of files was incorrect.
1340 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1341 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1344 4. Give more explanation in the error message when the command for a transport
1345 filter fails to execute.
1347 5. There are several places where Exim runs a non-Exim command in a
1348 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1349 was being done only for the command run by the queryprogram router. It is
1350 now done for all such subprocesses. The other cases are: ${run, transport
1351 filters, and the commands run by the lmtp and pipe transports.
1353 6. Added CONFIGURE_GROUP build-time option.
1355 7. Some older OS have a limit of 256 on the maximum number of file
1356 descriptors. Exim was using setrlimit() to set 1000 as a large value
1357 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1358 systems. I've change it so that if it can't get 1000, it tries for 256.
1360 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1361 was an oversight, and furthermore, ever since the addition of extra
1362 controls (e.g. 4.43/32), the checks on when to allow different forms of
1363 "control" were broken. There should now be diagnostics for all cases when a
1364 control that does not make sense is encountered.
1366 9. Added the /retain_sender option to "control=submission".
1368 10. $recipients is now available in the predata ACL (oversight).
1370 11. Tidy the search cache before the fork to do a delivery from a message
1371 received from the command line. Otherwise the child will trigger a lookup
1372 failure and thereby defer the delivery if it tries to use (for example) a
1373 cached ldap connection that the parent has called unbind on.
1375 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1376 of $address_data from the recipient verification was clobbered by the
1377 sender verification.
1379 13. The value of address_data from a sender verification is now available in
1380 $sender_address_data in subsequent conditions in the ACL statement.
1382 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1384 15. Added a new option "connect=<time>" to callout options, to set a different
1387 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1388 was its contents. (It was OK if the option was not defined at all.)
1390 17. A "Completed" log line is now written for messages that are removed from
1391 the spool by the -Mrm option.
1393 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1394 information about exactly what failed.
1396 19. Added -dd to debug only the daemon process.
1398 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1399 handles timeouts, both on the server side and network timeouts. Renamed the
1400 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1402 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1403 It is now set to "smtps".
1405 22. $host_address is now set to the target address during the checking of
1406 ignore_target_hosts.
1408 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1409 being passed; this would have caused $sender_host_name to have been used if
1410 matching the list had actually called for a host name (not very likely,
1411 since this list is usually IP addresses). A host name is now passed as
1414 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1415 code that specifies a non-restarting handler (typically sigaction() in
1416 modern systems) in an attempt to fix a rare and obscure crash bug.
1418 25. Narrowed the window for a race in the daemon that could cause it to ignore
1419 SIGCHLD signals. This is not a major problem, because they are used only to
1420 wake it up if nothing else does.
1422 26. A malformed maildirsize file could cause Exim to calculate negative values
1423 for the mailbox size or file count. Odd effects could occur as a result.
1424 The maildirsize information is now recalculated if the size or filecount
1427 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1428 support for a long time. Removed HAVE_SYS_VFS_H.
1430 28. Installed the latest version of exipick from John Jetmore.
1432 29. In an address list, if the pattern was not a regular expression, an empty
1433 subject address (from a bounce message) matched only if the pattern was an
1434 empty string. Non-empty patterns were not even tested. This was the wrong
1435 because it is perfectly reasonable to use an empty address as part of a
1436 database query. An empty address is now tested by patterns that are
1437 lookups. However, all the other forms of pattern expect the subject to
1438 contain a local part and a domain, and therefore, for them, an empty
1439 address still always fails if the pattern is not itself empty.
1441 30. Exim went into a mad DNS loop when attempting to do a callout where the
1442 host was specified on an smtp transport, and looking it up yielded more
1443 than one IP address.
1445 31. Re-factored the code for checking spool and log partition space into a
1446 function that finds that data and another that does the check. The former
1447 is then used to implement four new variables: $spool_space, $log_space,
1448 $spool_inodes, and $log_inodes.
1450 32. The RFC2047 encoding function was originally intended for short strings
1451 such as real names; it was not keeping to the 75-character limit for
1452 encoded words that the RFC imposes. It now respects the limit, and
1453 generates multiple encoded words if necessary. To be on the safe side, I
1454 have increased the buffer size for the ${rfc2047: expansion operator from
1457 33. It is now permitted to omit both strings after an "if" condition; if the
1458 condition is true, the result is "true". As before, when the second string
1459 is omitted, a false condition yields an empty string. This makes it less
1460 cumbersome to write custom ACL and router conditions.
1462 34. Failure to deliver a bounce message always caused it to be frozen, even if
1463 there was an errors_to setting on the router. The errors_to setting is now
1466 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1467 canonical form (fully expanded) before being placed in
1468 $sender_host_address.
1470 36. The table in the code that translates DNS record types into text (T_A to
1471 "A" for instance) was missing entries for NS and CNAME. It is just possible
1472 that this could have caused confusion if both these types were looked up
1473 for the same domain, because the text type is used as part of Exim's
1474 per-process caching. But the chance of anyone hitting this buglet seems
1477 37. The dnsdb lookup has been extended in a number of ways.
1479 (1) There is a new type, "zns", which walks up the domain tree until it
1480 finds some nameserver records. It should be used with care.
1482 (2) There is a new type, "mxh", which is like "mx" except that it returns
1483 just the host names, not the priorities.
1485 (3) It is now possible to give a list of domains (or IP addresses) to be
1486 looked up. The behaviour when one of the lookups defers can be
1487 controlled by a keyword.
1489 (4) It is now possible to specify the separator character for use when
1490 multiple records are returned.
1492 38. The dnslists ACL condition has been extended: it is now possible to supply
1493 a list of IP addresses and/or domains to be looked up in a particular DNS
1496 39. Added log_selector=+queue_time_overall.
1498 40. When running the queue in the test harness, wait just a tad after forking a
1499 delivery process, to get repeatability of debugging output.
1501 41. Include certificate and key file names in error message when GnuTLS fails
1502 to set them up, because the GnuTLS error message doesn't include the name
1503 of the failing file when there is a problem reading it.
1505 42. Allow both -bf and -bF in the same test run.
1507 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1509 44. The "Exiscan patch" is now merged into the mainline Exim source.
1511 45. Sometimes the final signoff response after QUIT could fail to get
1512 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1513 < 0 before doing a fflush(). This bug looks as though it goes back to the
1514 introduction of TLS in release 3.20, but "sometimes" must have been rare
1515 because the tests only now provoked it.
1517 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1518 (this can affect the format of dates).
1520 47. exim_tidydb, when checking for the continued existence of a message for
1521 which it has found a message-specific retry record, was not finding
1522 messages that were in split spool directories. Consequently, it was
1523 deleting retry records that should have stayed in existence.
1525 48. Steve fixed some bugs in eximstats.
1527 49. The SPA authentication driver was not abandoning authentication and moving
1528 on to the next authenticator when an expansion was forced to fail,
1529 contradicting the general specification for all authenticators. Instead it
1530 was generating a temporary error. It now behaves as specified.
1532 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1533 (the order specifies the preference for clients). The order is now AES256,
1534 AES128, 3DES, ARCFOUR128.
1536 51. Small patch to Sieve code - explicitly set From: when generating an
1539 52. Exim crashed if a remote delivery caused a very long error message to be
1540 recorded - for instance if somebody sent an entire SpamAssassin report back
1541 as a large number of 550 error lines. This bug was coincidentally fixed by
1542 increasing the size of one of Exim's internal buffers (big_buffer) that
1543 happened as part of the Exiscan merge. However, to be on the safe side, I
1544 have made the code more robust (and fixed the comments that describe what
1547 53. Now that there can be additional text after "Completed" in log lines (if
1548 the queue_time_overall log selector is set), a one-byte patch to exigrep
1549 was needed to allow it to recognize "Completed" as not the last thing in
1552 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1553 patch that reportedly fixes this has been added. I am not expert enough to
1554 create a test for it. This is what the patch creator wrote:
1556 "I found a little strange behaviour of ldap code when working with
1557 Windows 2003 AD Domain, where users was placed in more than one
1558 Organization Units. When I tried to give exim partial DN, the exit code
1559 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1560 But simultaneously result of request was absolutely normal ldap result,
1561 so I produce this patch..."
1563 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1564 REFERENCE, so I have modified the code to exclude the patch when that macro
1567 55. Some experimental protocols are using DNS PTR records for new purposes. The
1568 keys for these records are domain names, not reversed IP addresses. The
1569 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1570 leaves it alone. Component reversal etc. now happens only for IP addresses.
1573 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1575 57. Double the size of the debug message buffer (to 2048) so that more of very
1576 long debug lines gets shown.
1578 58. The exicyclog utility now does better if the number of log files to keep
1579 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1581 59. Two changes related to the smtp_active_hostname option:
1583 (1) $smtp_active_hostname is now available as a variable.
1584 (2) The default for smtp_banner uses $smtp_active_hostname instead
1585 of $primary_hostname.
1587 60. The host_aton() function is supposed to be passed a string that is known
1588 to be a valid IP address. However, in the case of IPv6 addresses, it was
1589 not checking this. This is a hostage to fortune. Exim now panics and dies
1590 if the condition is not met. A case was found where this could be provoked
1591 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1592 components; fortuitously, this particular loophole had already been fixed
1593 by change 4.50/55 above.
1595 If there are any other similar loopholes, the new check in host_aton()
1596 itself should stop them being exploited. The report I received stated that
1597 data on the command line could provoke the exploit when Exim was running as
1598 exim, but did not say which command line option was involved. All I could
1599 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1600 running as the user.
1603 61. There was a buffer overflow vulnerability in the SPA authentication code
1604 (which came originally from the Samba project). I have added a test to the
1605 spa_base64_to_bits() function which I hope fixes it.
1608 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1609 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1611 63. The daemon start-up calls getloadavg() while still root for those OS that
1612 need the first call to be done as root, but it missed one case: when
1613 deliver_queue_load_max is set with deliver_drop_privilege. This is
1614 necessary for the benefit of the queue runner, because there is no re-exec
1615 when deliver_drop_privilege is set.
1617 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1618 This has been fixed.
1620 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1621 was in use, was not putting the data itself into the right store pool;
1622 consequently, it could be overwritten for a subsequent message in the same
1623 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1626 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1628 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1629 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1630 the function were treating the return as a boolean value, which happened to
1631 work because 0=false and not-0=true, but is not correct code.
1633 68. The host_aton() function was not handling scoped IPv6 addresses (those
1634 with, for example, "%eth0" on the end) correctly.
1636 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1637 negated items (that is, ~something) in unsigned ints. Some compilers
1638 apparently mutter when there is no cast.
1640 70. If an address verification called from an ACL failed, and did not produce a
1641 user-specific message (i.e. there was only a "system" message), nothing was
1642 put in $acl_verify_message. In this situation, it now puts the system
1645 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1646 session; change 4.31/43 added the unwanted input to the log line - except
1647 that it did not do this in the start of session case. It now does.
1649 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1650 This is wrong and can cause the other end to generate a synchronization
1651 error if it is another Exim or anything else that does the synchronization
1652 check. A QUIT command is no longer sent after a timeout.
1654 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1655 during host lookups.
1657 74. The defer_ok option of callout verification was not working if it was used
1658 when verifying addresses in header lines, that is, for this case:
1660 verify = header_sender/callout=defer_ok
1662 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1663 those file descriptors could be used for SMTP connections. If anything
1664 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1665 could be sent to the SMTP client, causing chaos. The daemon now opens
1666 stdin, stdout, and stderr to /dev/null when it puts itself into the
1669 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1670 log by default. The user can override this with suitable Perl magic.
1672 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1673 the log message when discard triggers, was not working for the DATA ACL or
1674 for the non-SMTP ACL.
1676 78. Error message wording change in sieve.c.
1678 79. If smtp_accept_max_per_host was set, the number of connections could be
1679 restricted to fewer than expected, because the daemon was trying to set up
1680 a new connection before checking whether the processes handling previous
1681 connections had finished. The check for completed processes is now done
1682 earlier. On busy systems, this bug wouldn't be noticed because something
1683 else would have woken the daemon, and it would have reaped the completed
1686 80. If a message was submitted locally by a user whose login name contained one
1687 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1688 It caused a spool format error. I have fixed the spool reading code. A
1689 related problem was that the "from" clause in the Received: line became
1690 illegal because of the space(s). It is now covered by ${quote_local_part.
1692 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1695 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1697 83. Patch from Sieve maintainer to fix unterminated string problem in
1698 "vacation" handling.
1700 84. Some minor changes to the Linux configuration files to help with other
1701 OS variants using glibc.
1703 85. One more patch for Sieve to update vacation handling to latest spec.
1706 ----------------------------------------------------
1707 See the note above about the 4.44 and 4.50 releases.
1708 ----------------------------------------------------
1714 1. Change 4.43/35 introduced a bug that caused file counts to be
1715 incorrectly computed when quota_filecount was set in an appendfile
1718 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1719 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1722 3. Give more explanation in the error message when the command for a transport
1723 filter fails to execute.
1725 4. There are several places where Exim runs a non-Exim command in a
1726 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1727 was being done only for the command run by the queryprogram router. It is
1728 now done for all such subprocesses. The other cases are: ${run, transport
1729 filters, and the commands run by the lmtp and pipe transports.
1731 5. Some older OS have a limit of 256 on the maximum number of file
1732 descriptors. Exim was using setrlimit() to set 1000 as a large value
1733 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1734 systems. I've change it so that if it can't get 1000, it tries for 256.
1736 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1737 was an oversight, and furthermore, ever since the addition of extra
1738 controls (e.g. 4.43/32), the checks on when to allow different forms of
1739 "control" were broken. There should now be diagnostics for all cases when a
1740 control that does not make sense is encountered.
1742 7. $recipients is now available in the predata ACL (oversight).
1744 8. Tidy the search cache before the fork to do a delivery from a message
1745 received from the command line. Otherwise the child will trigger a lookup
1746 failure and thereby defer the delivery if it tries to use (for example) a
1747 cached ldap connection that the parent has called unbind on.
1749 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1750 of $address_data from the recipient verification was clobbered by the
1751 sender verification.
1753 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1754 was its contents. (It was OK if the option was not defined at all.)
1756 11. A "Completed" log line is now written for messages that are removed from
1757 the spool by the -Mrm option.
1759 12. $host_address is now set to the target address during the checking of
1760 ignore_target_hosts.
1762 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1763 being passed; this would have caused $sender_host_name to have been used if
1764 matching the list had actually called for a host name (not very likely,
1765 since this list is usually IP addresses). A host name is now passed as
1768 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1769 code that specifies a non-restarting handler (typically sigaction() in
1770 modern systems) in an attempt to fix a rare and obscure crash bug.
1772 15. Narrowed the window for a race in the daemon that could cause it to ignore
1773 SIGCHLD signals. This is not a major problem, because they are used only to
1774 wake it up if nothing else does.
1776 16. A malformed maildirsize file could cause Exim to calculate negative values
1777 for the mailbox size or file count. Odd effects could occur as a result.
1778 The maildirsize information is now recalculated if the size or filecount
1781 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1782 support for a long time. Removed HAVE_SYS_VFS_H.
1784 18. Updated exipick to current release from John Jetmore.
1786 19. Allow an empty sender to be matched against a lookup in an address list.
1787 Previously the only cases considered were a regular expression, or an
1790 20. Exim went into a mad DNS lookup loop when doing a callout where the
1791 host was specified on the transport, if the DNS lookup yielded more than
1794 21. The RFC2047 encoding function was originally intended for short strings
1795 such as real names; it was not keeping to the 75-character limit for
1796 encoded words that the RFC imposes. It now respects the limit, and
1797 generates multiple encoded words if necessary. To be on the safe side, I
1798 have increased the buffer size for the ${rfc2047: expansion operator from
1801 22. Failure to deliver a bounce message always caused it to be frozen, even if
1802 there was an errors_to setting on the router. The errors_to setting is now
1805 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1806 canonical form (fully expanded) before being placed in
1807 $sender_host_address.
1809 24. Updated eximstats to version 1.33
1811 25. Include certificate and key file names in error message when GnuTLS fails
1812 to set them up, because the GnuTLS error message doesn't include the name
1813 of the failing file when there is a problem reading it.
1815 26. Expand error message when OpenSSL has problems setting up cert/key files.
1818 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1819 (this can affect the format of dates).
1821 28. exim_tidydb, when checking for the continued existence of a message for
1822 which it has found a message-specific retry record, was not finding
1823 messages that were in split spool directories. Consequently, it was
1824 deleting retry records that should have stayed in existence.
1826 29. eximstats updated to version 1.35
1827 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1828 1.35 - bugfix such that pie charts by volume are generated correctly
1830 30. The SPA authentication driver was not abandoning authentication and moving
1831 on to the next authenticator when an expansion was forced to fail,
1832 contradicting the general specification for all authenticators. Instead it
1833 was generating a temporary error. It now behaves as specified.
1835 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1836 (the order specifies the preference for clients). The order is now AES256,
1837 AES128, 3DES, ARCFOUR128.
1839 31. Small patch to Sieve code - explicitly set From: when generating an
1842 32. Exim crashed if a remote delivery caused a very long error message to be
1843 recorded - for instance if somebody sent an entire SpamAssassin report back
1844 as a large number of 550 error lines. This bug was coincidentally fixed by
1845 increasing the size of one of Exim's internal buffers (big_buffer) that
1846 happened as part of the Exiscan merge. However, to be on the safe side, I
1847 have made the code more robust (and fixed the comments that describe what
1850 33. Some experimental protocols are using DNS PTR records for new purposes. The
1851 keys for these records are domain names, not reversed IP addresses. The
1852 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1853 leaves it alone. Component reversal etc. now happens only for IP addresses.
1856 34. The host_aton() function is supposed to be passed a string that is known
1857 to be a valid IP address. However, in the case of IPv6 addresses, it was
1858 not checking this. This is a hostage to fortune. Exim now panics and dies
1859 if the condition is not met. A case was found where this could be provoked
1860 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1861 components; fortuitously, this particular loophole had already been fixed
1862 by change 4.50/55 or 4.44/33 above.
1864 If there are any other similar loopholes, the new check in host_aton()
1865 itself should stop them being exploited. The report I received stated that
1866 data on the command line could provoke the exploit when Exim was running as
1867 exim, but did not say which command line option was involved. All I could
1868 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1869 running as the user.
1872 35. There was a buffer overflow vulnerability in the SPA authentication code
1873 (which came originally from the Samba project). I have added a test to the
1874 spa_base64_to_bits() function which I hope fixes it.
1877 36. The daemon start-up calls getloadavg() while still root for those OS that
1878 need the first call to be done as root, but it missed one case: when
1879 deliver_queue_load_max is set with deliver_drop_privilege. This is
1880 necessary for the benefit of the queue runner, because there is no re-exec
1881 when deliver_drop_privilege is set.
1883 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1884 was in use, was not putting the data itself into the right store pool;
1885 consequently, it could be overwritten for a subsequent message in the same
1886 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1889 38. Sometimes the final signoff response after QUIT could fail to get
1890 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1891 < 0 before doing a fflush(). This bug looks as though it goes back to the
1892 introduction of TLS in release 3.20, but "sometimes" must have been rare
1893 because the tests only now provoked it.
1899 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1900 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1901 Now it returns an integer. A number of calls were still expecting a T/F
1902 return. Fortuitously, in all cases, the tests worked in OK situations,
1903 which is the norm. However, things would have gone wrong on any write
1904 failures on the smtp file descriptor. This function is used when sending
1905 messages over SMTP and also when doing verify callouts.
1907 2. When Exim is called to do synchronous delivery of a locally submitted
1908 message (the -odf or -odi options), it no longer closes stderr before doing
1911 3. Implemented the mua_wrapper option.
1913 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1915 5. Implemented the functions header_remove(), header_testname(),
1916 header_add_at_position(), and receive_remove_recipient(), and exported them
1919 6. If an ACL "warn" statement specified the addition of headers, Exim already
1920 inserted X-ACL-Warn: at the start if there was no header name. However, it
1921 was not making this test for the second and subsequent header lines if
1922 there were newlines in the string. This meant that an invalid header could
1923 be inserted if Exim was badly configured.
1925 7. Allow an ACL "warn" statement to add header lines at the start or after all
1926 the Received: headers, as well as at the end.
1928 8. Added the rcpt_4xx retry error code.
1930 9. Added postmaster_mailfrom=xxx to callout verification option.
1932 10. Added mailfrom=xxxx to the callout verification option, for verify=
1935 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1936 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1938 12. Inserted some casts to stop certain compilers warning when using pointer
1939 differences as field lengths or precisions in printf-type calls (mostly
1940 affecting debugging statements).
1942 13. Added optional readline() support for -be (dynamically loaded).
1944 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1945 same clock tick as a message's arrival, so that its received time was the
1946 same as the "first fail" time on the retry record, and that message
1947 remained on the queue past the ultimate address timeout, every queue runner
1948 would try a delivery (because it was past the ultimate address timeout) but
1949 after another failure, the ultimate address timeout, which should have then
1950 bounced the address, did not kick in. This was a "< instead of <=" error;
1951 in most cases the first failure would have been in the next clock tick
1952 after the received time, and all would be well.
1954 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1955 being recognized when the domain list was tested by the match_domain
1956 condition in an expansion string.
1958 16. Added the ${str2b64: operator.
1960 17. Exim was always calling setrlimit() to set a large limit for the number of
1961 processes, without checking whether the existing limit was already
1962 adequate. (It did check for the limit on file descriptors.) Furthermore,
1963 errors from getrlimit() and setrlimit() were being ignored. Now they are
1964 logged to the main and panic logs, but Exim does carry on, to try to do its
1965 job under whatever limits there are.
1967 18. Imported PCRE 5.0.
1969 19. Trivial typo in log message " temporarily refused connection" (the leading
1972 20. If the log selector return_path_on_delivery was set and an address was
1973 redirected to /dev/null, the delivery process crashed because it assumed
1974 that a return path would always be set for a "successful" delivery. In this
1975 case, the whole delivery is bypassed as an optimization, and therefore no
1978 21. Internal re-arrangement: the function for sending a challenge and reading
1979 a response while authentication was assuming a zero-terminated challenge
1980 string. It's now changed to take a pointer and a length, to allow for
1981 binary data in such strings.
1983 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1985 23. Exim was not respecting finduser_retries when seeking the login of the
1986 uid under which it was called; it was always trying 10 times. (The default
1987 setting of finduser_retries is zero.) Also, it was sleeping after the final
1988 failure, which is pointless.
1990 24. Implemented tls_on_connect_ports.
1992 25. Implemented acl_smtp_predata.
1994 26. If the domain in control=submission is set empty, Exim assumes that the
1995 authenticated id is a complete email address when it generates From: or
1996 Sender: header lines.
1998 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1999 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2000 chown and chgrp in /bin and hostname in /usr/bin.
2002 28. Exim was keeping the "process log" file open after each use, just as it
2003 does for the main log. This opens the possibility of it remaining open for
2004 long periods when the USR1 signal hits a daemon. Occasional processlog
2005 errors were reported, that could have been caused by this. Anyway, it seems
2006 much more sensible not to leave this file open at all, so that is what now
2009 29. The long-running daemon process does not normally write to the log once it
2010 has entered its main loop, and it closes the log before doing so. This is
2011 so that log files can straightforwardly be renamed and moved. However,
2012 there are a couple of unusual error situations where the daemon does write
2013 log entries, and I had neglected to close the log afterwards.
2015 30. The text of an SMTP error response that was received during a remote
2016 delivery was being truncated at 512 bytes. This is too short for some of
2017 the long messages that one sometimes sees. I've increased the limit to
2020 31. It is now possible to make retry rules that apply only when a message has a
2021 specific sender, in particular, an empty sender.
2023 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2024 it possible to be selective about when SMTP synchronization is enforced.
2026 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2028 32. Implemented hosts_connection_nolog.
2030 33. Added an ACL for QUIT.
2032 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2035 35. Added mailbox_size and mailbox_filecount to appendfile.
2037 36. Added control = no_multiline_responses to ACLs.
2039 37. There was a bug in the logic of the code that waits for the clock to tick
2040 in the case where the clock went backwards by a substantial amount such
2041 that the microsecond fraction of "now" was more than the microsecond
2042 fraction of "then" (but the whole seconds number was less).
2044 38. Added support for the libradius Radius client library this is found on
2045 FreeBSD (previously only the radiusclient library was supported).
2051 1. When certain lookups returned multiple values in the form name=value, the
2052 quoting of the values was not always being done properly. Specifically:
2053 (a) If the value started with a double quote, but contained no whitespace,
2055 (b) If the value contained whitespace other than a space character (i.e.
2056 tabs or newlines or carriage returns) it was not quoted.
2057 This fix has been applied to the mysql and pgsql lookups by writing a
2058 separate quoting function and calling it from the lookup code. The fix
2059 should probably also be applied to nisplus, ibase and oracle lookups, but
2060 since I cannot test any of those, I have not disturbed their existing code.
2062 2. A hit in the callout cache for a specific address caused a log line with no
2063 reason for rejecting RCPT. Now it says "Previous (cached) callout
2064 verification failure".
2066 3. There was an off-by-one bug in the queryprogram router. An over-long
2067 return line was truncated at 256 instead of 255 characters, thereby
2068 overflowing its buffer with the terminating zero. As well as fixing this, I
2069 have increased the buffer size to 1024 (and made a note to document this).
2071 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2072 when Exim is waiting for an SMTP response from a remote server, Exim
2073 restarts its select() call on the socket, thereby resetting its timeout.
2074 This is not a problem when such interrupts are rare. Somebody set up a cron
2075 job to run exiwhat every 2 minutes, which is less than the normal select()
2076 timeout (5 or 10 minutes). This meant that the select() timeout never
2077 kicked in because it was always reset. I have fixed this by comparing the
2078 time when an interrupt arrives with the time at the start of the first call
2079 to select(). If more time than the timeout has elapsed, the interrupt is
2080 treated as a timeout.
2082 5. Some internal re-factoring in preparation for the addition of Sieve
2083 extensions (by MH). In particular, the "personal" test is moved to a
2084 separate function, and given an option for scanning Cc: and Bcc: (which is
2085 not set for Exim filters).
2087 6. When Exim created an email address using the login of the caller as the
2088 local part (e.g. when creating a From: or Sender: header line), it was not
2089 quoting the local part when it contained special characters such as @.
2091 7. Installed new OpenBSD configuration files.
2093 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2094 try to make them clearer.
2096 9. Callout options, other than the timeout value, were being ignored when
2097 verifying sender addresses in header lines. For example, when using
2099 verify = header_sender/callout=no_cache
2101 the cache was (incorrectly) being used.
2103 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2104 only the Cygwin environment.
2106 11. When return_path_on_delivery was set as a log selector, if different remote
2107 addresses in the same message used different return paths and parallel
2108 remote delivery occurred, the wrong values would sometimes be logged.
2109 (Whenever a remote delivery process finished, the return path value from
2110 the most recently started remote delivery process was logged.)
2112 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2113 lines when AUTH and/or TLS are in use. This is the "received protocol"
2114 field. Exim used to use "asmtp" for authenticated SMTP, without any
2115 indication (in the protocol name) for TLS use. Now it follows the RFC and
2116 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2117 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2118 names appear in log lines as well as in Received: header lines.
2120 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2121 extensions, and comparison tests, and to fix some bugs.
2123 14. Changes to the "personal" filter test:
2125 (1) The test was buggy in that it was just doing the equivalent of
2126 "contains" tests on header lines. For example, if a user's address was
2127 anne@some.where, the "personal" test would incorrectly be true for
2129 To: susanne@some.where
2131 This test is now done by extracting each address from the header in turn,
2132 and checking the entire address. Other tests that are part of "personal"
2133 are now done using regular expressions (for example, to check local parts
2134 of addresses in From: header lines).
2136 (2) The list of non-personal local parts in From: addresses has been
2137 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2138 taken from the Sieve specification recommendations.
2140 (3) If the message contains any header line starting with "List-" it is
2141 treated as non-personal.
2143 (4) The test for "circular" in the Subject: header line has been removed
2144 because it now seems ill-conceived.
2146 15. Minor typos in src/EDITME comments corrected.
2148 16. Installed latest exipick from John Jetmore.
2150 17. If headers_add on a router specified a text string that was too long for
2151 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2152 of string_sprintf() is now avoided.
2154 18. $message_body_size was not set (it was always zero) when running the DATA
2155 ACL and the local_scan() function.
2157 19. For the "mail" command in an Exim filter, no default was being set for
2158 the once_repeat time, causing a random time value to be used if "once" was
2159 specified. (If the value happened to be <= 0, no repeat happened.) The
2160 default is now 0s, meaning "never repeat". The "vacation" command was OK
2161 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2162 (I found it when inspecting the code).
2164 20. There is now an overall timeout for performing a callout verification. It
2165 defaults to 4 times the callout timeout, which applies to individual SMTP
2166 commands during the callout. The overall timeout applies when there is more
2167 than one host that can be tried. The timeout is checked before trying the
2168 next host. This prevents very long delays if there are a large number of
2169 hosts and all are timing out (e.g. when the network connections are timing
2170 out). The value of the overall timeout can be changed by specifying an
2171 additional sub-option for "callout", called "maxwait". For example:
2173 verify = sender/callout=5s,maxwait=20s
2175 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2176 to the end before writing, but this should make it even safer).
2178 22. Exim was forgetting that it had advertised PIPELINING for the second and
2179 subsequent messages on an SMTP connection. It was also not resetting its
2180 memory on STARTTLS and an internal HELO.
2182 23. When Exim logs an SMTP synchronization error within a session, it now
2183 records whether PIPELINING has been advertised or not.
2185 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2186 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2187 rather than long int.
2189 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2191 26. Added the never_mail option to autoreply.
2197 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2198 crash if the getsockname() call failed; this can happen if a connection is
2199 closed very soon after it is established. The problem was simply in the
2200 order in which certain operations were done, causing Exim to try to write
2201 to the SMTP stream before it had set up the file descriptor. The bug has
2202 been fixed by making things happen in the correct order.
2208 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2209 before the connection was closed, thus losing the rejection response.
2211 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2212 some early Solaris releases, but causes trouble in current releases where
2213 socklen_t is defined.
2215 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2218 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2221 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2222 that releasing the top of it at the end releases what was used for sub-
2223 expansions (unless the block got too big). However, discard this block if
2224 the first thing is a variable or header, so that we can use its block when
2225 it is dynamic (useful for very large $message_headers, for example).
2227 6. Lookups now cache *every* query, not just the most recent. A new, separate
2228 store pool is used for this. It can be recovered when all lookup caches are
2229 flushed. Lookups now release memory at the end of their result strings.
2230 This has involved some general refactoring of the lookup sources.
2232 7. Some code has been added to the store_xxx() functions to reduce the amount
2233 of flapping under certain conditions.
2235 8. log_incoming_interface used to affect only the <= reception log lines. Now
2236 it causes the local interface and port to be added to several more SMTP log
2237 lines, for example "SMTP connection from", and rejection lines.
2239 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2241 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2243 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2244 could be overwritten at the end of the current message (or the start of a
2245 new message if it was set in a HELO ACL). The value is now preserved for
2246 the duration of the SMTP connection.
2248 12. If a transport had a headers_rewrite setting, and a matching header line
2249 contained an unqualified address, that address was qualified, even if it
2250 did not match any rewriting rules. The underlying bug was that the values
2251 of the flags that permit the existence of unqualified sender and recipient
2252 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2253 non-local messages, and by -bnq for local messages) were not being
2254 preserved with the message after it was received.
2256 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2257 "next input=" as part of the text comprising the host identity instead of
2258 the correct text. The code was using the same buffer for two different
2259 strings. However, depending on which order the printing function evaluated
2260 its arguments, the bug did not always show up. Under Linux, for example, my
2261 test suite worked just fine.
2263 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2264 doesn't work with some older versions of Perl. It has been changed to "my",
2265 which in any case is probably the better facility to use.
2267 15. A really picky compiler found some instances of statements for creating
2268 error messages that either had too many or two few arguments for the format
2271 16. The size of the buffer for calls to the DNS resolver has been increased
2272 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2273 for addresses that have a lot of PTR records. This alleviates a problem; it
2274 does not fully solve it.
2276 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2277 buffer now truncates the list and logs the incident, which is the same
2278 action as happens when Exim is looking up a host name and its aliases.
2279 Previously in this situation something unpredictable would happen;
2280 sometimes it was "internal error: store_reset failed".
2282 18. If a server dropped the connection unexpectedly when an Exim client was
2283 using GnuTLS and trying to read a response, the client delivery process
2284 crashed while trying to generate an error log message.
2286 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2287 string, for example:
2289 warn message = H1: something\nH2: something
2291 the text was added as a single header line from Exim's point of view
2292 though it ended up OK in the delivered message. However, searching for the
2293 second and subsequent header lines using $h_h2: did not work. This has been
2294 fixed. Similarly, if a system filter added multiple headers in this way,
2295 the routers could not see them.
2297 20. Expanded the error message when iplsearch is called with an invalid key to
2298 suggest using net-iplsearch in a host list.
2300 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2301 ACLs are no longer actually imposed (and a message to that effect is
2304 22. If a "gecos" field in a passwd entry contained escaped characters, in
2305 particular, if it contained a \" sequence, Exim got it wrong when building
2306 a From: or a Sender: header from that name. A second bug also caused
2307 incorrect handling when an unquoted " was present following a character
2308 that needed quoting.
2310 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2311 was not being matched caselessly.
2313 24. Arranged for all hyphens in the exim.8 source to be escaped with
2316 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2317 properly. Recipient callout cache records were still being keyed to include
2318 the sender, even when use_sender was set false. This led to far more
2319 callouts that were necessary. The sender is no longer included in the key
2320 when use_sender is false.
2322 26. Added "control = submission" modifier to ACLs.
2324 27. Added the ${base62d: operator to decode base 62 numbers.
2326 28. dnsdb lookups can now access SRV records.
2328 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2329 the configuration file.
2331 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2332 (-v) mode. This makes the output for a verbose queue run more intelligible.
2334 31. Added a use_postmaster feature to recipient callouts.
2336 32. Added the $body_zerocount variable, containing the number of binary zero
2337 bytes in the message body.
2339 33. The time of last modification of the "new" subdirectory is now used as the
2340 "mailbox time last read" when there is a quota error for a maildir
2343 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2345 35. Added +ignore_unknown as a special item in host lists.
2347 36. Code for decoding IPv6 addresses in host lists is now included, even if
2348 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2349 address was recognized as an IP address, but was then not correctly decoded
2350 into binary, causing unexpected and incorrect effects when compared with
2357 1. Very minor rewording of debugging text in manualroute to say "list of
2358 hosts" instead of "hostlist".
2360 2. If verify=header_syntax was set, and a header line with an unqualified
2361 address (no domain) and a large number of spaces between the end of the
2362 name and the colon was received, the reception process suffered a buffer
2363 overflow, and (when I tested it) crashed. This was caused by some obsolete
2364 code that should have been removed. The fix is to remove it!
2366 3. When running in the test harness, delay a bit after writing a bounce
2367 message to get a bit more predictability in the log output.
2369 4. Added a call to search_tidyup() just before forking a reception process. In
2370 theory, someone could use a lookup in the expansion of smtp_accept_max_
2371 per_host which, without the tidyup, could leave open a database connection.
2373 5. Added the variables $recipient_data and $sender_data which get set from a
2374 lookup success in an ACL "recipients" or "senders" condition, or a router
2375 "senders" option, similar to $domain_data and $local_part_data.
2377 6. Moved the writing of debug_print from before to after the "senders" test
2380 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2381 problems for message scanning, either using a data ACL, or using
2382 local_scan() because the Received: header was not generated till after they
2383 were called (in order to set the time as the time of reception completion).
2384 I have revised the way this works. The header is now generated after the
2385 body is received, but before the ACL or local_scan() are called. After they
2386 are run, the timestamp in the header is updated.
2392 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2393 before starting a queue runner without re-exec. This happened only when
2394 deliver_drop_privilege was set or when the Exim user was set to root. The
2395 effect of the bug was that timeouts during subsequent deliveries caused
2396 crashes instead of being properly handled. The handler is now left at its
2397 default (and expected) setting.
2399 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2400 message, again when deliver_drop_privilege is set or Exim is run as root.
2401 The bug described in (1) was not present in this case, but the tidying up
2402 of the other signals was missing. I have made the two cases consistent.
2404 3. The ignore_target_hosts setting on a manualroute router was being ignored
2405 for hosts that were looked up using the /MX notation.
2407 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2410 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2411 operated on the sender address. After changing the $sender_address to <>
2412 for the sender address verify, Exim was re-instated it as the original
2413 (before rewriting) address, but remembering that it had rewritten it, so it
2414 wasn't rewriting it again. This bug also had the effect of breaking the
2415 sender address verification caching when the sender address was rewritten.
2417 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2418 This has been changed so that if the ip literal address matches
2419 ignore_target_hosts, the router declines.
2421 7. Added expansion conditions match_domain, match_address, and match_local_
2422 part (NOT match_host).
2424 8. The placeholder for the Received: header didn't have a length field set.
2426 9. Added code to Exim itself and to exim_lock to test for a specific race
2427 condition that could lead to file corruption when using MBX delivery. The
2428 issue is with the lockfile that is created in /tmp. If this file is removed
2429 after a process has opened it but before that process has acquired a lock,
2430 there is the potential for a second process to recreate the file and also
2431 acquire a lock. This could lead to two Exim processes writing to the file
2432 at the same time. The added code performs the same test as UW imapd; it
2433 checks after acquiring the lock that its file descriptor still refers to
2434 the same named file.
2436 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2437 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2440 11. Added the smtp_active_hostname option. If used, this will typically be made
2441 to depend on the incoming interface address. Because $interface_address is
2442 not set up until the daemon has forked a reception process, error responses
2443 that can happen earlier (such as "too many connections") no longer contain
2446 12. If an expansion in a condition on a "warn" statement fails because a lookup
2447 defers, the "warn" statement is abandoned, and the next ACL statement is
2448 processed. Previously this caused the whole ACL to be aborted.
2450 13. Added the iplsearch lookup type.
2452 14. Added ident_timeout as a log selector.
2454 15. Added tls_certificate_verified as a log selector.
2456 16. Added a global option tls_require_ciphers (compare the smtp transport
2457 option of the same name). This controls incoming TLS connections.
2459 17. I finally figured out how to make tls_require_ciphers do a similar thing
2460 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2461 before starting the TLS session.
2463 18. Tabs are now shown as \t in -bP output.
2465 19. If the log selector return_path_on_delivery was set, Exim crashed when
2466 bouncing a message because it had too many Received: header lines.
2468 20. If two routers both had headers_remove settings, and the first one included
2469 a superfluous trailing colon, the final name in the first list and the
2470 first name in the second list were incorrectly joined into one item (with a
2471 colon in the middle).
2477 1. Added -C and -D options to the exinext utility, mainly to make it easier
2478 to include in the automated testing, but these could be helpful when
2479 multiple configurations are in use.
2481 2. The exinext utility was not formatting the output nicely when there was
2482 an alternate port involved in the retry record key, nor when there was a
2483 message id as well (for retries that were specific to a specific message
2484 and a specific host). It was also confused by IPv6 addresses, because of
2485 the additional colons they contain. I have fixed the IPv4 problem, and
2486 patched it up to do a reasonable job for IPv6.
2488 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2489 delivery, the log line now contains "pipelined" if PIPELINING was used.
2491 4. An SMTP transport process used to panic and die if the bind() call to set
2492 an explicit outgoing interface failed. This has been changed; it is now
2493 treated in the same way as a connect() failure.
2495 5. A reference to $sender_host_name in the part of a conditional expansion
2496 that was being skipped was still causing a DNS lookup. This no longer
2499 6. The def: expansion condition was not recognizing references to header lines
2500 that used bh_ and bheader_.
2502 7. Added the _cache feature to named lists.
2504 8. The code for checking quota_filecount in the appendfile transport was
2505 allowing one more file than it should have been.
2507 9. For compatibility with Sendmail, the command line option
2515 and sets the incoming protocol and host name (for trusted callers). The
2516 host name and its colon can be omitted when only the protocol is to be set.
2517 Note the Exim already has two private options, -pd and -ps, that refer to
2518 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2519 "s", but I don't think that's a major issue.
2521 10. A number of refactoring changes to the code, none of which should affect
2524 (a) The number of logging options was getting close to filling up the
2525 32-bit word that was used as a bit map. I have split them into two classes:
2526 those that are passed in the argument to log_write(), and those that are
2527 only ever tested independently outside of that function. These are now in
2528 separate 32-bit words, so there is plenty of room for expansion again.
2529 There is no change in the user interface or the logging behaviour.
2531 (b) When building, for example, log lines, the code previously used a
2532 macro that called string_cat() twice, in order to add two strings. This is
2533 not really sufficiently general. Furthermore, there was one instance where
2534 it was actually wrong because one of the argument was used twice, and in
2535 one call a function was used. (As it happened, calling the function twice
2536 did not affect the overall behaviour.) The macro has been replaced by a
2537 function that can join an arbitrary number of extra strings onto a growing
2540 (c) The code for expansion conditions now uses a table and a binary chop
2541 instead of a serial search (which was left over from when there were very
2542 few conditions). Also, it now recognizes conditions like "pam" even when
2543 the relevant support is not compiled in: a suitably worded error message is
2544 given if an attempt is made to use such a condition.
2546 11. Added ${time_interval:xxxxx}.
2548 12. A bug was causing one of the ddress fields not to be passed back correctly
2549 from remote delivery subprocesses. The field in question was not being
2550 subsequently used, so this caused to problems in practice.
2552 13. Added new log selectors queue_time and deliver_time.
2554 14. Might have fixed a bug in maildirsizefile handling that threw up
2555 "unexpected character" debug warnings, and recalculated the data
2556 unnecessarily. In any case, I expanded the warning message to give more
2559 15. Added the message "Restricted characters in address" to the statements in
2560 the default ACL that block characters like @ and % in local parts.
2562 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2563 Three changes have been made:
2565 (a) There was a serious bug; a negative response to MAIL caused the whole
2566 recipient domain to be cached as invalid, thereby blocking all messages
2567 to all local parts at the same domain, from all senders. This bug has
2568 been fixed. The domain is no longer cached after a negative response to
2569 MAIL if the sender used is not empty.
2571 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2574 (c) A new callout option, "use_sender" has been added for people who want
2575 the modified behaviour.
2581 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2584 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2585 indeed breaks things for older releases.
2587 3. Added additional logging to the case where there is a problem reading data
2588 from a filter that is running in a subprocess using a pipe, in order to
2589 try to track down a specific problem.
2591 4. Testing facility fudge: when running in the test harness and attempting
2592 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2593 getting "No route to host". Convert this to a timeout.
2595 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2598 6. Some OS don't have socklen_t but use size_t instead. This affects the
2599 fifth argument of getsockopt() amongst other things. This is now
2600 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2601 can be set for individual OS. I have set it for SunOS5, OSF1, and
2602 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2603 some earlier ones do not.
2605 7. Change 4.30/15 was not doing the test caselessly.
2607 8. The standard form for an IPv6 address literal was being rejected by address
2608 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2609 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2610 this, as well as the form without the "IPv6" on the front (but only when
2611 address literals are enabled, of course).
2613 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2615 10. Exim crashed if a message with an empty sender address specified by -f
2616 encountered a router with an errors_to setting. This could be provoked only
2617 by a command such as
2621 where an empty string was supplied; "<>" did not hit this bug.
2623 11. Installed PCRE release 4.5.
2625 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2626 remained set. It is now erased.
2628 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2629 times from message ids (which are base 36 rather than the normal 62).
2631 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2632 were being counted as actual protocol errors, and logged if the log
2633 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2634 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2635 and DATA following a set of rejected RCPTs do not count as protocol errors.
2636 In other words, Exim assumes they were pipelined, though this may not
2637 actually be the case. Of course, in all cases the client gets an
2638 appropriate error code.
2640 15. If a lookup fails in an ACL condition, a message about the failure may
2641 be available; it is used if testing the ACL cannot continue, because most
2642 such messages specify what the cause of the deferral is. However, some
2643 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2644 that caused an old message to be retained and used if a later statement
2645 caused a defer, replacing the real cause of the deferral.
2647 16. If an IP address had so many PTR records that the DNS lookup buffer
2648 was not large enough to hold them, Exim could crash while trying to process
2649 the truncated data. It now detects and logs this case.
2651 17. Further to 4.21/58, another change has been made: if (and only if) the
2652 first line of a message (the first header line) ends with CRLF, a bare LF
2653 in a subsequent header line has a space inserted after it, so as not to
2654 terminate the header.
2656 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2657 unnecessarily, used atoi() instead of strtol(), and didn't check the
2658 termination when getting file sizes from file names by regex.
2660 19. Completely re-implemented the support for maildirsize files, in the light
2661 of a number of problems with the previous contributed implementation
2662 (4.30/29). In particular:
2664 . If the quota is zero, the maildirsize file is maintained, but no quota is
2667 . If the maildir directory does not exist, it is created before any attempt
2668 to write a maildirsize file.
2670 . The quota value in the file is just a cache; if the quota is changed in
2671 the transport, the new value overrides.
2673 . A regular expression is available for excluding directories from the
2676 20. The autoreply transport checks the characters in options that define the
2677 message's headers; it allows continued headers, but it was checking with
2678 isspace() after an embedded newline instead of explicitly looking for a
2681 21. If all the "regular" hosts to which an address was routed had passed their
2682 expiry times, and had not reached their retry times, the address was
2683 bounced, even if fallback hosts were defined. Now Exim should go on to try
2686 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2687 equivalent code in the SMTP transport. Some hosts send humungous responses
2688 to HELO/EHLO, more than 1024 it seems.
2690 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2691 gives compiler warnings in some environments. I've now done it "properly",
2694 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2695 (because of problems with the built-in one) was declared to return uschar *
2696 instead of char *, causing compiler failure.
2698 25. Fixed a file descriptor leak when processing alias/forward files.
2700 26. Fixed a minor format string issue in dbfn.c.
2702 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2704 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2705 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2708 29. When a sender address is verified, it is cached, to save repeating the test
2709 when there is more than one recipient in a message. However, when the
2710 verification involves a callout, it is possible for different callout
2711 options to be set for different recipients. It is too complicated to keep
2712 track of this in the cache, so now Exim always runs a verification when a
2713 callout is required, relying on the callout cache for the optimization.
2714 The overhead is duplication of the address routing, but this should not be
2717 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2718 to be verified with callout=postmaster, and the main callout worked but the
2719 postmaster check failed, the verification correctly failed. However, if a
2720 subsequent RCPT command asked for sender verification *without* the
2721 postmaster check, incorrect caching caused this verification also to fail,
2724 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2725 it was not caching the DNS options (qualify_single, search_parents) that
2726 were used when the lookup failed. A subsequent lookup with different
2727 options therefore always gave the same answer, though there were cases
2728 where it should not have. (Example: a "domains = !$mx_any" option on a
2729 dnslookup router: the "domains" option is always processed without any
2730 widening, but the router might have qualify_single set.) Now Exim uses the
2731 cached value only when the same options are set.
2733 32. Added John Jetmore's "exipick" utility to the distribution.
2735 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2736 than a timeout (e.g. a certificate is required, and is not provided), an
2737 Exim server now closes the connection immediately. Previously it waited for
2738 the client to close - but if the client is SSL, it seems that they each
2739 wait for each other, leading to a delay before one of them times out.
2741 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2742 maintained 0.8.x compatibility because I don't think many are using it, and
2743 it is clearly obsolete.
2745 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2748 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2749 client certificate was expired. A simple patch fixes this, though I don't
2750 understand the full logic of why the verify callback is called multiple
2753 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2754 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2755 which causes problems with some clients (such as the Certicom SSL Plus
2756 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2757 disables the coutermeasure allowing Eudora to connect."
2759 38. Exim was not checking that a write() to a log file succeeded. This could
2760 lead to Bad Things if a log got too big, in particular if it hit a file
2761 size limit. Exim now panics and dies if it cannot write to a log file, just
2762 as it does if it cannot open a log file.
2764 39. Modified OS/Makefile-Linux so that it now contains
2766 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2768 The two -D definitions ensure that Exim is compiled with large file
2769 support, which makes it possible to handle log files that are bigger than
2772 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2773 instance) a domain was checked against a named list that involved a lookup,
2774 causing $domain_data to be set, then another domain was checked against the
2775 same list, then the first domain was re-checked, the value of $domain_data
2776 after the final check could be wrong. In particular, if the second check
2777 failed, it could be set empty. This bug probably also applied to
2780 41. The strip_trailing_dot option was not being applied to the address given
2781 with the -f command-line option.
2783 42. The code for reading a message's header from the spool was incrementing
2784 $received_count, but never initializing it. This meant that the value was
2785 incorrect (doubled) while delivering a message in the same process in which
2786 it was received. In the most common configuration of Exim, this never
2787 happens - a fresh exec is done - but it can happen when
2788 deliver_drop_privilege is set.
2790 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2791 it now includes up to 150 characters of the unexpected data in the log
2794 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2795 and building data strings. The size of both of these buffers was 10 000
2796 bytes - far larger than anybody would *ever* want, thought I. Needless to
2797 say, somebody hit the limit. I have increased the maximum line length to
2798 20 000 and the maximum data length of concatenated lines to 100 000. I have
2799 also fixed two bugs, because there was no checking on these buffers. Tsk,
2800 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2801 buffer is too small.
2803 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2804 lsearch lookups. Now it does.
2806 46. When parsing a route_list item in a manualroute router, a fixed-length
2807 buffer was used for the list of hosts. I made this 1024 bytes long,
2808 thinking that nobody would ever have a list of hosts that long. Wrong.
2809 Somebody had a whole pile of complicated expansion conditions, and the
2810 string was silently truncated, leading to an expansion error. It turns out
2811 that it is easier to change to an unlimited length (owing to other changes
2812 that have happened since this code was originally written) than to build
2813 structure for giving a limitation error. The length of the item that
2814 expands into the list of hosts is now unlimited.
2816 47. The lsearch lookup could not handle data where the length of text line was
2817 more than 4095 characters. Such lines were truncated, leading to shortened
2818 data being returned. It should now handle lines of any length.
2820 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2821 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2824 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2825 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2826 no longer happens in comment lines. A list of replacements is now placed
2827 at the head of all of the source files, except those whose only change is
2828 to replace PERL_COMMAND in the very first #! line.
2830 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2831 messages on the queue, with a bottom-up merge sort, using code contributed
2832 by Michael Haardt. This should make operations like -bp somewhat faster on
2833 large queues. It won't affect queue runners, except when queue_run_in_order
2836 51. Installed eximstats 1.31 in the distribution.
2838 52. Added support for SRV lookups to the dnslookup router.
2840 53. If an ACL referred to $message_body or $message_body_end, the value was not
2841 reset for any messages that followed in the same SMTP session.
2843 54. The store-handling optimization for building very long strings was not
2844 differentiating between the different store pools. I don't think this
2845 actually made any difference in practice, but I've tidied it.
2847 55. While running the routers to verify a sender address, $sender_address
2848 was still set to the sender address. This is wrong, because when routing to
2849 send a bounce to the sender, it would be empty. Therefore, I have changed
2850 it so that, while verifying a sender address, $sender_address is set to <>.
2851 (There is no change to what happens when verifying a recipient address.)
2853 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2854 target A or AAAA records (if not already returned) without resetting the
2855 qualify_single or search_parents options of the DNS resolver. These are
2856 inappropriate in this case because the targets of MX and SRV records must
2857 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2858 target that, when qualified, matched something in the local domain. These
2859 two options are now turned off when doing these lookups.
2861 57. It seems that at least some releases of Reiserfs (which does not have the
2862 concept of a fixed number of inodes) returns zero and not -1 for the
2863 number of available inodes. This interacted badly with check_spool_inodes,
2864 which assumed that -1 was the "no such thing" setting. What I have done is
2865 to check that the total number of inodes is greater than zero before doing
2866 the test of how many are available.
2868 58. When a "warn" ACL statement has a log_message modifier, the message is
2869 remembered, and not repeated. This is to avoid a lot of repetition when a
2870 message has many recipients that cause the same warning to be written.
2871 Howewer, Exim was preserving the list of already written lines for an
2872 entire SMTP session, which doesn't seem right. The memory is now reset if a
2873 new message is started.
2875 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2876 debugging output unless log_rewrite was also set.
2878 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2879 of (int)(handle) when we know that handle contains (void *)(-1).
2881 61. The Exim daemon panic-logs an error return when it closes the incoming
2882 connection. However "connection reset by peer" seems to be common, and
2883 isn't really an error worthy of noting specially, so that particular error
2886 62. When Exim is trying to find all the local interfaces, it used to panic and
2887 die if the ioctl to get the interface flags failed. However, it seems that
2888 on at least one OS (Solaris 9) it is possible to have an interface that is
2889 included in the list of interfaces, but for which you get a failure error
2890 for this call. This happens when the interface is not "plumbed" into a
2891 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2892 failure of the "get flags" call assumes that the interface is down.
2894 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2895 makes life easier for people who are doing arithmetic on fields extracted
2896 from dates, where you often get leading zeros that should not be
2897 interpreted as octal.
2899 64. Added qualify_domain to the redirect router, to override the global
2902 65. If a pathologically long header line contained very many addresses (the
2903 report of this problem mentioned 10 000) and each of them was rewritten,
2904 Exim could use up a very large amount of memory. (It kept on making new
2905 copies of the header line as it rewrote, and never released the old ones.)
2906 At the expense of a bit more processing, the header rewriting function has
2907 been changed so that it no longer eats memory in this way.
2909 66. The generation of the Received: header has been moved from the time that a
2910 message starts to be received, to the time that it finishes. The timestamp
2911 in the Received: header should now be very close to that of the <= log
2912 line. There are two side-effects of this change:
2914 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2915 logged header lines no longer include the local Received: line, because
2916 it has not yet been created. The same applies to a copy of the message
2917 that is returned to a non-SMTP sender when a message is rejected.
2919 (b) When a filter file is tested using -bf, no additional Received: header
2920 is added to the test message. After some thought, I decided that this
2923 This change does not affect the value of $received_for. It is still set
2924 after address rewriting, but before local_scan() is called.
2926 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2928 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2929 gave an unhelpful panic error message, and a defer error. I have managed to
2930 change this behaviour so that it now rejects any supplied certificate,
2931 which seems right, as the list of acceptable certificates is empty.
2933 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2934 gave an unhelpful defer error. I have not managed to make this reject any
2935 supplied certificates, but the error message it gives is "no certificate
2936 supplied", which is not helpful.
2938 70. exigrep's output now also includes lines that are not associated with any
2939 message, but which match the given pattern. Implemented by a patch from
2940 Martin Sluka, which also tidied up the Perl a bit.
2942 71. Recipient callout verification, like sender verification, was using <> in
2943 the MAIL FROM command. This isn't really the right thing, since the actual
2944 sender may affect whether the remote host accepts the recipient or not. I
2945 have changed it to use the actual sender in the callout; this means that
2946 the cache record is now keyed on a recipient/sender pair, not just the
2947 recipient address. There doesn't seem to be a real danger of callout loops,
2948 since a callout by the remote host to check the sender would use <>.
2949 [SEE ABOVE: changed after hitting problems.]
2951 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2952 temporary errors. However, in the case of such a code being given after
2953 the end of a data transmission (i.e. after ".") Exim was failing to write
2954 a retry record for the message. (Yes, there was some broken host that was
2955 actually sending 8xx at this point.)
2957 73. An unknown lookup type in a host list could cause Exim to panic-die when
2958 the list was checked. (An example that provoked this was putting <; in the
2959 middle of a list instead of at the start.) If this happened during a DATA
2960 ACL check, a -D file could be left lying around. This kind of configuration
2961 error no longer causes Exim to die; instead it causes a defer errror. The
2962 incident is still logged to the main and panic logs.
2964 74. Buglet left over from Exim 3 conversion. The message "too many messages
2965 in one connection" was written to the rejectlog but not the mainlog, except
2966 when address rewriting (yes!) was being logged.
2968 75. Added write_rejectlog option.
2970 76. When a system filter was run not as root (that is, when system_filter_user
2971 was set), the values of the $n variables were not being returned to the
2972 main process; thus, they were not subsequently available in the $sn
2975 77. Added +return_path_on_delivery log selector.
2977 78. A connection timeout was being treated differently from recipients deferred
2978 when testing hosts_max_try with a message that was older than the host's
2979 retry timeout. (The host should not be counted, thus allowing all hosts to
2980 be tried at least once before bouncing.) This may have been the cause of an
2981 occasionally reported bug whereby a message would remain on the queue
2982 longer than the retry timeout, but would be bounced if a delivery was
2983 forced. I say "may" because I never totally pinned down the problem;
2984 setting up timeout/retry tests is difficult. See also the next item.
2986 79. The ultimate address timeout was not being applied to errors that involved
2987 a combination of host plus message (for example, a timeout on a MAIL
2988 command). When an address resolved to a number of possible hosts, and they
2989 were not all tried for each delivery (e.g. because of hosts_max_try), a
2990 message could remain on the queue longer than the retry timeout.
2992 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2995 81. Fixed an obscure SMTP outgoing bug which required at least the following
2996 conditions: (a) there was another message waiting for the same server;
2997 (b) the server returned 5xx to all RCPT commands in the first message so
2998 that the message was not completed; (c) the server dropped the connection
2999 or gave a negative response to the RSET that Exim sends to abort the
3000 transaction. The observed case was a dropped connection after DATA that had
3001 been sent in pipelining mode. That is, the server had advertised PIPELINING
3002 but was not implementing it correctly. The effect of the bug was incorrect
3003 behaviour, such as trying another host, and this could lead to a crash.
3009 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3010 and daemon.c were passed as pointers to ints; they should have been
3011 pointers to socklen_t variables (which are typically unsigned ints).
3013 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3016 3. Fixed a really odd bug that affected only the testing scheme; patching a
3017 certain fixed string in the binary changed the value of another string that
3018 happened to be identical to the end of the original first string.
3020 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3021 name", it returns that address as the IP address. On some operating
3022 systems (e.g. Solaris), it also passes back the IP address string as the
3023 "host name". However, on others (e.g. Linux), it passes back an empty
3024 string. Exim wasn't checking for this, and was changing the host name to an
3025 empty string, assuming it had been canonicized.
3027 5. Although rare, it is permitted to have more than one PTR record for a given
3028 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3029 all the names associated with an address, because they do in Solaris.
3030 However, it seems that they do not in Linux for data that comes from the
3031 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3032 I found this out when I moved to a new Linux workstation and tried to run
3033 the Exim test suite.
3035 To get round this problem I have changed the code so that it now does its
3036 own call to the DNS to look up PTR records when searching for a host name.
3037 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3038 addresses that are only in /etc/hosts are still found.
3040 This behaviour is, however, controlled by an option called host_lookup_
3041 order, which defaults to "bydns:byaddr". If people want to use the other
3042 order, or indeed, just use one or the other means of lookup, they can
3043 specify it in this variable.
3045 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3046 some operating systems, this comes back from gethostbyaddr() as an empty
3047 string, and this is what Exim used to test for. However, it seems that in
3048 other systems, "." is yielded. Exim now tests for this case too.
3050 7. The values of check_spool_space and check_log_space are now held internally
3051 as a number of kilobytes instead of an absolute number of bytes. If a
3052 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3053 kilobyte. This means that much larger values can be stored.
3055 8. Exim monitor: an attempt to get the action menu when not actually pointing
3056 at a message produces an empty menu entitled "No message selected". This
3057 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3058 no entries in it ("Shell widget menu has zero width and/or height"). So I
3059 have added a single, blank menu entry in this case.
3061 9. Added ${quote_local_part.
3063 10. MIME decoding is now applied to the contents of Subject: header lines when
3066 11. Now that a reference to $sender_host_address automatically causes a reverse
3067 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3068 host lookup before query-style lookups in lists that might use this
3069 variable. This has therefore been abolished, and the "net-" prefix is no
3070 longer necessary for query-style lookups.
3072 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3073 have been a typo for LFLAGS, so it has been changed.
3075 13. The install script calls Exim with "-C /dev/null" in order to find the
3076 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3077 to be output. Howeve, since Exim outputs its version number before the
3078 error, it didn't break the script. It just looked ugly. I fixed this by
3079 always allowing "-C /dev/null" if the caller is root.
3081 14. Ignore overlarge ACL variable number when reading spool file - insurance
3082 against a later release with more variables having written the file.
3084 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3085 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3086 this, as well as the form without the "IPv6" on the front.
3088 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3089 OS/Makefile-Darwin file.
3091 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3092 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3094 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3095 variable. However, some OS use ints for inodes. I've added cast to long int
3096 to get rid of the compiler warning.
3098 19. I had forgotten to lock out "/../" in configuration file names when
3099 ALT_CONFIG_PREFIX was set.
3101 20. Routers used for verification do not need to specify transports. However,
3102 if such a router generated a host list, and callout was configured, Exim
3103 crashed, because it could not find a port number from the (non-existent)
3104 transport. It now assumes port 25 in this circumstance.
3106 21. Added the -t option to exigrep.
3108 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3109 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3110 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3111 them is set, LOOKUP_LSEARCH is forced.
3113 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3115 24. Added sender and host information to the "rejected by local_scan()" log
3116 line; previously there was no indication of these.
3118 25. Added .include_if_exists.
3120 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3121 Linux. It turns out that not all file systems support this. Apparently some
3122 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3123 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3124 not supported on the file descriptor, is now ignored when Exim is trying to
3125 sync a directory. This applies only to Linux.
3127 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3129 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3130 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3133 29. Added support for maildirsize files from supplied patch (modified a bit).
3135 30. The use of :fail: followed by an empty string could lead Exim to respond to
3136 sender verification failures with (e.g.):
3138 550 Verification failed for <xxx>
3139 550 Sender verify failed
3141 where the first response line was missing the '-' that indicates it is not
3142 the final line of the response.
3144 31. The loop for finding the name of the user that called Exim had a hardwired
3145 limit of 10; it now uses the value of finduser_retries, which is used for
3146 all other user lookups.
3148 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3151 33. Exim was neglecting to zero errno before one call of strtol() when
3152 expanding a string and expecting an integer value. On some systems this
3153 resulted in spurious "integer overflow" errors. Also, it was casting the
3154 result into an int without checking.
3156 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3157 did not work. The code looks as if it has *never* worked, though it appears
3158 to have been documented since at least releast 1.62. I have made it work.
3160 35. The "timeout_DNS" error in retry rules, also documented since at least
3161 1.62, also never worked. As it isn't clear exactly what this means, and
3162 clearly it isn't a major issue, I have abolished the feature by treating it
3163 as "timeout", and writing a warning to the main and panic logs.
3165 36. The display of retry rules for -brt wasn't always showing the error code
3168 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3169 timeout_connect_A, timeout_connect_MX.
3171 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3172 to the empty sender.
3174 39. The daemon was not analysing the content of -oX till after it had closed
3175 stderr and disconnected from the controlling terminal. This meant that any
3176 syntax errors were only noted on the panic log, and the return code from
3177 the command was 0. By re-arranging the code a little, I've made the
3178 decoding happen first, so such errors now appear on stderr, and the return
3179 code is 1. However, the actual setting up of the sockets still happens in
3180 the disconnected process, so errors there are still only recorded on the
3183 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3184 connections (as happens on some IP stacks) was logged at start up time as
3185 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3186 it from "IPv6 and IPv4", which means that two separate sockets are being
3189 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3190 says whether AF_INET or AF_INET6 was passed as an argument.
3192 42. Exiwhat output was messed up when time zones were included in log
3195 43. Exiwhat now gives more information about the daemon's listening ports,
3196 and whether -tls-on-connect was used.
3198 44. The "port" option of the smtp transport is now expanded.
3200 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3201 silently ignored. Now an error message is written to the main and panic
3204 46. There's a new ACL modifier called "logwrite" which writes to a log file
3205 as soon as it is encountered.
3207 47. Added $local_user_uid and $local_user_gid at routing time.
3209 48. Exim crashed when trying to verify a sender address that was being
3212 49. Exim was recognizing only a space character after ".include". It now also
3213 recognizes a tab character.
3215 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3216 extracting the relevant information from the specification. The man page no
3217 longer contains scrambled data for the -d option, and I've added a section
3218 at the front about calling Exim under different names.
3220 51. Added "extra_headers" argument to the "mail" command in filter files.
3222 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3225 53. Installed eximstats 1.29.
3227 54. Added transport_filter_timeout as a generic transport option.
3229 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3230 Cc: header lines. This was required by RFC 822, but it not required by RFC
3233 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3234 incoming messages that did not have them. Now it does so only if the
3235 message originates locally, that is, if there is no associated remote host
3236 address. When Resent- header lines are present, this applies to the Resent-
3237 lines rather than the non-Resent- lines.
3239 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3240 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3242 58. Messages for configuration errors now include the name of the main
3243 configuration file - useful now that there may be more than one file in a
3244 list (.included file names were always shown).
3246 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3247 for those rare installations that do not start the daemon as root or run it
3248 setuid root. I've cut out the call to initgroups() if the daemon is not
3251 60. The Exim user and group can now be bound into the binary as text strings
3252 that are looked up at the start of Exim's processing.
3254 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3256 62. Added $mailstore_basename variable.
3258 63. Installed patch to sieve.c from Michael Haardt.
3260 64. When Exim failed to open the panic log after failing to open the main log,
3261 the original message it was trying to log was written to stderr and debug
3262 output, but if they were not available (the usual case in production), it
3263 was lost. Now it is written to syslog before the two lines that record the
3264 failures to open the logs.
3266 65. Users' Exim filters run in subprocesses under the user's uid. It is
3267 possible for a "deliver" command or an alias in a "personal" command to
3268 provoke an address rewrite. If logging of address rewriting is configured,
3269 this fails because the process is not running as root or exim. There may be
3270 a better way of dealing with this, but for the moment (because 4.30 needs
3271 to be released), I have disabled address rewrite logging when running a
3272 filter in a non-root, non-exim process.
3278 1. The buildconfig auxiliary program wasn't quoting the value set for
3279 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3280 not defined. This bug was masked in 4.22 by the effect that was fixed in
3283 2. Some messages that were rejected after a message id was allocated were
3284 shown as "incomplete" by exigrep. It no longer does this for messages that
3285 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3287 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3288 have allow_domain_literals set, the ID did not get logged in the <= line.
3289 Domain literals are now always recognized in Message-ID: header lines.
3291 4. The first argument for a ${extract expansion item is the key name or field
3292 number. Leading and trailing spaces in this item were not being ignored,
3293 causing some misleading effects.
3295 5. When deliver_drop_privilege was set, single queue runner processes started
3296 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3297 same command in the process it spins off) were not dropping privilege.
3299 6. When the daemon running as "exim" started a queue runner, it always
3300 re-executed Exim in the spun-off process. This is a waste of effort when
3301 deliver_drop_privilege is set. The new process now just calls the
3302 queue-runner function directly.
3308 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3311 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3312 ignored. Though the use of -oP was forcing the writing of a pid file, it
3313 was always written to the default place.
3315 3. If the message "no IP address found for host xxxx" is generated during
3316 incoming verification, it is now followed by identification of the incoming
3317 connection (so you can more easily find what provoked it).
3319 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3321 5. Added some features to "harden" Exim a bit more against certain attacks:
3323 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3324 be put in Local/Makefile. This is like the never_users runtime option,
3325 but it cannot be overridden. The default setting is "root".
3327 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3328 prefix string with which any file named in a -C command line option
3331 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3332 is retained for -C and -D only if the caller of Exim is root. Without
3333 it, the exim user may also use -C and -D and retain privilege.
3335 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3336 command line option is disabled.
3338 6. Macro names set by the -D option must start with an upper case letter, just
3339 like macro names defined in the configuration file.
3341 7. Added "dereference=" facility to LDAP.
3343 8. Two instances of the typo "uknown" in the source files are fixed.
3345 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3346 the Configure-Makefile script screwed up while processing it.
3348 10. Incorporated PCRE 4.4.
3350 11. The SMTP synchronization check was not operating right at the start of an
3351 SMTP session. For example, it could not catch a HELO sent before the client
3352 waited for the greeting. There is now a check for outstanding input at the
3353 point when the greeting is written. Because of the duplex, asynchronous
3354 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3355 way, but not yet received, when the check is performed.
3357 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3358 on TCP/IP sockets, because this apparently causes some broken clients to
3361 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3362 unchanged) from the Cygwin maintainer.
3364 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3365 was supported instead of testing for mbx. Effectively a typo.
3367 15. The spa authenticator server code was not checking that the input it
3368 received was valid base64.
3370 16. The debug output line for the "set" modifier in ACLs was not showing the
3371 name of the variable that was being set.
3373 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3375 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3376 lookup on its own. Something else was needed to trigger the lookup. For
3377 example, a match in host_lookup or the need for a host name in a host list.
3378 Now, if $sender_host_name is referenced and the host name has not yet been
3379 looked up, a lookup is performed. If the lookup fails, the variable remains
3380 empty, and $host_lookup_failed is set to "1".
3382 19. Added "eqi" as a case-independent comparison operator.
3384 20. The saslauthd authentication condition could segfault if neither service
3385 nor realm was specified.
3387 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3388 error message that was logged was misleading, and incoming SMTP
3389 connections were dropped. The message is now more accurate, and temporary
3390 errors are given to SMTP connections.
3392 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3393 (except RSET) with a 421 error, until QUIT is received. However, it was
3394 failing to send a response to QUIT.
3396 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3397 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3398 ACL is now run after the helo_try_verify_hosts code.
3400 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3401 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3402 case-independent, so other case variants are also recognized). Apparently
3403 some systems use these upper case variants.
3405 25. If more than two messages were waiting for the same host, and a transport
3406 filter was specified for the transport, Exim sent two messages over the
3407 same TCP/IP connection, and then failed with "socket operation on non-
3408 socket" when it tried to send the third.
3410 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3412 27. The extern definition of crypt16() in expand.c was not being excluded when
3413 the OS had its own crypt16() function.
3415 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3416 as a preferred synonym for return_size_limit, both as an option and as an
3419 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3421 30. Changed the default configuration ACL to relax the local part checking rule
3422 for addresses that are not in any local domains. For these addresses,
3423 slashes and pipe symbols are allowed within local parts, but the sequence
3424 /../ is explicitly forbidden.
3426 31. SPA server authentication was not clearing the challenge buffer before
3429 32. log_message in a "warn" ACL statement was writing to the reject log as
3430 well as to the main log, which contradicts the documentation and doesn't
3431 seem right (because no rejection is happening). So I have stopped it.
3433 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3434 However, I am unable to do any testing of this.
3436 34. Fixed an infelicity in the appendfile transport. When checking directories
3437 for a mailbox, to see if any needed to be created, it was accidentally
3438 using path names with one or more superfluous leading slashes; tracing
3439 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3441 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3442 contents are added to the log line that is written for every discarded
3443 recipient. (Previously a log_message setting was ignored.)
3445 36. The ${quote: operator now quotes the string if it is empty.
3447 37. The install script runs exim in order to find its version number. If for
3448 some reason other than non-existence or emptiness, which it checks, it
3449 could not run './exim', it was installing it with an empty version number,
3450 i.e. as "exim-". This error state is now caught, and the installation is
3453 38. An argument was missing from the function that creates an error message
3454 when Exim fails to connect to the socket for saslauthd authentication.
3455 This could cause Exim to crash, or give a corrupted message.
3457 39. Added isip, isip4, and isip6 to ${if conditions.
3459 40. The ACL variables $acl_xx are now saved with the message, and can be
3460 accessed later in routers, transports, and filters.
3462 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3463 strings in the file are not string-expanded.
3465 42. If a MAIL command specified a SIZE value that was too large to fit into an
3466 int variable, the check against message_size_limit failed. Such values are
3467 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3468 day this will have to be increased, but I don't think I want to be around
3469 when emails are that large.
3476 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3477 iconv() is not standard in FreeBSD.
3479 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3480 IPv6 enabled. The observed symptom was a segmentation fault on return from
3481 the function os_common_find_running_interfaces() in src/os.c.
3483 3. In the check_special_case() function in daemon.c I had used "errno" as an
3484 argument name, which causes warnings on some systems. This was basically a
3485 typo, since it was named "eno" in the comments!
3487 4. The code that waits for the clock to tick (at a resolution of some fraction
3488 of a second) so as to ensure message-id uniqueness was always waiting for
3489 at least one whole tick, when it could have waited for less. [This is
3490 almost certainly not relevant at current processor speeds, where it is
3491 unlikely to ever wait at all. But we try to future-proof.]
3493 5. The function that sleeps for a time interval that includes fractions of a
3494 second contained a race. It did not block SIGALRM between setting the
3495 timer, and suspending (a couple of lines later). If the interval was short
3496 and the sigsuspend() was delayed until after it had expired, the suspension
3497 never ended. On busy systems this could lead to processes getting stuck for
3500 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3501 process, before it forks any delivery processes. The open lookup caching
3502 mechanism meant that the open file or database connection was passed into
3503 the delivery process. The problem was that delivery processes always tidy
3504 up cached lookup data. This could cause a problem for the next delivery
3505 process started by the queue runner, because the external queue runner
3506 process does not know about the closure. So the next delivery process
3507 still has data in the lookup cache. In the case of a file lookup, there was
3508 no problem because closing a file descriptor in a subprocess doesn't affect
3509 the parent. However, if the lookup was caching a connection to a database,
3510 the connection was closed, and the second delivery process was likely to
3511 see errors such as "PGSQL: query failed: server closed the connection
3512 unexpectedly". The problem has been fixed by closing all cached lookups
3513 in a queue runner before running a delivery process.
3515 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3516 seem to have the "const" qualifier which it has on other OS. I've
3519 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3520 *of the same type* (client or server) with the same public name that an
3521 error should be diagnosed.
3523 9. When Exim looked up a host name for an IP address, but failed to find the
3524 original IP address when looking up the host name (a safety check), it
3525 output the message "<ip address> does not match any IP for NULL", which was
3526 confusing, to say the least. The bug was that the host name should have
3527 appeared instead of "NULL".
3529 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3530 user that is built into the binary, and the -C or -D options is used, root
3531 privilege is dropped before the configuration file is read. In addition,
3532 logging is switched to stderr instead of the normal log files. If the
3533 configuration then re-defines the Exim user, the unprivileged environment
3534 is probably not what is expected, so Exim logs a panic warning message (but
3537 However, if deliver_drop_privilege is set, the unprivileged state may well
3538 be exactly what is intended, so the warning has been cut out in that case,
3539 and Exim is allowed to try to write to its normal log files.
3545 1. smtp_return_error_details was not giving details for temporary sender
3546 or receiver verification errors.
3548 2. Diagnose a configuration error if two authenticators have the same public
3551 3. Exim used not to create the message log file for a message until the first
3552 delivery attempt. This could be confusing when incoming messages were held
3553 for policy or load reasons. The message log file is now created at the time
3554 the message is received, and an initial "Received" line is written to it.
3556 4. The automatically generated man page for command line options had a minor
3557 bug that caused no ill effects; however, a more serious problem was that
3558 the procedure for building the man page automatically didn't always
3559 operate. Consequently, release 4.20 contains an out-of-date version. This
3560 shouldn't happen again.
3562 5. When building Exim with embedded Perl support, the script that builds the
3563 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3564 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3566 6. The freeze_tell option was not being used for messages that were frozen on
3567 arrival, either by an ACL or by local_scan().
3569 7. Added the smtp_incomplete_transaction log selector.
3571 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3572 was accepting AUTH without a new EHLO.
3574 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3575 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3578 10. Logging of TCP/IP connections (when configured) now happens in the main
3579 daemon process instead of the child process, so that the TCP/IP connection
3580 count is more accurate (but it can never be perfect).
3582 11. The use of "drop" in a nested ACL was not being handled correctly in the
3583 outer ACL. Now, if condition failure induced by the nested "drop" causes
3584 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3585 or "require"), the connection is dropped.
3587 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3588 that yield "discard" can now be used with an "accept" or a "discard" verb,
3589 but an error is generated for any others (because I can't see a useful way
3590 to define what should happen).
3592 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3593 are now processed in the same way as lines in the Exim configuration file.
3594 In particular, continuation lines are supported.
3596 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3598 15. Added -ti meaning -t -i.
3600 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3601 domains, and warn by logging if others are found.
3603 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3604 returned from GIFCONF when Exim is trying to find the list of interfaces on
3605 a host. The code in os.c has been modified to copy each ifreq to an aligned
3606 structure in all cases.
3608 Also, in some cases, the returned ifreq's were being copied to a 'struct
3609 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3610 means the last couple of bytes of an IPv6 address could be chopped if the
3611 ifreq contained only a normal sockaddr (14 bytes storage).
3613 18. Named domain lists were not supported in the hosts_treat_as_local option.
3614 An entry such as +xxxx was not recognized, and was treated as a literal
3617 19. Ensure that header lines added by a DATA ACL are included in the reject log
3618 if the ACL subsequently rejects the message.
3620 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3621 MD5 (which is deprecated).
3623 21. When testing a filter file using -bf, Exim was writing a message when it
3624 took the sender from a "From " line in the message, but it was not doing so
3625 when it took $return_path from a Return-Path: header line. It now does.
3627 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3628 with a valid header line field name (a series of printing characters
3629 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3631 23. Changed "disc" in the source to "disk" to conform to the documentation and
3632 the book and for uniformity.
3634 24. Ignore Sendmail's -Ooption=value command line item.
3636 25. When execve() failed while trying to run a command in a pipe transport,
3637 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3638 could be confused with a return value of 69 from the command itself. This
3639 has been changed to 127, the value the shell returns if it is asked to run
3640 a non-existent command. The wording for the related log line suggests a
3641 non-existent command as the problem.
3643 26. If received_header_text expands to an empty string, do not add a Received:
3644 header line to the message. (Well, it adds a token one on the spool, but
3645 marks it "old" so that it doesn't get used or transmitted.)
3647 27. Installed eximstats 1.28 (addition of -nt option).
3649 28. There was no check for failure on the call to getsockname() in the daemon
3650 code. This can fail if there is a shortage of resources on the system, with
3651 ENOMEM, for example. A temporary error is now given on failure.
3653 29. Contrary to the C standard, it seems that in some environments, the
3654 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3655 program. Exim now does this explicitly; it affects the formatting of
3656 timestamps using strftime().
3658 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3659 complaints. I've now initialized all the variables, to avoid this.
3661 32. Header lines added by a system filter were not being "seen" during
3662 transport-time rewrites.
3664 33. The info_callback() function passed to OpenSSL is set up with type void
3665 (*)(SSL *, int, int), as described somewhere. However, when calling the
3666 function (actually a macro) that sets it up, the type void(*)() is
3667 expected. I've put in a cast to prevent warnings from picky compilers.
3669 34. If a DNS black list lookup found a CNAME record, but there were no A
3670 records associated with the domain it pointed at, Exim crashed.
3672 35. If a DNS black list lookup returned more than one A record, Exim ignored
3673 all but the first. It now scans all returned addresses if a particular IP
3674 value is being sought. In this situation, the contents of the
3675 $dnslist_value variable are a list of all the addresses, separated by a
3678 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3679 to accept a host name and all its aliases if the forward lookup for any of
3680 them yielded the IP address of the incoming connection. Now it accepts only
3681 those names whose forward lookup yields the correct IP address. Any other
3682 names are discarded. This closes a loophole whereby a rogue DNS
3683 administrator could create reverse DNS records to break through a
3684 wildcarded host restriction in an ACL.
3686 37. If a user filter or a system filter that ran in a subprocess used any of
3687 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3688 the wrong values were passed to the pipe command ($thisaddress had the
3689 value of $0, $0 had the value of $1, etc). This bug was introduced by
3690 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3692 38. Improved the line breaking for long SMTP error messages from ACLs.
3693 Previously, if there was no break point between 40 and 75 characters, Exim
3694 left the rest of the message alone. Two changes have been made: (a) I've
3695 reduced the minimum length to 35 characters; (b) if it can't find a break
3696 point between 35 and 75 characters, it looks ahead and uses the first one
3697 that it finds. This may give the occasional overlong line, but at least the
3698 remaining text gets split now.
3700 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3701 file descriptors might be low, and that setting 1000 would always raise it.
3702 It turns out that in some environments, the limit is already over 1000 and
3703 that lowering it causes trouble. So now Exim takes care not to decrease it.
3705 40. When delivering a message, the value of $return_path is set to $sender_
3706 address at the start of routing (routers may change the value). By an
3707 oversight, this default was not being set up when an address was tested by
3708 -bt or -bv, which affected the outcome if any router or filter referred to
3711 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3712 log only when "message" or "log_message" are set. However, if one of the
3713 conditions was an address verification, or a call to a nested ACL, the
3714 messages generated by the underlying test were being passed through. This
3715 no longer happens. The underlying message is available in $acl_verify_
3716 message for both "message" and "log_message" expansions, so it can be
3717 passed through if needed.
3719 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3720 new expansion $bh_ to give the encoded byte string without charset
3721 translation. Translation happens only if iconv() is available; HAVE_ICONV
3722 indicates this at build time. HEADERS_CHARSET gives the charset to
3723 translate to; headers_charset can change it in the configuration, and
3724 "headers charset" can change it in an individual filter file.
3726 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3727 that creates RFC 2047 encoded "words" labels them as that charset instead
3728 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3729 expansion operator; (ii) when Exim creates a From: line for a local
3730 message; (iii) when a header line is rewritten to include a "phrase" part.
3732 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3733 buggy, causing it to skip the first lines of messages whose message ID
3734 ended in 'D'. This would not have bitten before Exim release 4.14, because
3735 message IDs were unlikely to end in 'D' before then. The effect was to have
3736 incorrect size information for certain domains.
3738 45. #include "config.h" was missing at the start of the crypt16.c module. This
3739 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3742 46. If there was a timeout during a "random" callout check, Exim treated it as
3743 a failure of the random address, and carried on sending RSET and the real
3744 address. If the delay was just some slowness somewhere, the response to the
3745 original RCPT would be taken as a response to RSET and so on, causing
3746 mayhem of various kinds.
3748 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3749 when I implemented it. It didn't allow for the fact that some option values
3750 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3751 the right test for positive values.
3753 48. Domain names in DNS records are case-independent. Exim always looks them up
3754 in lower case. Some resolvers return domain names in exactly the case they
3755 appear in the zone file, that is, they may contain uppercase letters. Not
3756 all resolvers do this - some return always lower case. Exim was treating a
3757 change of case by a resolver as a change of domain, similar to a widening
3758 of a domain abbreviation. This triggered its re-routing code and so it was
3759 trying to route what was effectively the same domain again. This normally
3760 caused routing to fail (because the router wouldn't handle the domain
3761 twice). Now Exim checks for this case specially, and just changes the
3762 casing of the domain that it ultimately uses when it transmits the message
3765 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3768 50. If a filter generated a file delivery with a non-absolute name (possible if
3769 no home directory exists for the router), the forbid_file option was not
3772 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3773 the existing equality matching.
3775 52. Exim was using ints instead of ino_t variables in some places where it was
3776 dealing with inode numbers.
3778 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3779 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3780 called TMPDIR, and if it finds it is different, it changes its value.
3782 54. The smtp_printf() function is now made available to local_scan() so
3783 additional output lines can be written before returning. There is also an
3784 smtp_fflush() function to enable the detection of a dropped connection.
3785 The variables smtp_input and smtp_batched_input are exported to
3788 55. Changed the default runtime configuration: the message "Unknown user"
3789 has been removed from the ACL, and instead placed on the localuser router,
3790 using the cannot_route_message feature. This means that any verification
3791 failures that generate their own messages won't get overridden. Similarly,
3792 the "Unrouteable address" message that was in the ACL for unverifiable
3793 relay addresses has also been removed.
3795 56. Added hosts_avoid_esmtp to the smtp transport.
3797 57. The exicyclog script was not checking for the esoteric option
3798 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3799 will work only if exicyclog is run under the appropriate euid.
3801 58. Following a discussion on the list, the rules by which Exim recognises line
3802 endings on incoming messages have been changed. The -dropcr and drop_cr
3803 options are now no-ops, retained only for backwards compatibility. The
3804 following line terminators are recognized: LF CRLF CR. However, special
3805 processing applies to CR:
3807 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3808 nor a local message in the state where . is a terminator.
3810 (ii) If a bare CR is encountered in a header line, an extra space is added
3811 after the line terminator so as not to end the header. The reasoning
3812 behind this is that bare CRs in header lines are most likely either
3813 to be mistakes, or people trying to play silly games.
3815 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3816 was being incorrectly given as 18 bytes larger than it should have been.
3817 This is a VOB (very old bug).
3819 60. This may never have affected anything current, but just in case it has:
3820 When the local host is found other than at the start of a list of hosts,
3821 the local host, those with the same MX, and any that follow, are discarded.
3822 When the list in question was part of a longer list of hosts, the following
3823 hosts (not currently being processed) were also being discarded. This no
3824 longer happens. I'm not sure if this situation could ever has previously
3827 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3830 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3831 header. This is something that is recommended in a new Internet Draft, and
3832 is something that is documented as being done by Sendmail. There are two
3833 possible values. For messages generated by the autoreply transport, Exim
3836 Auto-Submitted: auto-replied
3838 whereas for all other generated messages (e.g. bounces) it adds
3840 Auto-Submitted: auto-generated
3842 63. The "personal" condition in filters now includes a test for the
3843 Auto-Submitted: header. If it contains the string "auto-" the message it
3844 not considered personal.
3846 64. Added rcpt_include_affixes as a generic transport option.
3848 65. Added queue_only_override (default true).
3850 66. Added the syslog_duplication option.
3852 67. If what should have been the first header line of a message consisted of
3853 a space followed by a colon, Exim was mis-interpreting it as a header line.
3854 It isn't of course - it is syntactically invalid and should therefore be
3855 treated as the start of the message body. The misbehaviour could have
3856 caused a number of strange effects, including loss of data in subsequent
3857 header lines, and spool format errors.
3859 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3860 client host had authenticated. This control can now be exercised by an ACL
3861 for more flexibility.
3863 69. By default, callouts do not happen when testing with -bh. There is now a
3864 variant, -bhc, which does actually run the callout code, including
3865 consulting and updating the callout cache.
3867 70. Added support for saslauthd authentication, courtesy of Alexander
3870 71. If statvfs() failed on the spool or log directories while checking their
3871 size for availability, Exim confusingly gave the error "space shortage".
3872 Furthermore, in debugging mode it crashed with a floating point exception.
3873 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3874 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3875 serious problem, Exim now writes to the main and panic logs when this
3876 happens, with details of the failure. It then refuses to accept the
3877 incoming message, giving the message "spool directory problem" or "log
3878 directory problem" with a 421 code for SMTP messages.
3880 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3881 0, 1, and 2 exist, because some OS complain for execs without them (see
3882 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3883 descriptors. However, the code omitted to check that the open succeeded,
3884 causing mysterious errors if for some reason the permissions on /dev/null
3885 got screwed. Now Exim writes a message to the main and panic logs, and
3886 bombs out if it can't open /dev/null.
3888 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3889 interact so that it is all more flexible. It is supposed to remain
3890 backwards compatible. Also added extra_local_interfaces.
3892 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3893 to bomb out with an assertion failure - to the client this appears as a
3894 connection drop. This problem occurs in the part of the code that was taken
3895 from the Samba project. Fortunately, the assertion is in a very simple
3896 function, so I have fixed this by reproducing the function inline in the
3897 one place where it is called, and arranging for authentication to fail
3898 instead of killing the process with assert().
3900 75. The SPA client code was not working when the server requested OEM rather
3901 than Unicode encoding.
3903 76. Added code to make require_files with a specific uid setting more usable in
3904 the case where statting the file as root fails - usually a non-root-mounted
3905 NFS file system. When this happens and the failure is EACCES, Exim now
3906 forks a subprocess and does the per-uid checking as the relevant uid.
3908 77. Added process_log_path.
3910 78. If log_file_path was not explicitly set, a setting of check_log_space or
3911 check_log_inodes was ignored.
3913 79. If a space check for the spool or log partitions fails, the incident is now
3914 logged. Of course, in the latter case the data may get lost...
3916 80. Added the %p formatting code to string_format() so that it can be used to
3917 print addresses in debug_print(). Adjusted all the address printing in the
3918 debugging in store.c to use %p rather than %d.
3920 81. There was a concern that a line of code in smtp_in.c could overflow a
3921 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3922 initially expressed, the concern was not well-founded, because trailing
3923 spaces are removed early. However, if the trailing spaces were followed by
3924 a NULL, they did not get removed, so the overflow was possible. Two fixes
3927 (a) I re-wrote the offending code in a cleaner fashion.
3928 (b) If an incoming SMTP command contains a NULL character, it is rejected
3931 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3932 runs initgroups(), so that if the Exim user is in any additional groups,
3933 they will be used during message reception.
3939 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob