1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.278 2005/12/15 17:58:23 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
20 PH/03 Added disable_ipv6.
22 PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
29 PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
32 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
38 PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
43 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
46 PH/09 Moved a debug statement in filter processing to avoid a race problem when
49 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
52 JJ/02 exipick: Added support for new ACL variable spool format introduced
59 PH/01 Two changes to the default runtime configuration:
61 (1) Move the checks for relay_from_hosts and authenticated clients from
62 after to before the (commented out) DNS black list checks.
64 (2) Add control=submission to the relay_from_hosts and authenticated
65 clients checks, on the grounds that messages accepted by these
66 statements are most likely to be submissions.
68 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
70 (1) Generate an error if the third argument for the ${prvs expansion is
73 (2) Treat a missing third argument of ${prvscheck as if it were an empty
76 (3) Reset the variables that are obtained from the first argument of
77 ${prvscheck and used in the second argument before leaving the code,
78 because their memory is reclaimed, so using them afterwards may do
81 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
82 one (it's much easier than Tom thought :-).
84 (5) Because of (4), we can now allow for the use of $prvscheck_result
85 inside the third argument.
87 PH/03 For some reason, the default setting of PATH when running a command from
88 a pipe transport was just "/usr/bin". I have changed it to
91 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
92 anything to be listed in the output from -bV.
94 PH/05 When a filter generated an autoreply, the entire To: header line was
95 quoted in the delivery log line, like this:
97 => >A.N.Other <ano@some.domain> <original@ddress> ...
99 This has been changed so that it extracts the operative address. There
100 may be more than one such address. If so, they are comma-separated, like
103 => >ano@some.domain,ona@other.domain <original@ddress> ...
105 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
106 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
107 not being looked up in the rDNS to get a host name, Exim was showing the
108 IP address twice in Received: lines, even though the IP addresses were
109 identical. For example:
111 Received: from [1.2.3.4] (helo=[1.2.3.4])
113 However, if the real host name was known, it was omitting the HELO data
114 if it matched the actual IP address. This has been tidied up so that it
115 doesn't show the same IP address twice.
117 PH/07 When both +timestamp and +memory debugging was on, the value given by
118 $tod_xxx expansions could be wrong, because the tod_stamp() function was
119 called by the debug printing, thereby overwriting the timestamp buffer.
120 Debugging no longer uses the tod_stamp() function when +timestamp is set.
122 PH/08 When the original message was included in an autoreply transport, it
123 always said "this is a copy of the message, including all the headers",
124 even if body_only or headers_only was set. It now gives an appropriate
127 PH/09 Applied a patch from the Sieve maintainer which:
129 o fixes some comments
130 o adds the (disabled) notify extension core
131 o adds some debug output for the result of if/elsif tests
132 o points to the current vacation draft in the documentation
133 and documents the missing references header update
137 o fixes a bug in processing the envelope test (when testing
138 multiple envelope elements, the last element determinted the
141 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
142 Electronic Mail") by including:
144 Auto-submitted: auto-generated
146 in the messages that it generates (bounce messages and others, such as
147 warnings). In the case of bounce messages for non-SMTP mesages, there was
148 also a typo: it was using "Auto_submitted" (underscore instead of
149 hyphen). Since every message generated by Exim is necessarily in response
150 to another message, thes have all been changed to:
152 Auto-Submitted: auto-replied
154 in accordance with these statements in the RFC:
156 The auto-replied keyword:
158 - SHOULD be used on messages sent in direct response to another
159 message by an automatic process,
161 - MUST NOT be used on manually-generated messages,
163 - MAY be used on Delivery Status Notifications (DSNs) and Message
164 Disposition Notifications (MDNs),
166 - MUST NOT be used on messages generated by automatic or periodic
167 processes, except for messages which are automatic responses to
170 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
171 to the default Received: header definition.
173 PH/12 Added log selector acl_warn_skipped (default on).
175 PH/13 After a successful wildlsearch lookup, discard the values of numeric
176 variables because (a) they are in the wrong storage pool and (b) even if
177 they were copied, it wouldn't work properly because of the caching.
179 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
180 checking when decoding. Apparently there are clients that generate
181 overlong encoded strings. Why am I not surprised?
183 PH/15 If the first argument of "${if match_address" was not empty, but did not
184 contain an "@" character, Exim crashed. Now it writes a panic log message
185 and treats the condition as false.
187 PH/16 In autoreply, treat an empty string for "once" the same as unset.
189 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
190 extension "envelope-auth". The code is finished and in agreement with
191 other implementations, but there is no documentation so far and in fact,
192 nobody wrote the draft yet. This extension is currently #undef'ed, thus
193 not changing the active code.
195 Print executed "if" and "elsif" statements when debugging is used. This
196 helps a great deal to understand what a filter does.
198 Document more things not specified clearly in RFC3028. I had all this
199 sorted out, when out of a sudden new issues came to my mind. Oops."
201 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
204 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
205 canonical form (as documented). However, after a host name lookup from
206 the IP address, check_host() was doing a simple string comparison with
207 addresses acquired from the DNS when checking that the found name did
208 have the original IP as one of its addresses. Since any found IPv6
209 addresses are likely to be in abbreviated form, the comparison could
210 fail. Luckily, there already exists a function for doing the comparison
211 by converting both addresses to binary, so now that is used instead of
214 PH/20 There was another similar case to PH/19, when a complete host name was
215 given in a host list; looking up its IP address could give an abbreviated
216 form, whereas the current host's name might or might not be abbreviated.
217 The same fix has been applied.
223 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
224 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
227 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
228 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
230 PH/03 Typo: missing ".o" in src/pcre/Makefile.
232 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
233 header line, restrict the check to what is listed in RFCs 2369 and 2929.
234 Also, for "Auto-Submitted", treat anything other than "no" as
235 non-personal, in accordance with RFC 3834. (Previously it treated
236 anything starting "auto-" as non-personal.)
238 TF/01 The control=submission/name=... option had a problem with syntax
239 errors if the name included a slash character. The /name= option
240 now slurps the rest of the string, so it can include any characters
241 but it must come last in the list of options (after /sender_retain
244 PH/05 Some modifications to the interface to the fake nameserver for the new
247 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
248 into the default Received: header string.
255 TK/01 Added the "success_on_redirect" address verification option. See
256 NewStuff for rationale and an example.
258 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
260 PH/02 Patch to exigrep to allow it to work on syslog lines.
262 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
263 fread() to skip over the body file's header line, because in Cygwin the
264 header line is locked and is inaccessible.
266 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
267 co-exist for some time) to make it clear that it is the Exim ID that is
268 referenced, not the Message-ID: header line.
270 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
271 string_format() function, because snprintf() does not exist on all
274 PH/06 The use of forbid_filter_existstest now also locks out the use of the
275 ${stat: expansion item.
277 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
278 protocol synchronization error", to keep the pedants happy.
280 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
281 well as for IRIX systems, when gcc is being used. See the host.c source
284 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
286 PH/10 Named domain lists were not working if used in a queue_smtp_domains
289 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
290 transport and to the smtp transport in LMTP mode.
292 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
294 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
295 run a filter in a subprocess. This could lead to confusion in subsequent
296 lookups in the parent process. There should also be a search_tidyup() at
297 the end of the subprocess.
299 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
300 only if the host matched helo_try_verify_hosts, which caused the
301 verification to occur when the EHLO/HELO command was issued. The ACL just
302 tested the remembered result. Now, if a previous verification attempt has
303 not happened, "verify = helo" does it there and then.
305 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
307 TK/03 Fix log output including CR from clamd.
309 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
310 exist provoked a memory error which could cause a segfault.
312 PH/15 Installed PCRE 6.2
314 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
316 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
317 of the problem. Specifically, suggested +O2 rather than +O1 for the
320 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
322 PH/20 If a delivery was routed to a non-standard port by means of an SRV
323 record, the port was not correctly logged when the outgoing_port log
324 selector was set (it logged the transort's default port).
326 PH/21 Added support for host-specific ports to manualroute, queryprogram,
327 fallback_hosts, and "hosts" in the smtp transport.
329 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
330 host errors such as "Connection refused".
332 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
333 authentication with radiusclient 0.4.9:
335 - Error returned from rc_read_config was caught wrongly
336 - Username/password not passed on to radius server due to wrong length.
338 The presumption is that some radiusclient API changes for 4.51/PH/17
339 were not taken care of correctly. The code is still untested by me (my
340 Linux distribution still has 0.3.2 of radiusclient), but it was
341 contributed by a Radius user.
343 PH/24 When doing a callout, the value of $domain wasn't set correctly when
344 expanding the "port" option of the smtp transport.
346 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
347 while reading a MIME header. Thanks to Tom Hughes for a patch.
349 PH/24 Include config.h inside local_scan.h so that configuration settings are
352 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
353 that in an ACL for RCPT (for example), you can examine exactly what was
356 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
357 commands, but it was not correctly comparing the address with the actual
358 client host address. Thus, it would show the EHLO address in Received:
359 header lines when this was not necessary.
361 PH/27 Added the % operator to ${eval:}.
363 PH/28 Exim tries to create and chdir to its spool directory when it starts;
364 it should be ignoring failures (because with -C, for example, it has lost
365 privilege). It wasn't ignoring creation failures other than "already
368 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
371 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
372 because an input line was too long, either on its own, or by virtue of
373 too many continuations, the temporary file was not being removed, and the
374 return code was incorrect.
376 PH/31 Missing "BOOL" in function definition in filtertest.c.
378 PH/32 Applied Sieve patches from the maintainer.
380 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
382 PH/33 Added "verify = not_blind".
384 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
385 Local/Makefile (with some defaults set). These are used in built scripts
386 such as exicyclog, but they have never been used in the exim_install
387 script (though there are many overriding facilities there). I have
388 arranged that the exim_install script now takes note of these two
391 PH/35 Installed configuration files for Dragonfly.
393 PH/36 When a locally submitted message by a trusted user did not contain a
394 From: header, and the sender address was obtained from -f or from an SMTP
395 MAIL command, and the trusted user did not use -F to supply a sender
396 name, $originator_name was incorrectly used when constructing a From:
397 header. Furthermore, $originator_name was used for submission mode
398 messages from external hosts without From: headers in a similar way,
399 which is clearly wrong.
401 PH/37 Added control=suppress_local_fixups.
403 PH/38 When log_selector = +received_sender was set, and the addition of the
404 sender made the log line's construction buffer exactly full, or one byte
405 less than full, an overflow happened when the terminating "\n" was
408 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
409 when the result of a list match is failure because a DNS lookup failed.
411 PH/40 RM_COMMAND is now used in the building process.
413 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
414 the "build-* directories that it finds.
416 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
417 address in a domain literal was a prefix of an interface address.
419 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
420 when verifying a sender address, unless rewrite_headers is false.
422 PH/44 Wrote a long comment about why errors_to addresses are verified as
423 recipients, not senders.
425 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
426 the ratelimit ACL was added.
428 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
430 PH/46 Added extra information about PostgreSQL errors to the error string.
432 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
433 avoiding the need to install special zones in a real server. This is
434 backwards compatible; if it can't find the fake resolver, it drops back.
435 Thus, both old and new test suites can be run.
437 TF/02 Added util/ratelimit.pl
439 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
440 clock is set back in time.
442 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
443 Candler <B.Candler@pobox.com>.
445 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
446 OK for addresses that are the result of redirections.
448 PH/48 A number of further additions for the benefit of the new test suite,
449 including a fake gethostbyname() that interfaces to the fake DNS resolver
452 TF/06 The fix for widen_domains has also been applied to qualify_single and
453 search_parents which are the other dnslookup options that can cause
456 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
459 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
461 TF/07 Exim produced the error message "an SRV record indicated no SMTP
462 service" if it encountered an MX record with an empty target hostname.
463 The message is now "an MX or SRV record indicated no SMTP service".
465 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
466 if the DNS of the sending site is misconfigured. This is quite a
467 common situation. This change restores the behaviour of treating a
468 helo verification defer as a failure.
470 PH/51 If self=fail was set on a router, the bounce message did not include the
471 actual error message.
477 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
479 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
480 command itself ended in error, the underlying message about the transport
481 filter timeout was being overwritten with the pipe command error. Now the
482 underlying error message should be appended to the second error message.
484 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
486 PH/02 Reduce the amount of output that "make" produces by default. Full output
487 can still be requested.
489 PH/03 The warning log line about a condition test deferring for a "warn" verb
490 was being output only once per connection, rather than after each
491 occurrence (because it was using the same function as for successful
492 "warn" verbs). This seems wrong, so I have changed it.
494 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
495 it should not have, which might have caused a crash in the right
496 circumstances, but probably never did.
498 PH/04 Installed a modified version of Tony Finch's patch to make submission
499 mode fix the return path as well as the Sender: header line, and to
500 add a /name= option so that you can make the user's friendly name appear
503 TF/03 Added the control = fakedefer ACL modifier.
505 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
506 Mark Lowes for thorough testing.
508 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
510 TK/03 Merged latest SRS patch from Miles Wilton.
512 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
513 with the definition in sysexits.h (which is #included earlier).
514 Fortunately, Exim does not actually use EX_OK. The code used to try to
515 preserve the sysexits.h value, by assumimg that macro definitions were
516 scanned for macro replacements. I have been disabused of this notion,
517 so now the code just undefines EX_OK before #including unistd.h.
519 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
520 in the smtp transport. When a block could not be written in a single
521 write() function, the timeout was being re-applied to each part-write.
522 This seems wrong - if the receiver was accepting one byte at a time it
523 would take for ever. The timeout is now adjusted when this happens. It
524 doesn't have to be particularly precise.
526 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
527 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
529 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
530 without a domain if the check to <postmaster@domain> fails.
532 SC/01 Eximstats: added -xls and the ability to specify output files
533 (patch written by Frank Heydlauf).
535 SC/02 Eximstats: use FileHandles for outputing results.
537 SC/03 Eximstats: allow any combination of xls, txt, and html output.
539 SC/04 Eximstats: fixed display of large numbers with -nvr option
541 SC/05 Eximstats: fixed merging of reports with empty tables.
543 SC/06 Eximstats: added the -include_original_destination flag
545 SC/07 Eximstats: removed tabs and trailing whitespace.
547 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
549 TK/06 MBOX spool code: Add real "From " MBOX separator line
550 so the .eml file is really in mbox format (even though
551 most programs do not really care). Patch from Alex Miller.
553 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
554 The latter is generated from $received_to and is only set if the
555 message has one envelope recipient. SA can use these headers,
556 obviously out-of-the-box. Patch from Alex Miller.
558 PH/08 The ${def test on a variable was returning false if the variable's
559 value was "0", contrary to what the specification has always said!
560 The result should be true unless the variable is empty.
562 PH/09 The syntax error of a character other than { following "${if
563 def:variable_name" (after optional whitespace) was not being diagnosed.
564 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
565 accidental colon was present, for example, could give incorrect results.
567 PH/10 Tidied the code in a number of places where the st_size field of a stat()
568 result is used (not including appendfile, where other changes are about
571 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
572 This involved changing a lot of size variables from int to off_t. It
573 should work with maildirs and everything.
575 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
576 spamd dying while we are connected to it.
578 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
581 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
582 function for BDB 4.3.
584 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
586 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
587 the sources for PCRE have been re-arranged and more files are now
590 PH/15 The code I had for printing potentially long long variables in PH/11
591 above was not the best (it lost precision). The length of off_t variables
592 is now inspected at build time, and an appropriate printing format (%ld
593 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
594 to be "long long int" or "long int". This is needed for the internal
595 formatting function string_vformat().
597 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
598 the configuration file to be ":syslog", then the script "guesses" where
599 the logs files are, rather than using the compiled in default. In our
600 case the guess is not the same as the compiled default, so the script
601 suddenly stopped working when I started to use syslog. The patch checks
602 to see if log_file_path is "". If so, it attempts to read it from exim
603 with no configuration file to get the compiled in version, before it
604 falls back to the previous guessing code."
606 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
607 implementing BATV in an Exim configuration. See NewStuff for the gory
610 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
611 Makefile that are specific to HP-UX.
613 PH/18 If the "use_postmaster" option was set for a recipient callout together
614 with the "random" option, the postmaster address was used as the MAIL
615 FROM address for the random test, but not for the subsequent recipient
616 test. It is now used for both.
618 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
619 patch removes a few documentation additions to RFC 3028, because the
620 latest draft now contains them. It adds the new en;ascii-case comparator
621 and a new error check for 8bit text in MIME parts. Comparator and
622 require names are now matched exactly. I enabled the subaddress
623 extension, but it is not well tested yet (read: it works for me)."
625 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
626 rework some of the code of TK/09 above to avoid the hardwired use of
627 "%lld" and "long long". Replaced the call to snprintf() with a call to
630 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
631 records point to non-existent hosts", "retry timeout exceeded", and
632 "retry time not reached for any host after a long failure period".
634 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
635 experimental DomainKeys support:
637 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
638 (2) On an error such as an illegally used "control", the wrong name for
639 the control was given.
641 These problems did NOT occur unless DomainKeys support was compiled.
643 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
645 PH/24 Added ${if match_ip condition.
647 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
648 databases so that it will be absolutely obvious if a crash occurs in the
649 DB library. This is a regular occurrence (often caused by mis-matched
652 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
653 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
654 picked up on a user's system that detects such things. There doesn't seem
655 to be a gcc warning option for this - only an attribute that has to be
656 put on the function's prototype. It seems that in Fedora Core 4 they have
657 set this on a number of new functions. No doubt there will be more in due
660 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
661 specify a transport. However, if an address that was verified by such a
662 router was the subject of a callout, Exim crashed because it tried to
663 read the rcpt_include_affixes from the non-existent transport. Now it
664 just assumes that the setting of that option is false. This bug was
665 introduced by 4.51/PH/31.
667 PH/28 Changed -d+all to exclude +memory, because that information is very
668 rarely of interest, but it makes the output a lot bigger. People tend to
669 do -d+all out of habit.
671 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
672 code in os-type was giving problems when libc.so lives in lib64, like on
675 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
676 aren't the modern standard, and it seems that some systems' include files
677 don't always have them. Exim was already checking for some of the newer
678 ones like T_AAAA, and defining it itself. I've added checks for all the
679 record types that Exim uses.
681 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
682 not automatically generating a new one, as it is supposed to. This
683 prevented TLS from working. If the file did exist, but contained invalid
684 data, a new version was generated, as expected. It was only the case of a
685 non-existent file that was broken.
687 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
688 with a change in libdomainkeys > 0.64.
690 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
691 from DNS. If the selector record carries the flag, it now has
692 precedence over the domain-wide flag.
694 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
696 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
697 the use of a port name that isn't defined in /etc/services) occurred, the
698 message was deferred as in a normal delivery, and thus remained on the
699 spool, instead of being failed because of the mua_wrapper setting. This
700 is now fixed, and I tidied up some of the mua_wrapper messages at the
703 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
704 the messages in a hash of arrays rather than using individual hashes.
705 This is a bit cleaner and results in dramatic memory savings, albeit
706 at a slight CPU cost.
708 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
709 as requested by Marc Sherman.
711 SC/10 Eximstats: added histograms for user specified patterns as requested
714 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
716 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
717 fopen() in the content-scanning modules that did not already have it.
723 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
724 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
726 TK/02 Fix ACL "control" statement not being available in MIME ACL.
728 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
730 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
731 to test Sieve filters that use "vacation".
733 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
734 that changes the way the GnuTLS parameters are stored in the cache file.
735 The new format can be generated externally. For backward compatibility,
736 if the data in the cache doesn't make sense, Exim assumes it has read an
737 old-format file, and it generates new data and writes a new file. This
738 means that you can't go back to an older release without removing the
741 PH/03 A redirect router that has both "unseen" and "one_time" set does not
742 work if there are any delivery delays because "one_time" forces the
743 parent to be marked "delivered", so its unseen clone is never tried
744 again. For this reason, Exim now forbids the simultaneous setting of
747 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
748 redirected to themselves ("homonym" addresses). Read the long ChangeLog
749 entry if you want to know the details. The fix, however, neglected to
750 consider the case when local delivery batching is involved. The test for
751 "previously delivered" was not happening when checking to see if an
752 address could be batched with a previous (undelivered) one; under
753 certain circumstances this could lead to multiple deliveries to the same
756 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
757 in its include files, and this causes problems building Exim.
759 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
760 header_syntax) but Exim was just ignoring anything given after a slash.
761 In particular, this caused confusion with an attempt to use "verify =
762 reverse_host_lookup/defer_ok". An error is now given when options are
763 supplied for verify items that do not have them. (Maybe reverse_host_
764 lookup should have a defer_ok option, but that's a different point.)
766 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
767 defined by RFC 821) to 2048, because there were problems with some AUTH
768 commands, and RFC 1869 says the size should be increased for extended
769 SMTP commands that take arguments.
771 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
774 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
775 "unknown" error; now it says that the functionality isn't in the binary.
777 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
778 an address' error message when a string expansion fails (syntax or
779 whatever). Otherwise the password may appear in the log. Following change
780 PH/42 below, there is no longer a chance of it appearing in a bounce
783 PH/11 Installed exipick version 20050225.0 from John Jetmore.
785 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
786 of its addresses was ever tried. (Bugzilla bug #2.)
788 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
789 the result incorrectly in the debug output. (It correctly added a newline
790 to what was transported.)
792 TF/01 Added $received_time.
794 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
795 commented out examples of how to interface to a virus scanner and to
796 SpamAssassin. Also added commented examples of av_scanner and
797 spamd_address settings.
799 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
800 and controls are allowed in which ACLs. There were a couple of minor
801 errors. Some of the entries in the conditions table (which is a table of
802 where they are NOT allowed) were getting very unwieldy; rewrote them as a
803 negation of where the condition IS allowed.
805 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
807 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
808 header file does not have a version number, so I've had to invent a new
809 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
810 API. The code is untested by me (my Linux distribution still has 0.3.2 of
811 radiusclient), but it was contributed by a Radius user.
813 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
814 files or directories, for OpenSSL.
816 PH/19 When an Exim process that is running as root has to create an Exim log
817 file, it does so in a subprocess that runs as exim:exim so as to get the
818 ownership right at creation (otherwise, other Exim processes might see
819 the file with the wrong ownership). There was no test for failure of this
820 fork() call, which would lead to the process getting stuck as it waited
821 for a non-existent subprocess. Forks do occasionally fail when resources
822 run out. I reviewed all the other calls to fork(); they all seem to check
825 PH/20 When checking for unexpected SMTP input at connect time (before writing
826 the banner), Exim was not dealing correctly with a non-positive return
827 from the read() function. If the client had disconnected by this time,
828 the result was a log entry for a synchronization error with an empty
829 string after "input=" when read() returned zero. If read() returned -1
830 (an event I could not check), uninitialized data bytes were printed.
831 There were reports of junk text (parts of files, etc) appearing after
834 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
836 PH/22 Added support for macro redefinition, and (re)definition in between
837 driver and ACL definitions.
839 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
840 forgetting to use the resulting value; it was using the unexpanded value.
842 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
843 hadn't been configured. The fix is from Juergen Kreileder, who
844 understands it better than I do:
846 "Here's what I see happening with three configured cyrus_sasl
847 authenticators configured (plain, login, cram-md5):
849 On startup auth_cyrus_sasl_init() gets called for each of these.
850 This means three calls to sasl_listmech() without a specified mech_list.
851 => SASL tests which mechs of all available mechs actually work
852 => three warnings about OTP not working
853 => the returned list contains: plain, login, cram-md5, digest-md5, ...
855 With the patch, sasl_listmech() also gets called three times. But now
856 SASL's mech_list option is set to the server_mech specified in the the
857 authenticator. Or in other words, the answer from sasl_listmech()
858 gets limited to just the mech you're testing for (which is different
860 => the return list contains just 'plain' or 'login', 'cram-md5' or
861 nothing depending on the value of ob->server_mech.
863 I've just tested the patch: Authentication still works fine,
864 unavailable mechs specified in the exim configuration are still
865 caught, and the auth.log warnings about OTP are gone."
867 PH/25 When debugging is enabled, the contents of the command line are added
868 to the debugging output, even when log_selector=+arguments is not
871 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
872 answer is "GNU", and only if the return is "GNU/something" is the answer
875 PH/27 $acl_verify_message is now set immediately after the failure of a
876 verification in an ACL, and so is available in subsequent modifiers. In
877 particular, the message can be preserved by coding like this:
879 warn !verify = sender
880 set acl_m0 = $acl_verify_message
882 Previously, $acl_verify_message was set only while expanding "message"
883 and "log_message" when a very denied access.
885 PH/28 Modified OS/os.c-Linux with
887 -#ifndef OS_LOAD_AVERAGE
888 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
890 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
891 nomenclature these days.)
893 PH/29 Installed patch from the Sieve maintainer that adds the options
894 sieve_useraddress and sieve_subaddress to the redirect router.
896 PH/30 In these circumstances:
897 . Two addresses routed to the same list of hosts;
898 . First host does not offer TLS;
899 . First host accepts first address;
900 . First host gives temporary error to second address;
901 . Second host offers TLS and a TLS session is established;
902 . Second host accepts second address.
903 Exim incorrectly logged both deliveries with the TLS parameters (cipher
904 and peerdn, if requested) that were in fact used only for the second
907 PH/31 When doing a callout as part of verifying an address, Exim was not paying
908 attention to any local part prefix or suffix that was matched by the
909 router that accepted the address. It now behaves in the same way as it
910 does for delivery: the affixes are removed from the local part unless
911 rcpt_include_affixes is set on the transport.
913 PH/32 Add the sender address, as F=<...>, to the log line when logging a
914 timeout during the DATA phase of an incoming message.
916 PH/33 Sieve envelope tests were broken for match types other than :is. I have
917 applied a patch sanctioned by the Sieve maintainer.
919 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
920 the uid or gid is negative. A case of a negative gid caused this to be
921 noticed. The fix allows for either to be negative.
923 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
924 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
927 PH/36 The change PH/12 above was broken. Fixed it.
929 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
930 the grounds that routing the same address twice would always produce the
931 same answer. This might have been true once, but it is certainly no
932 longer true now. Routing a child address may depend on the previous
933 routing that produced that child. Some complicated redirection strategies
934 went wrong when messages had multiple recipients, and made Exim's
935 behaviour dependent on the order in which the addresses were given.
937 I have moved the duplicate checking until after the routing is complete.
938 Exim scans the addresses that are assigned to local and remote
939 transports, and removes any duplicates. This means that more work will be
940 done, as duplicates will always all be routed, but duplicates are
941 presumably rare, so I don't expect this is of any significance.
943 For deliveries to pipes, files, and autoreplies, the duplicate checking
944 still happens during the routing process, since they are not going to be
947 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
948 It corrects a timeout issue with spamd. This is Ian's comment: "The
949 background is that sometimes spamd either never reads data from a
950 connection it has accepted, or it never writes response data. The exiscan
951 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
952 blindly assumes that writes won't block so it may never time out."
954 PH/39 Allow G after quota size as well as K and M.
956 PH/40 The value set for $authenticated_id in an authenticator may not contain
957 binary zeroes or newlines because the value is written to log lines and
958 to spool files. There was no check on this. Now the value is run through
959 the string_printing() function so that such characters are converted to
960 printable escape sequences.
962 PH/41 $message_linecount is a new variable that contains the total number of
963 lines in the message. Compare $body_linecount, which is the count for the
966 PH/42 Exim no longer gives details of delivery errors for specific addresses in
967 bounce and delay warning messages, except in certain special cases, which
970 (a) An SMTP error message from a remote host;
971 (b) A message specified in a :fail: redirection;
972 (c) A message specified in a "fail" command in a system filter;
973 (d) A message specified in a FAIL return from the queryprogram router;
974 (e) A message specified by the cannot_route_message router option.
976 In these cases only, Exim does include the error details in bounce and
977 warning messages. There are also a few cases where bland messages such
978 as "unrouteable address" or "local delivery error" are given.
980 PH/43 $value is now also set for the "else" part of a ${run expansion.
982 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
983 being worked on, but at least Exim now implements the latest version to
986 PH/45 In a pipe transport, although a timeout while waiting for the pipe
987 process to complete was treated as a delivery failure, a timeout while
988 writing the message to the pipe was logged, but erroneously treated as a
989 successful delivery. Such timeouts include transport filter timeouts. For
990 consistency with the overall process timeout, these timeouts are now
991 treated as errors, giving rise to delivery failures by default. However,
992 there is now a new Boolean option for the pipe transport called
993 timeout_defer, which, if set TRUE, converts the failures into defers for
994 both kinds of timeout. A transport filter timeout is now identified in
997 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
998 systems where "make" and "gmake" are different, calling "gmake" at top
999 level broke things. I've arranged for the value of $(MAKE) to be passed
1000 from the Makefile to this script so that it can call the same version of
1004 A note about Exim versions 4.44 and 4.50
1005 ----------------------------------------
1007 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1008 changes of various kinds. As a consequence, a big documentation update was
1009 needed. This delayed the release for rather longer than seemed good, especially
1010 in the light of a couple of (minor) security issues. Therefore, the changes
1011 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1012 release. So 4.44 and 4.50 are in effect two different branches that both start
1015 I have left the 4.50 change log unchanged; it contains all the changes since
1016 4.43. The change log for 4.44 is below; many of its items are identical to
1017 those for 4.50. This seems to be the most sensible way to preserve the
1018 historical information.
1024 1. Minor wording change to the doc/README.SIEVE file.
1026 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1027 computation of the current number of files was incorrect.
1029 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1030 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1033 4. Give more explanation in the error message when the command for a transport
1034 filter fails to execute.
1036 5. There are several places where Exim runs a non-Exim command in a
1037 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1038 was being done only for the command run by the queryprogram router. It is
1039 now done for all such subprocesses. The other cases are: ${run, transport
1040 filters, and the commands run by the lmtp and pipe transports.
1042 6. Added CONFIGURE_GROUP build-time option.
1044 7. Some older OS have a limit of 256 on the maximum number of file
1045 descriptors. Exim was using setrlimit() to set 1000 as a large value
1046 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1047 systems. I've change it so that if it can't get 1000, it tries for 256.
1049 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1050 was an oversight, and furthermore, ever since the addition of extra
1051 controls (e.g. 4.43/32), the checks on when to allow different forms of
1052 "control" were broken. There should now be diagnostics for all cases when a
1053 control that does not make sense is encountered.
1055 9. Added the /retain_sender option to "control=submission".
1057 10. $recipients is now available in the predata ACL (oversight).
1059 11. Tidy the search cache before the fork to do a delivery from a message
1060 received from the command line. Otherwise the child will trigger a lookup
1061 failure and thereby defer the delivery if it tries to use (for example) a
1062 cached ldap connection that the parent has called unbind on.
1064 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1065 of $address_data from the recipient verification was clobbered by the
1066 sender verification.
1068 13. The value of address_data from a sender verification is now available in
1069 $sender_address_data in subsequent conditions in the ACL statement.
1071 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1073 15. Added a new option "connect=<time>" to callout options, to set a different
1076 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1077 was its contents. (It was OK if the option was not defined at all.)
1079 17. A "Completed" log line is now written for messages that are removed from
1080 the spool by the -Mrm option.
1082 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1083 information about exactly what failed.
1085 19. Added -dd to debug only the daemon process.
1087 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1088 handles timeouts, both on the server side and network timeouts. Renamed the
1089 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1091 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1092 It is now set to "smtps".
1094 22. $host_address is now set to the target address during the checking of
1095 ignore_target_hosts.
1097 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1098 being passed; this would have caused $sender_host_name to have been used if
1099 matching the list had actually called for a host name (not very likely,
1100 since this list is usually IP addresses). A host name is now passed as
1103 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1104 code that specifies a non-restarting handler (typically sigaction() in
1105 modern systems) in an attempt to fix a rare and obscure crash bug.
1107 25. Narrowed the window for a race in the daemon that could cause it to ignore
1108 SIGCHLD signals. This is not a major problem, because they are used only to
1109 wake it up if nothing else does.
1111 26. A malformed maildirsize file could cause Exim to calculate negative values
1112 for the mailbox size or file count. Odd effects could occur as a result.
1113 The maildirsize information is now recalculated if the size or filecount
1116 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1117 support for a long time. Removed HAVE_SYS_VFS_H.
1119 28. Installed the latest version of exipick from John Jetmore.
1121 29. In an address list, if the pattern was not a regular expression, an empty
1122 subject address (from a bounce message) matched only if the pattern was an
1123 empty string. Non-empty patterns were not even tested. This was the wrong
1124 because it is perfectly reasonable to use an empty address as part of a
1125 database query. An empty address is now tested by patterns that are
1126 lookups. However, all the other forms of pattern expect the subject to
1127 contain a local part and a domain, and therefore, for them, an empty
1128 address still always fails if the pattern is not itself empty.
1130 30. Exim went into a mad DNS loop when attempting to do a callout where the
1131 host was specified on an smtp transport, and looking it up yielded more
1132 than one IP address.
1134 31. Re-factored the code for checking spool and log partition space into a
1135 function that finds that data and another that does the check. The former
1136 is then used to implement four new variables: $spool_space, $log_space,
1137 $spool_inodes, and $log_inodes.
1139 32. The RFC2047 encoding function was originally intended for short strings
1140 such as real names; it was not keeping to the 75-character limit for
1141 encoded words that the RFC imposes. It now respects the limit, and
1142 generates multiple encoded words if necessary. To be on the safe side, I
1143 have increased the buffer size for the ${rfc2047: expansion operator from
1146 33. It is now permitted to omit both strings after an "if" condition; if the
1147 condition is true, the result is "true". As before, when the second string
1148 is omitted, a false condition yields an empty string. This makes it less
1149 cumbersome to write custom ACL and router conditions.
1151 34. Failure to deliver a bounce message always caused it to be frozen, even if
1152 there was an errors_to setting on the router. The errors_to setting is now
1155 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1156 canonical form (fully expanded) before being placed in
1157 $sender_host_address.
1159 36. The table in the code that translates DNS record types into text (T_A to
1160 "A" for instance) was missing entries for NS and CNAME. It is just possible
1161 that this could have caused confusion if both these types were looked up
1162 for the same domain, because the text type is used as part of Exim's
1163 per-process caching. But the chance of anyone hitting this buglet seems
1166 37. The dnsdb lookup has been extended in a number of ways.
1168 (1) There is a new type, "zns", which walks up the domain tree until it
1169 finds some nameserver records. It should be used with care.
1171 (2) There is a new type, "mxh", which is like "mx" except that it returns
1172 just the host names, not the priorities.
1174 (3) It is now possible to give a list of domains (or IP addresses) to be
1175 looked up. The behaviour when one of the lookups defers can be
1176 controlled by a keyword.
1178 (4) It is now possible to specify the separator character for use when
1179 multiple records are returned.
1181 38. The dnslists ACL condition has been extended: it is now possible to supply
1182 a list of IP addresses and/or domains to be looked up in a particular DNS
1185 39. Added log_selector=+queue_time_overall.
1187 40. When running the queue in the test harness, wait just a tad after forking a
1188 delivery process, to get repeatability of debugging output.
1190 41. Include certificate and key file names in error message when GnuTLS fails
1191 to set them up, because the GnuTLS error message doesn't include the name
1192 of the failing file when there is a problem reading it.
1194 42. Allow both -bf and -bF in the same test run.
1196 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1198 44. The "Exiscan patch" is now merged into the mainline Exim source.
1200 45. Sometimes the final signoff response after QUIT could fail to get
1201 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1202 < 0 before doing a fflush(). This bug looks as though it goes back to the
1203 introduction of TLS in release 3.20, but "sometimes" must have been rare
1204 because the tests only now provoked it.
1206 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1207 (this can affect the format of dates).
1209 47. exim_tidydb, when checking for the continued existence of a message for
1210 which it has found a message-specific retry record, was not finding
1211 messages that were in split spool directories. Consequently, it was
1212 deleting retry records that should have stayed in existence.
1214 48. Steve fixed some bugs in eximstats.
1216 49. The SPA authentication driver was not abandoning authentication and moving
1217 on to the next authenticator when an expansion was forced to fail,
1218 contradicting the general specification for all authenticators. Instead it
1219 was generating a temporary error. It now behaves as specified.
1221 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1222 (the order specifies the preference for clients). The order is now AES256,
1223 AES128, 3DES, ARCFOUR128.
1225 51. Small patch to Sieve code - explicitly set From: when generating an
1228 52. Exim crashed if a remote delivery caused a very long error message to be
1229 recorded - for instance if somebody sent an entire SpamAssassin report back
1230 as a large number of 550 error lines. This bug was coincidentally fixed by
1231 increasing the size of one of Exim's internal buffers (big_buffer) that
1232 happened as part of the Exiscan merge. However, to be on the safe side, I
1233 have made the code more robust (and fixed the comments that describe what
1236 53. Now that there can be additional text after "Completed" in log lines (if
1237 the queue_time_overall log selector is set), a one-byte patch to exigrep
1238 was needed to allow it to recognize "Completed" as not the last thing in
1241 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1242 patch that reportedly fixes this has been added. I am not expert enough to
1243 create a test for it. This is what the patch creator wrote:
1245 "I found a little strange behaviour of ldap code when working with
1246 Windows 2003 AD Domain, where users was placed in more than one
1247 Organization Units. When I tried to give exim partial DN, the exit code
1248 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1249 But simultaneously result of request was absolutely normal ldap result,
1250 so I produce this patch..."
1252 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1253 REFERENCE, so I have modified the code to exclude the patch when that macro
1256 55. Some experimental protocols are using DNS PTR records for new purposes. The
1257 keys for these records are domain names, not reversed IP addresses. The
1258 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1259 leaves it alone. Component reversal etc. now happens only for IP addresses.
1262 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1264 57. Double the size of the debug message buffer (to 2048) so that more of very
1265 long debug lines gets shown.
1267 58. The exicyclog utility now does better if the number of log files to keep
1268 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1270 59. Two changes related to the smtp_active_hostname option:
1272 (1) $smtp_active_hostname is now available as a variable.
1273 (2) The default for smtp_banner uses $smtp_active_hostname instead
1274 of $primary_hostname.
1276 60. The host_aton() function is supposed to be passed a string that is known
1277 to be a valid IP address. However, in the case of IPv6 addresses, it was
1278 not checking this. This is a hostage to fortune. Exim now panics and dies
1279 if the condition is not met. A case was found where this could be provoked
1280 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1281 components; fortuitously, this particular loophole had already been fixed
1282 by change 4.50/55 above.
1284 If there are any other similar loopholes, the new check in host_aton()
1285 itself should stop them being exploited. The report I received stated that
1286 data on the command line could provoke the exploit when Exim was running as
1287 exim, but did not say which command line option was involved. All I could
1288 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1289 running as the user.
1292 61. There was a buffer overflow vulnerability in the SPA authentication code
1293 (which came originally from the Samba project). I have added a test to the
1294 spa_base64_to_bits() function which I hope fixes it.
1297 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1298 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1300 63. The daemon start-up calls getloadavg() while still root for those OS that
1301 need the first call to be done as root, but it missed one case: when
1302 deliver_queue_load_max is set with deliver_drop_privilege. This is
1303 necessary for the benefit of the queue runner, because there is no re-exec
1304 when deliver_drop_privilege is set.
1306 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1307 This has been fixed.
1309 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1310 was in use, was not putting the data itself into the right store pool;
1311 consequently, it could be overwritten for a subsequent message in the same
1312 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1315 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1317 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1318 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1319 the function were treating the return as a boolean value, which happened to
1320 work because 0=false and not-0=true, but is not correct code.
1322 68. The host_aton() function was not handling scoped IPv6 addresses (those
1323 with, for example, "%eth0" on the end) correctly.
1325 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1326 negated items (that is, ~something) in unsigned ints. Some compilers
1327 apparently mutter when there is no cast.
1329 70. If an address verification called from an ACL failed, and did not produce a
1330 user-specific message (i.e. there was only a "system" message), nothing was
1331 put in $acl_verify_message. In this situation, it now puts the system
1334 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1335 session; change 4.31/43 added the unwanted input to the log line - except
1336 that it did not do this in the start of session case. It now does.
1338 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1339 This is wrong and can cause the other end to generate a synchronization
1340 error if it is another Exim or anything else that does the synchronization
1341 check. A QUIT command is no longer sent after a timeout.
1343 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1344 during host lookups.
1346 74. The defer_ok option of callout verification was not working if it was used
1347 when verifying addresses in header lines, that is, for this case:
1349 verify = header_sender/callout=defer_ok
1351 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1352 those file descriptors could be used for SMTP connections. If anything
1353 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1354 could be sent to the SMTP client, causing chaos. The daemon now opens
1355 stdin, stdout, and stderr to /dev/null when it puts itself into the
1358 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1359 log by default. The user can override this with suitable Perl magic.
1361 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1362 the log message when discard triggers, was not working for the DATA ACL or
1363 for the non-SMTP ACL.
1365 78. Error message wording change in sieve.c.
1367 79. If smtp_accept_max_per_host was set, the number of connections could be
1368 restricted to fewer than expected, because the daemon was trying to set up
1369 a new connection before checking whether the processes handling previous
1370 connections had finished. The check for completed processes is now done
1371 earlier. On busy systems, this bug wouldn't be noticed because something
1372 else would have woken the daemon, and it would have reaped the completed
1375 80. If a message was submitted locally by a user whose login name contained one
1376 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1377 It caused a spool format error. I have fixed the spool reading code. A
1378 related problem was that the "from" clause in the Received: line became
1379 illegal because of the space(s). It is now covered by ${quote_local_part.
1381 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1384 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1386 83. Patch from Sieve maintainer to fix unterminated string problem in
1387 "vacation" handling.
1389 84. Some minor changes to the Linux configuration files to help with other
1390 OS variants using glibc.
1392 85. One more patch for Sieve to update vacation handling to latest spec.
1395 ----------------------------------------------------
1396 See the note above about the 4.44 and 4.50 releases.
1397 ----------------------------------------------------
1403 1. Change 4.43/35 introduced a bug that caused file counts to be
1404 incorrectly computed when quota_filecount was set in an appendfile
1407 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1408 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1411 3. Give more explanation in the error message when the command for a transport
1412 filter fails to execute.
1414 4. There are several places where Exim runs a non-Exim command in a
1415 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1416 was being done only for the command run by the queryprogram router. It is
1417 now done for all such subprocesses. The other cases are: ${run, transport
1418 filters, and the commands run by the lmtp and pipe transports.
1420 5. Some older OS have a limit of 256 on the maximum number of file
1421 descriptors. Exim was using setrlimit() to set 1000 as a large value
1422 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1423 systems. I've change it so that if it can't get 1000, it tries for 256.
1425 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1426 was an oversight, and furthermore, ever since the addition of extra
1427 controls (e.g. 4.43/32), the checks on when to allow different forms of
1428 "control" were broken. There should now be diagnostics for all cases when a
1429 control that does not make sense is encountered.
1431 7. $recipients is now available in the predata ACL (oversight).
1433 8. Tidy the search cache before the fork to do a delivery from a message
1434 received from the command line. Otherwise the child will trigger a lookup
1435 failure and thereby defer the delivery if it tries to use (for example) a
1436 cached ldap connection that the parent has called unbind on.
1438 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1439 of $address_data from the recipient verification was clobbered by the
1440 sender verification.
1442 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1443 was its contents. (It was OK if the option was not defined at all.)
1445 11. A "Completed" log line is now written for messages that are removed from
1446 the spool by the -Mrm option.
1448 12. $host_address is now set to the target address during the checking of
1449 ignore_target_hosts.
1451 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1452 being passed; this would have caused $sender_host_name to have been used if
1453 matching the list had actually called for a host name (not very likely,
1454 since this list is usually IP addresses). A host name is now passed as
1457 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1458 code that specifies a non-restarting handler (typically sigaction() in
1459 modern systems) in an attempt to fix a rare and obscure crash bug.
1461 15. Narrowed the window for a race in the daemon that could cause it to ignore
1462 SIGCHLD signals. This is not a major problem, because they are used only to
1463 wake it up if nothing else does.
1465 16. A malformed maildirsize file could cause Exim to calculate negative values
1466 for the mailbox size or file count. Odd effects could occur as a result.
1467 The maildirsize information is now recalculated if the size or filecount
1470 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1471 support for a long time. Removed HAVE_SYS_VFS_H.
1473 18. Updated exipick to current release from John Jetmore.
1475 19. Allow an empty sender to be matched against a lookup in an address list.
1476 Previously the only cases considered were a regular expression, or an
1479 20. Exim went into a mad DNS lookup loop when doing a callout where the
1480 host was specified on the transport, if the DNS lookup yielded more than
1483 21. The RFC2047 encoding function was originally intended for short strings
1484 such as real names; it was not keeping to the 75-character limit for
1485 encoded words that the RFC imposes. It now respects the limit, and
1486 generates multiple encoded words if necessary. To be on the safe side, I
1487 have increased the buffer size for the ${rfc2047: expansion operator from
1490 22. Failure to deliver a bounce message always caused it to be frozen, even if
1491 there was an errors_to setting on the router. The errors_to setting is now
1494 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1495 canonical form (fully expanded) before being placed in
1496 $sender_host_address.
1498 24. Updated eximstats to version 1.33
1500 25. Include certificate and key file names in error message when GnuTLS fails
1501 to set them up, because the GnuTLS error message doesn't include the name
1502 of the failing file when there is a problem reading it.
1504 26. Expand error message when OpenSSL has problems setting up cert/key files.
1507 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1508 (this can affect the format of dates).
1510 28. exim_tidydb, when checking for the continued existence of a message for
1511 which it has found a message-specific retry record, was not finding
1512 messages that were in split spool directories. Consequently, it was
1513 deleting retry records that should have stayed in existence.
1515 29. eximstats updated to version 1.35
1516 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1517 1.35 - bugfix such that pie charts by volume are generated correctly
1519 30. The SPA authentication driver was not abandoning authentication and moving
1520 on to the next authenticator when an expansion was forced to fail,
1521 contradicting the general specification for all authenticators. Instead it
1522 was generating a temporary error. It now behaves as specified.
1524 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1525 (the order specifies the preference for clients). The order is now AES256,
1526 AES128, 3DES, ARCFOUR128.
1528 31. Small patch to Sieve code - explicitly set From: when generating an
1531 32. Exim crashed if a remote delivery caused a very long error message to be
1532 recorded - for instance if somebody sent an entire SpamAssassin report back
1533 as a large number of 550 error lines. This bug was coincidentally fixed by
1534 increasing the size of one of Exim's internal buffers (big_buffer) that
1535 happened as part of the Exiscan merge. However, to be on the safe side, I
1536 have made the code more robust (and fixed the comments that describe what
1539 33. Some experimental protocols are using DNS PTR records for new purposes. The
1540 keys for these records are domain names, not reversed IP addresses. The
1541 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1542 leaves it alone. Component reversal etc. now happens only for IP addresses.
1545 34. The host_aton() function is supposed to be passed a string that is known
1546 to be a valid IP address. However, in the case of IPv6 addresses, it was
1547 not checking this. This is a hostage to fortune. Exim now panics and dies
1548 if the condition is not met. A case was found where this could be provoked
1549 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1550 components; fortuitously, this particular loophole had already been fixed
1551 by change 4.50/55 or 4.44/33 above.
1553 If there are any other similar loopholes, the new check in host_aton()
1554 itself should stop them being exploited. The report I received stated that
1555 data on the command line could provoke the exploit when Exim was running as
1556 exim, but did not say which command line option was involved. All I could
1557 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1558 running as the user.
1561 35. There was a buffer overflow vulnerability in the SPA authentication code
1562 (which came originally from the Samba project). I have added a test to the
1563 spa_base64_to_bits() function which I hope fixes it.
1566 36. The daemon start-up calls getloadavg() while still root for those OS that
1567 need the first call to be done as root, but it missed one case: when
1568 deliver_queue_load_max is set with deliver_drop_privilege. This is
1569 necessary for the benefit of the queue runner, because there is no re-exec
1570 when deliver_drop_privilege is set.
1572 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1573 was in use, was not putting the data itself into the right store pool;
1574 consequently, it could be overwritten for a subsequent message in the same
1575 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1578 38. Sometimes the final signoff response after QUIT could fail to get
1579 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1580 < 0 before doing a fflush(). This bug looks as though it goes back to the
1581 introduction of TLS in release 3.20, but "sometimes" must have been rare
1582 because the tests only now provoked it.
1588 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1589 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1590 Now it returns an integer. A number of calls were still expecting a T/F
1591 return. Fortuitously, in all cases, the tests worked in OK situations,
1592 which is the norm. However, things would have gone wrong on any write
1593 failures on the smtp file descriptor. This function is used when sending
1594 messages over SMTP and also when doing verify callouts.
1596 2. When Exim is called to do synchronous delivery of a locally submitted
1597 message (the -odf or -odi options), it no longer closes stderr before doing
1600 3. Implemented the mua_wrapper option.
1602 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1604 5. Implemented the functions header_remove(), header_testname(),
1605 header_add_at_position(), and receive_remove_recipient(), and exported them
1608 6. If an ACL "warn" statement specified the addition of headers, Exim already
1609 inserted X-ACL-Warn: at the start if there was no header name. However, it
1610 was not making this test for the second and subsequent header lines if
1611 there were newlines in the string. This meant that an invalid header could
1612 be inserted if Exim was badly configured.
1614 7. Allow an ACL "warn" statement to add header lines at the start or after all
1615 the Received: headers, as well as at the end.
1617 8. Added the rcpt_4xx retry error code.
1619 9. Added postmaster_mailfrom=xxx to callout verification option.
1621 10. Added mailfrom=xxxx to the callout verification option, for verify=
1624 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1625 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1627 12. Inserted some casts to stop certain compilers warning when using pointer
1628 differences as field lengths or precisions in printf-type calls (mostly
1629 affecting debugging statements).
1631 13. Added optional readline() support for -be (dynamically loaded).
1633 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1634 same clock tick as a message's arrival, so that its received time was the
1635 same as the "first fail" time on the retry record, and that message
1636 remained on the queue past the ultimate address timeout, every queue runner
1637 would try a delivery (because it was past the ultimate address timeout) but
1638 after another failure, the ultimate address timeout, which should have then
1639 bounced the address, did not kick in. This was a "< instead of <=" error;
1640 in most cases the first failure would have been in the next clock tick
1641 after the received time, and all would be well.
1643 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1644 being recognized when the domain list was tested by the match_domain
1645 condition in an expansion string.
1647 16. Added the ${str2b64: operator.
1649 17. Exim was always calling setrlimit() to set a large limit for the number of
1650 processes, without checking whether the existing limit was already
1651 adequate. (It did check for the limit on file descriptors.) Furthermore,
1652 errors from getrlimit() and setrlimit() were being ignored. Now they are
1653 logged to the main and panic logs, but Exim does carry on, to try to do its
1654 job under whatever limits there are.
1656 18. Imported PCRE 5.0.
1658 19. Trivial typo in log message " temporarily refused connection" (the leading
1661 20. If the log selector return_path_on_delivery was set and an address was
1662 redirected to /dev/null, the delivery process crashed because it assumed
1663 that a return path would always be set for a "successful" delivery. In this
1664 case, the whole delivery is bypassed as an optimization, and therefore no
1667 21. Internal re-arrangement: the function for sending a challenge and reading
1668 a response while authentication was assuming a zero-terminated challenge
1669 string. It's now changed to take a pointer and a length, to allow for
1670 binary data in such strings.
1672 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1674 23. Exim was not respecting finduser_retries when seeking the login of the
1675 uid under which it was called; it was always trying 10 times. (The default
1676 setting of finduser_retries is zero.) Also, it was sleeping after the final
1677 failure, which is pointless.
1679 24. Implemented tls_on_connect_ports.
1681 25. Implemented acl_smtp_predata.
1683 26. If the domain in control=submission is set empty, Exim assumes that the
1684 authenticated id is a complete email address when it generates From: or
1685 Sender: header lines.
1687 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1688 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1689 chown and chgrp in /bin and hostname in /usr/bin.
1691 28. Exim was keeping the "process log" file open after each use, just as it
1692 does for the main log. This opens the possibility of it remaining open for
1693 long periods when the USR1 signal hits a daemon. Occasional processlog
1694 errors were reported, that could have been caused by this. Anyway, it seems
1695 much more sensible not to leave this file open at all, so that is what now
1698 29. The long-running daemon process does not normally write to the log once it
1699 has entered its main loop, and it closes the log before doing so. This is
1700 so that log files can straightforwardly be renamed and moved. However,
1701 there are a couple of unusual error situations where the daemon does write
1702 log entries, and I had neglected to close the log afterwards.
1704 30. The text of an SMTP error response that was received during a remote
1705 delivery was being truncated at 512 bytes. This is too short for some of
1706 the long messages that one sometimes sees. I've increased the limit to
1709 31. It is now possible to make retry rules that apply only when a message has a
1710 specific sender, in particular, an empty sender.
1712 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1713 it possible to be selective about when SMTP synchronization is enforced.
1715 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1717 32. Implemented hosts_connection_nolog.
1719 33. Added an ACL for QUIT.
1721 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1724 35. Added mailbox_size and mailbox_filecount to appendfile.
1726 36. Added control = no_multiline_responses to ACLs.
1728 37. There was a bug in the logic of the code that waits for the clock to tick
1729 in the case where the clock went backwards by a substantial amount such
1730 that the microsecond fraction of "now" was more than the microsecond
1731 fraction of "then" (but the whole seconds number was less).
1733 38. Added support for the libradius Radius client library this is found on
1734 FreeBSD (previously only the radiusclient library was supported).
1740 1. When certain lookups returned multiple values in the form name=value, the
1741 quoting of the values was not always being done properly. Specifically:
1742 (a) If the value started with a double quote, but contained no whitespace,
1744 (b) If the value contained whitespace other than a space character (i.e.
1745 tabs or newlines or carriage returns) it was not quoted.
1746 This fix has been applied to the mysql and pgsql lookups by writing a
1747 separate quoting function and calling it from the lookup code. The fix
1748 should probably also be applied to nisplus, ibase and oracle lookups, but
1749 since I cannot test any of those, I have not disturbed their existing code.
1751 2. A hit in the callout cache for a specific address caused a log line with no
1752 reason for rejecting RCPT. Now it says "Previous (cached) callout
1753 verification failure".
1755 3. There was an off-by-one bug in the queryprogram router. An over-long
1756 return line was truncated at 256 instead of 255 characters, thereby
1757 overflowing its buffer with the terminating zero. As well as fixing this, I
1758 have increased the buffer size to 1024 (and made a note to document this).
1760 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1761 when Exim is waiting for an SMTP response from a remote server, Exim
1762 restarts its select() call on the socket, thereby resetting its timeout.
1763 This is not a problem when such interrupts are rare. Somebody set up a cron
1764 job to run exiwhat every 2 minutes, which is less than the normal select()
1765 timeout (5 or 10 minutes). This meant that the select() timeout never
1766 kicked in because it was always reset. I have fixed this by comparing the
1767 time when an interrupt arrives with the time at the start of the first call
1768 to select(). If more time than the timeout has elapsed, the interrupt is
1769 treated as a timeout.
1771 5. Some internal re-factoring in preparation for the addition of Sieve
1772 extensions (by MH). In particular, the "personal" test is moved to a
1773 separate function, and given an option for scanning Cc: and Bcc: (which is
1774 not set for Exim filters).
1776 6. When Exim created an email address using the login of the caller as the
1777 local part (e.g. when creating a From: or Sender: header line), it was not
1778 quoting the local part when it contained special characters such as @.
1780 7. Installed new OpenBSD configuration files.
1782 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1783 try to make them clearer.
1785 9. Callout options, other than the timeout value, were being ignored when
1786 verifying sender addresses in header lines. For example, when using
1788 verify = header_sender/callout=no_cache
1790 the cache was (incorrectly) being used.
1792 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1793 only the Cygwin environment.
1795 11. When return_path_on_delivery was set as a log selector, if different remote
1796 addresses in the same message used different return paths and parallel
1797 remote delivery occurred, the wrong values would sometimes be logged.
1798 (Whenever a remote delivery process finished, the return path value from
1799 the most recently started remote delivery process was logged.)
1801 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1802 lines when AUTH and/or TLS are in use. This is the "received protocol"
1803 field. Exim used to use "asmtp" for authenticated SMTP, without any
1804 indication (in the protocol name) for TLS use. Now it follows the RFC and
1805 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1806 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1807 names appear in log lines as well as in Received: header lines.
1809 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1810 extensions, and comparison tests, and to fix some bugs.
1812 14. Changes to the "personal" filter test:
1814 (1) The test was buggy in that it was just doing the equivalent of
1815 "contains" tests on header lines. For example, if a user's address was
1816 anne@some.where, the "personal" test would incorrectly be true for
1818 To: susanne@some.where
1820 This test is now done by extracting each address from the header in turn,
1821 and checking the entire address. Other tests that are part of "personal"
1822 are now done using regular expressions (for example, to check local parts
1823 of addresses in From: header lines).
1825 (2) The list of non-personal local parts in From: addresses has been
1826 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1827 taken from the Sieve specification recommendations.
1829 (3) If the message contains any header line starting with "List-" it is
1830 treated as non-personal.
1832 (4) The test for "circular" in the Subject: header line has been removed
1833 because it now seems ill-conceived.
1835 15. Minor typos in src/EDITME comments corrected.
1837 16. Installed latest exipick from John Jetmore.
1839 17. If headers_add on a router specified a text string that was too long for
1840 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1841 of string_sprintf() is now avoided.
1843 18. $message_body_size was not set (it was always zero) when running the DATA
1844 ACL and the local_scan() function.
1846 19. For the "mail" command in an Exim filter, no default was being set for
1847 the once_repeat time, causing a random time value to be used if "once" was
1848 specified. (If the value happened to be <= 0, no repeat happened.) The
1849 default is now 0s, meaning "never repeat". The "vacation" command was OK
1850 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1851 (I found it when inspecting the code).
1853 20. There is now an overall timeout for performing a callout verification. It
1854 defaults to 4 times the callout timeout, which applies to individual SMTP
1855 commands during the callout. The overall timeout applies when there is more
1856 than one host that can be tried. The timeout is checked before trying the
1857 next host. This prevents very long delays if there are a large number of
1858 hosts and all are timing out (e.g. when the network connections are timing
1859 out). The value of the overall timeout can be changed by specifying an
1860 additional sub-option for "callout", called "maxwait". For example:
1862 verify = sender/callout=5s,maxwait=20s
1864 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1865 to the end before writing, but this should make it even safer).
1867 22. Exim was forgetting that it had advertised PIPELINING for the second and
1868 subsequent messages on an SMTP connection. It was also not resetting its
1869 memory on STARTTLS and an internal HELO.
1871 23. When Exim logs an SMTP synchronization error within a session, it now
1872 records whether PIPELINING has been advertised or not.
1874 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1875 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1876 rather than long int.
1878 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1880 26. Added the never_mail option to autoreply.
1886 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1887 crash if the getsockname() call failed; this can happen if a connection is
1888 closed very soon after it is established. The problem was simply in the
1889 order in which certain operations were done, causing Exim to try to write
1890 to the SMTP stream before it had set up the file descriptor. The bug has
1891 been fixed by making things happen in the correct order.
1897 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1898 before the connection was closed, thus losing the rejection response.
1900 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1901 some early Solaris releases, but causes trouble in current releases where
1902 socklen_t is defined.
1904 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1907 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1910 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1911 that releasing the top of it at the end releases what was used for sub-
1912 expansions (unless the block got too big). However, discard this block if
1913 the first thing is a variable or header, so that we can use its block when
1914 it is dynamic (useful for very large $message_headers, for example).
1916 6. Lookups now cache *every* query, not just the most recent. A new, separate
1917 store pool is used for this. It can be recovered when all lookup caches are
1918 flushed. Lookups now release memory at the end of their result strings.
1919 This has involved some general refactoring of the lookup sources.
1921 7. Some code has been added to the store_xxx() functions to reduce the amount
1922 of flapping under certain conditions.
1924 8. log_incoming_interface used to affect only the <= reception log lines. Now
1925 it causes the local interface and port to be added to several more SMTP log
1926 lines, for example "SMTP connection from", and rejection lines.
1928 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1930 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1932 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1933 could be overwritten at the end of the current message (or the start of a
1934 new message if it was set in a HELO ACL). The value is now preserved for
1935 the duration of the SMTP connection.
1937 12. If a transport had a headers_rewrite setting, and a matching header line
1938 contained an unqualified address, that address was qualified, even if it
1939 did not match any rewriting rules. The underlying bug was that the values
1940 of the flags that permit the existence of unqualified sender and recipient
1941 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1942 non-local messages, and by -bnq for local messages) were not being
1943 preserved with the message after it was received.
1945 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1946 "next input=" as part of the text comprising the host identity instead of
1947 the correct text. The code was using the same buffer for two different
1948 strings. However, depending on which order the printing function evaluated
1949 its arguments, the bug did not always show up. Under Linux, for example, my
1950 test suite worked just fine.
1952 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1953 doesn't work with some older versions of Perl. It has been changed to "my",
1954 which in any case is probably the better facility to use.
1956 15. A really picky compiler found some instances of statements for creating
1957 error messages that either had too many or two few arguments for the format
1960 16. The size of the buffer for calls to the DNS resolver has been increased
1961 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1962 for addresses that have a lot of PTR records. This alleviates a problem; it
1963 does not fully solve it.
1965 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1966 buffer now truncates the list and logs the incident, which is the same
1967 action as happens when Exim is looking up a host name and its aliases.
1968 Previously in this situation something unpredictable would happen;
1969 sometimes it was "internal error: store_reset failed".
1971 18. If a server dropped the connection unexpectedly when an Exim client was
1972 using GnuTLS and trying to read a response, the client delivery process
1973 crashed while trying to generate an error log message.
1975 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1976 string, for example:
1978 warn message = H1: something\nH2: something
1980 the text was added as a single header line from Exim's point of view
1981 though it ended up OK in the delivered message. However, searching for the
1982 second and subsequent header lines using $h_h2: did not work. This has been
1983 fixed. Similarly, if a system filter added multiple headers in this way,
1984 the routers could not see them.
1986 20. Expanded the error message when iplsearch is called with an invalid key to
1987 suggest using net-iplsearch in a host list.
1989 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1990 ACLs are no longer actually imposed (and a message to that effect is
1993 22. If a "gecos" field in a passwd entry contained escaped characters, in
1994 particular, if it contained a \" sequence, Exim got it wrong when building
1995 a From: or a Sender: header from that name. A second bug also caused
1996 incorrect handling when an unquoted " was present following a character
1997 that needed quoting.
1999 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2000 was not being matched caselessly.
2002 24. Arranged for all hyphens in the exim.8 source to be escaped with
2005 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2006 properly. Recipient callout cache records were still being keyed to include
2007 the sender, even when use_sender was set false. This led to far more
2008 callouts that were necessary. The sender is no longer included in the key
2009 when use_sender is false.
2011 26. Added "control = submission" modifier to ACLs.
2013 27. Added the ${base62d: operator to decode base 62 numbers.
2015 28. dnsdb lookups can now access SRV records.
2017 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2018 the configuration file.
2020 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2021 (-v) mode. This makes the output for a verbose queue run more intelligible.
2023 31. Added a use_postmaster feature to recipient callouts.
2025 32. Added the $body_zerocount variable, containing the number of binary zero
2026 bytes in the message body.
2028 33. The time of last modification of the "new" subdirectory is now used as the
2029 "mailbox time last read" when there is a quota error for a maildir
2032 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2034 35. Added +ignore_unknown as a special item in host lists.
2036 36. Code for decoding IPv6 addresses in host lists is now included, even if
2037 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2038 address was recognized as an IP address, but was then not correctly decoded
2039 into binary, causing unexpected and incorrect effects when compared with
2046 1. Very minor rewording of debugging text in manualroute to say "list of
2047 hosts" instead of "hostlist".
2049 2. If verify=header_syntax was set, and a header line with an unqualified
2050 address (no domain) and a large number of spaces between the end of the
2051 name and the colon was received, the reception process suffered a buffer
2052 overflow, and (when I tested it) crashed. This was caused by some obsolete
2053 code that should have been removed. The fix is to remove it!
2055 3. When running in the test harness, delay a bit after writing a bounce
2056 message to get a bit more predictability in the log output.
2058 4. Added a call to search_tidyup() just before forking a reception process. In
2059 theory, someone could use a lookup in the expansion of smtp_accept_max_
2060 per_host which, without the tidyup, could leave open a database connection.
2062 5. Added the variables $recipient_data and $sender_data which get set from a
2063 lookup success in an ACL "recipients" or "senders" condition, or a router
2064 "senders" option, similar to $domain_data and $local_part_data.
2066 6. Moved the writing of debug_print from before to after the "senders" test
2069 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2070 problems for message scanning, either using a data ACL, or using
2071 local_scan() because the Received: header was not generated till after they
2072 were called (in order to set the time as the time of reception completion).
2073 I have revised the way this works. The header is now generated after the
2074 body is received, but before the ACL or local_scan() are called. After they
2075 are run, the timestamp in the header is updated.
2081 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2082 before starting a queue runner without re-exec. This happened only when
2083 deliver_drop_privilege was set or when the Exim user was set to root. The
2084 effect of the bug was that timeouts during subsequent deliveries caused
2085 crashes instead of being properly handled. The handler is now left at its
2086 default (and expected) setting.
2088 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2089 message, again when deliver_drop_privilege is set or Exim is run as root.
2090 The bug described in (1) was not present in this case, but the tidying up
2091 of the other signals was missing. I have made the two cases consistent.
2093 3. The ignore_target_hosts setting on a manualroute router was being ignored
2094 for hosts that were looked up using the /MX notation.
2096 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2099 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2100 operated on the sender address. After changing the $sender_address to <>
2101 for the sender address verify, Exim was re-instated it as the original
2102 (before rewriting) address, but remembering that it had rewritten it, so it
2103 wasn't rewriting it again. This bug also had the effect of breaking the
2104 sender address verification caching when the sender address was rewritten.
2106 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2107 This has been changed so that if the ip literal address matches
2108 ignore_target_hosts, the router declines.
2110 7. Added expansion conditions match_domain, match_address, and match_local_
2111 part (NOT match_host).
2113 8. The placeholder for the Received: header didn't have a length field set.
2115 9. Added code to Exim itself and to exim_lock to test for a specific race
2116 condition that could lead to file corruption when using MBX delivery. The
2117 issue is with the lockfile that is created in /tmp. If this file is removed
2118 after a process has opened it but before that process has acquired a lock,
2119 there is the potential for a second process to recreate the file and also
2120 acquire a lock. This could lead to two Exim processes writing to the file
2121 at the same time. The added code performs the same test as UW imapd; it
2122 checks after acquiring the lock that its file descriptor still refers to
2123 the same named file.
2125 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2126 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2129 11. Added the smtp_active_hostname option. If used, this will typically be made
2130 to depend on the incoming interface address. Because $interface_address is
2131 not set up until the daemon has forked a reception process, error responses
2132 that can happen earlier (such as "too many connections") no longer contain
2135 12. If an expansion in a condition on a "warn" statement fails because a lookup
2136 defers, the "warn" statement is abandoned, and the next ACL statement is
2137 processed. Previously this caused the whole ACL to be aborted.
2139 13. Added the iplsearch lookup type.
2141 14. Added ident_timeout as a log selector.
2143 15. Added tls_certificate_verified as a log selector.
2145 16. Added a global option tls_require_ciphers (compare the smtp transport
2146 option of the same name). This controls incoming TLS connections.
2148 17. I finally figured out how to make tls_require_ciphers do a similar thing
2149 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2150 before starting the TLS session.
2152 18. Tabs are now shown as \t in -bP output.
2154 19. If the log selector return_path_on_delivery was set, Exim crashed when
2155 bouncing a message because it had too many Received: header lines.
2157 20. If two routers both had headers_remove settings, and the first one included
2158 a superfluous trailing colon, the final name in the first list and the
2159 first name in the second list were incorrectly joined into one item (with a
2160 colon in the middle).
2166 1. Added -C and -D options to the exinext utility, mainly to make it easier
2167 to include in the automated testing, but these could be helpful when
2168 multiple configurations are in use.
2170 2. The exinext utility was not formatting the output nicely when there was
2171 an alternate port involved in the retry record key, nor when there was a
2172 message id as well (for retries that were specific to a specific message
2173 and a specific host). It was also confused by IPv6 addresses, because of
2174 the additional colons they contain. I have fixed the IPv4 problem, and
2175 patched it up to do a reasonable job for IPv6.
2177 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2178 delivery, the log line now contains "pipelined" if PIPELINING was used.
2180 4. An SMTP transport process used to panic and die if the bind() call to set
2181 an explicit outgoing interface failed. This has been changed; it is now
2182 treated in the same way as a connect() failure.
2184 5. A reference to $sender_host_name in the part of a conditional expansion
2185 that was being skipped was still causing a DNS lookup. This no longer
2188 6. The def: expansion condition was not recognizing references to header lines
2189 that used bh_ and bheader_.
2191 7. Added the _cache feature to named lists.
2193 8. The code for checking quota_filecount in the appendfile transport was
2194 allowing one more file than it should have been.
2196 9. For compatibility with Sendmail, the command line option
2204 and sets the incoming protocol and host name (for trusted callers). The
2205 host name and its colon can be omitted when only the protocol is to be set.
2206 Note the Exim already has two private options, -pd and -ps, that refer to
2207 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2208 "s", but I don't think that's a major issue.
2210 10. A number of refactoring changes to the code, none of which should affect
2213 (a) The number of logging options was getting close to filling up the
2214 32-bit word that was used as a bit map. I have split them into two classes:
2215 those that are passed in the argument to log_write(), and those that are
2216 only ever tested independently outside of that function. These are now in
2217 separate 32-bit words, so there is plenty of room for expansion again.
2218 There is no change in the user interface or the logging behaviour.
2220 (b) When building, for example, log lines, the code previously used a
2221 macro that called string_cat() twice, in order to add two strings. This is
2222 not really sufficiently general. Furthermore, there was one instance where
2223 it was actually wrong because one of the argument was used twice, and in
2224 one call a function was used. (As it happened, calling the function twice
2225 did not affect the overall behaviour.) The macro has been replaced by a
2226 function that can join an arbitrary number of extra strings onto a growing
2229 (c) The code for expansion conditions now uses a table and a binary chop
2230 instead of a serial search (which was left over from when there were very
2231 few conditions). Also, it now recognizes conditions like "pam" even when
2232 the relevant support is not compiled in: a suitably worded error message is
2233 given if an attempt is made to use such a condition.
2235 11. Added ${time_interval:xxxxx}.
2237 12. A bug was causing one of the ddress fields not to be passed back correctly
2238 from remote delivery subprocesses. The field in question was not being
2239 subsequently used, so this caused to problems in practice.
2241 13. Added new log selectors queue_time and deliver_time.
2243 14. Might have fixed a bug in maildirsizefile handling that threw up
2244 "unexpected character" debug warnings, and recalculated the data
2245 unnecessarily. In any case, I expanded the warning message to give more
2248 15. Added the message "Restricted characters in address" to the statements in
2249 the default ACL that block characters like @ and % in local parts.
2251 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2252 Three changes have been made:
2254 (a) There was a serious bug; a negative response to MAIL caused the whole
2255 recipient domain to be cached as invalid, thereby blocking all messages
2256 to all local parts at the same domain, from all senders. This bug has
2257 been fixed. The domain is no longer cached after a negative response to
2258 MAIL if the sender used is not empty.
2260 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2263 (c) A new callout option, "use_sender" has been added for people who want
2264 the modified behaviour.
2270 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2273 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2274 indeed breaks things for older releases.
2276 3. Added additional logging to the case where there is a problem reading data
2277 from a filter that is running in a subprocess using a pipe, in order to
2278 try to track down a specific problem.
2280 4. Testing facility fudge: when running in the test harness and attempting
2281 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2282 getting "No route to host". Convert this to a timeout.
2284 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2287 6. Some OS don't have socklen_t but use size_t instead. This affects the
2288 fifth argument of getsockopt() amongst other things. This is now
2289 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2290 can be set for individual OS. I have set it for SunOS5, OSF1, and
2291 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2292 some earlier ones do not.
2294 7. Change 4.30/15 was not doing the test caselessly.
2296 8. The standard form for an IPv6 address literal was being rejected by address
2297 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2298 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2299 this, as well as the form without the "IPv6" on the front (but only when
2300 address literals are enabled, of course).
2302 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2304 10. Exim crashed if a message with an empty sender address specified by -f
2305 encountered a router with an errors_to setting. This could be provoked only
2306 by a command such as
2310 where an empty string was supplied; "<>" did not hit this bug.
2312 11. Installed PCRE release 4.5.
2314 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2315 remained set. It is now erased.
2317 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2318 times from message ids (which are base 36 rather than the normal 62).
2320 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2321 were being counted as actual protocol errors, and logged if the log
2322 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2323 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2324 and DATA following a set of rejected RCPTs do not count as protocol errors.
2325 In other words, Exim assumes they were pipelined, though this may not
2326 actually be the case. Of course, in all cases the client gets an
2327 appropriate error code.
2329 15. If a lookup fails in an ACL condition, a message about the failure may
2330 be available; it is used if testing the ACL cannot continue, because most
2331 such messages specify what the cause of the deferral is. However, some
2332 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2333 that caused an old message to be retained and used if a later statement
2334 caused a defer, replacing the real cause of the deferral.
2336 16. If an IP address had so many PTR records that the DNS lookup buffer
2337 was not large enough to hold them, Exim could crash while trying to process
2338 the truncated data. It now detects and logs this case.
2340 17. Further to 4.21/58, another change has been made: if (and only if) the
2341 first line of a message (the first header line) ends with CRLF, a bare LF
2342 in a subsequent header line has a space inserted after it, so as not to
2343 terminate the header.
2345 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2346 unnecessarily, used atoi() instead of strtol(), and didn't check the
2347 termination when getting file sizes from file names by regex.
2349 19. Completely re-implemented the support for maildirsize files, in the light
2350 of a number of problems with the previous contributed implementation
2351 (4.30/29). In particular:
2353 . If the quota is zero, the maildirsize file is maintained, but no quota is
2356 . If the maildir directory does not exist, it is created before any attempt
2357 to write a maildirsize file.
2359 . The quota value in the file is just a cache; if the quota is changed in
2360 the transport, the new value overrides.
2362 . A regular expression is available for excluding directories from the
2365 20. The autoreply transport checks the characters in options that define the
2366 message's headers; it allows continued headers, but it was checking with
2367 isspace() after an embedded newline instead of explicitly looking for a
2370 21. If all the "regular" hosts to which an address was routed had passed their
2371 expiry times, and had not reached their retry times, the address was
2372 bounced, even if fallback hosts were defined. Now Exim should go on to try
2375 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2376 equivalent code in the SMTP transport. Some hosts send humungous responses
2377 to HELO/EHLO, more than 1024 it seems.
2379 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2380 gives compiler warnings in some environments. I've now done it "properly",
2383 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2384 (because of problems with the built-in one) was declared to return uschar *
2385 instead of char *, causing compiler failure.
2387 25. Fixed a file descriptor leak when processing alias/forward files.
2389 26. Fixed a minor format string issue in dbfn.c.
2391 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2393 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2394 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2397 29. When a sender address is verified, it is cached, to save repeating the test
2398 when there is more than one recipient in a message. However, when the
2399 verification involves a callout, it is possible for different callout
2400 options to be set for different recipients. It is too complicated to keep
2401 track of this in the cache, so now Exim always runs a verification when a
2402 callout is required, relying on the callout cache for the optimization.
2403 The overhead is duplication of the address routing, but this should not be
2406 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2407 to be verified with callout=postmaster, and the main callout worked but the
2408 postmaster check failed, the verification correctly failed. However, if a
2409 subsequent RCPT command asked for sender verification *without* the
2410 postmaster check, incorrect caching caused this verification also to fail,
2413 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2414 it was not caching the DNS options (qualify_single, search_parents) that
2415 were used when the lookup failed. A subsequent lookup with different
2416 options therefore always gave the same answer, though there were cases
2417 where it should not have. (Example: a "domains = !$mx_any" option on a
2418 dnslookup router: the "domains" option is always processed without any
2419 widening, but the router might have qualify_single set.) Now Exim uses the
2420 cached value only when the same options are set.
2422 32. Added John Jetmore's "exipick" utility to the distribution.
2424 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2425 than a timeout (e.g. a certificate is required, and is not provided), an
2426 Exim server now closes the connection immediately. Previously it waited for
2427 the client to close - but if the client is SSL, it seems that they each
2428 wait for each other, leading to a delay before one of them times out.
2430 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2431 maintained 0.8.x compatibility because I don't think many are using it, and
2432 it is clearly obsolete.
2434 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2437 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2438 client certificate was expired. A simple patch fixes this, though I don't
2439 understand the full logic of why the verify callback is called multiple
2442 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2443 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2444 which causes problems with some clients (such as the Certicom SSL Plus
2445 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2446 disables the coutermeasure allowing Eudora to connect."
2448 38. Exim was not checking that a write() to a log file succeeded. This could
2449 lead to Bad Things if a log got too big, in particular if it hit a file
2450 size limit. Exim now panics and dies if it cannot write to a log file, just
2451 as it does if it cannot open a log file.
2453 39. Modified OS/Makefile-Linux so that it now contains
2455 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2457 The two -D definitions ensure that Exim is compiled with large file
2458 support, which makes it possible to handle log files that are bigger than
2461 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2462 instance) a domain was checked against a named list that involved a lookup,
2463 causing $domain_data to be set, then another domain was checked against the
2464 same list, then the first domain was re-checked, the value of $domain_data
2465 after the final check could be wrong. In particular, if the second check
2466 failed, it could be set empty. This bug probably also applied to
2469 41. The strip_trailing_dot option was not being applied to the address given
2470 with the -f command-line option.
2472 42. The code for reading a message's header from the spool was incrementing
2473 $received_count, but never initializing it. This meant that the value was
2474 incorrect (doubled) while delivering a message in the same process in which
2475 it was received. In the most common configuration of Exim, this never
2476 happens - a fresh exec is done - but it can happen when
2477 deliver_drop_privilege is set.
2479 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2480 it now includes up to 150 characters of the unexpected data in the log
2483 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2484 and building data strings. The size of both of these buffers was 10 000
2485 bytes - far larger than anybody would *ever* want, thought I. Needless to
2486 say, somebody hit the limit. I have increased the maximum line length to
2487 20 000 and the maximum data length of concatenated lines to 100 000. I have
2488 also fixed two bugs, because there was no checking on these buffers. Tsk,
2489 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2490 buffer is too small.
2492 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2493 lsearch lookups. Now it does.
2495 46. When parsing a route_list item in a manualroute router, a fixed-length
2496 buffer was used for the list of hosts. I made this 1024 bytes long,
2497 thinking that nobody would ever have a list of hosts that long. Wrong.
2498 Somebody had a whole pile of complicated expansion conditions, and the
2499 string was silently truncated, leading to an expansion error. It turns out
2500 that it is easier to change to an unlimited length (owing to other changes
2501 that have happened since this code was originally written) than to build
2502 structure for giving a limitation error. The length of the item that
2503 expands into the list of hosts is now unlimited.
2505 47. The lsearch lookup could not handle data where the length of text line was
2506 more than 4095 characters. Such lines were truncated, leading to shortened
2507 data being returned. It should now handle lines of any length.
2509 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2510 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2513 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2514 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2515 no longer happens in comment lines. A list of replacements is now placed
2516 at the head of all of the source files, except those whose only change is
2517 to replace PERL_COMMAND in the very first #! line.
2519 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2520 messages on the queue, with a bottom-up merge sort, using code contributed
2521 by Michael Haardt. This should make operations like -bp somewhat faster on
2522 large queues. It won't affect queue runners, except when queue_run_in_order
2525 51. Installed eximstats 1.31 in the distribution.
2527 52. Added support for SRV lookups to the dnslookup router.
2529 53. If an ACL referred to $message_body or $message_body_end, the value was not
2530 reset for any messages that followed in the same SMTP session.
2532 54. The store-handling optimization for building very long strings was not
2533 differentiating between the different store pools. I don't think this
2534 actually made any difference in practice, but I've tidied it.
2536 55. While running the routers to verify a sender address, $sender_address
2537 was still set to the sender address. This is wrong, because when routing to
2538 send a bounce to the sender, it would be empty. Therefore, I have changed
2539 it so that, while verifying a sender address, $sender_address is set to <>.
2540 (There is no change to what happens when verifying a recipient address.)
2542 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2543 target A or AAAA records (if not already returned) without resetting the
2544 qualify_single or search_parents options of the DNS resolver. These are
2545 inappropriate in this case because the targets of MX and SRV records must
2546 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2547 target that, when qualified, matched something in the local domain. These
2548 two options are now turned off when doing these lookups.
2550 57. It seems that at least some releases of Reiserfs (which does not have the
2551 concept of a fixed number of inodes) returns zero and not -1 for the
2552 number of available inodes. This interacted badly with check_spool_inodes,
2553 which assumed that -1 was the "no such thing" setting. What I have done is
2554 to check that the total number of inodes is greater than zero before doing
2555 the test of how many are available.
2557 58. When a "warn" ACL statement has a log_message modifier, the message is
2558 remembered, and not repeated. This is to avoid a lot of repetition when a
2559 message has many recipients that cause the same warning to be written.
2560 Howewer, Exim was preserving the list of already written lines for an
2561 entire SMTP session, which doesn't seem right. The memory is now reset if a
2562 new message is started.
2564 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2565 debugging output unless log_rewrite was also set.
2567 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2568 of (int)(handle) when we know that handle contains (void *)(-1).
2570 61. The Exim daemon panic-logs an error return when it closes the incoming
2571 connection. However "connection reset by peer" seems to be common, and
2572 isn't really an error worthy of noting specially, so that particular error
2575 62. When Exim is trying to find all the local interfaces, it used to panic and
2576 die if the ioctl to get the interface flags failed. However, it seems that
2577 on at least one OS (Solaris 9) it is possible to have an interface that is
2578 included in the list of interfaces, but for which you get a failure error
2579 for this call. This happens when the interface is not "plumbed" into a
2580 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2581 failure of the "get flags" call assumes that the interface is down.
2583 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2584 makes life easier for people who are doing arithmetic on fields extracted
2585 from dates, where you often get leading zeros that should not be
2586 interpreted as octal.
2588 64. Added qualify_domain to the redirect router, to override the global
2591 65. If a pathologically long header line contained very many addresses (the
2592 report of this problem mentioned 10 000) and each of them was rewritten,
2593 Exim could use up a very large amount of memory. (It kept on making new
2594 copies of the header line as it rewrote, and never released the old ones.)
2595 At the expense of a bit more processing, the header rewriting function has
2596 been changed so that it no longer eats memory in this way.
2598 66. The generation of the Received: header has been moved from the time that a
2599 message starts to be received, to the time that it finishes. The timestamp
2600 in the Received: header should now be very close to that of the <= log
2601 line. There are two side-effects of this change:
2603 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2604 logged header lines no longer include the local Received: line, because
2605 it has not yet been created. The same applies to a copy of the message
2606 that is returned to a non-SMTP sender when a message is rejected.
2608 (b) When a filter file is tested using -bf, no additional Received: header
2609 is added to the test message. After some thought, I decided that this
2612 This change does not affect the value of $received_for. It is still set
2613 after address rewriting, but before local_scan() is called.
2615 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2617 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2618 gave an unhelpful panic error message, and a defer error. I have managed to
2619 change this behaviour so that it now rejects any supplied certificate,
2620 which seems right, as the list of acceptable certificates is empty.
2622 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2623 gave an unhelpful defer error. I have not managed to make this reject any
2624 supplied certificates, but the error message it gives is "no certificate
2625 supplied", which is not helpful.
2627 70. exigrep's output now also includes lines that are not associated with any
2628 message, but which match the given pattern. Implemented by a patch from
2629 Martin Sluka, which also tidied up the Perl a bit.
2631 71. Recipient callout verification, like sender verification, was using <> in
2632 the MAIL FROM command. This isn't really the right thing, since the actual
2633 sender may affect whether the remote host accepts the recipient or not. I
2634 have changed it to use the actual sender in the callout; this means that
2635 the cache record is now keyed on a recipient/sender pair, not just the
2636 recipient address. There doesn't seem to be a real danger of callout loops,
2637 since a callout by the remote host to check the sender would use <>.
2638 [SEE ABOVE: changed after hitting problems.]
2640 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2641 temporary errors. However, in the case of such a code being given after
2642 the end of a data transmission (i.e. after ".") Exim was failing to write
2643 a retry record for the message. (Yes, there was some broken host that was
2644 actually sending 8xx at this point.)
2646 73. An unknown lookup type in a host list could cause Exim to panic-die when
2647 the list was checked. (An example that provoked this was putting <; in the
2648 middle of a list instead of at the start.) If this happened during a DATA
2649 ACL check, a -D file could be left lying around. This kind of configuration
2650 error no longer causes Exim to die; instead it causes a defer errror. The
2651 incident is still logged to the main and panic logs.
2653 74. Buglet left over from Exim 3 conversion. The message "too many messages
2654 in one connection" was written to the rejectlog but not the mainlog, except
2655 when address rewriting (yes!) was being logged.
2657 75. Added write_rejectlog option.
2659 76. When a system filter was run not as root (that is, when system_filter_user
2660 was set), the values of the $n variables were not being returned to the
2661 main process; thus, they were not subsequently available in the $sn
2664 77. Added +return_path_on_delivery log selector.
2666 78. A connection timeout was being treated differently from recipients deferred
2667 when testing hosts_max_try with a message that was older than the host's
2668 retry timeout. (The host should not be counted, thus allowing all hosts to
2669 be tried at least once before bouncing.) This may have been the cause of an
2670 occasionally reported bug whereby a message would remain on the queue
2671 longer than the retry timeout, but would be bounced if a delivery was
2672 forced. I say "may" because I never totally pinned down the problem;
2673 setting up timeout/retry tests is difficult. See also the next item.
2675 79. The ultimate address timeout was not being applied to errors that involved
2676 a combination of host plus message (for example, a timeout on a MAIL
2677 command). When an address resolved to a number of possible hosts, and they
2678 were not all tried for each delivery (e.g. because of hosts_max_try), a
2679 message could remain on the queue longer than the retry timeout.
2681 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2684 81. Fixed an obscure SMTP outgoing bug which required at least the following
2685 conditions: (a) there was another message waiting for the same server;
2686 (b) the server returned 5xx to all RCPT commands in the first message so
2687 that the message was not completed; (c) the server dropped the connection
2688 or gave a negative response to the RSET that Exim sends to abort the
2689 transaction. The observed case was a dropped connection after DATA that had
2690 been sent in pipelining mode. That is, the server had advertised PIPELINING
2691 but was not implementing it correctly. The effect of the bug was incorrect
2692 behaviour, such as trying another host, and this could lead to a crash.
2698 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2699 and daemon.c were passed as pointers to ints; they should have been
2700 pointers to socklen_t variables (which are typically unsigned ints).
2702 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2705 3. Fixed a really odd bug that affected only the testing scheme; patching a
2706 certain fixed string in the binary changed the value of another string that
2707 happened to be identical to the end of the original first string.
2709 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2710 name", it returns that address as the IP address. On some operating
2711 systems (e.g. Solaris), it also passes back the IP address string as the
2712 "host name". However, on others (e.g. Linux), it passes back an empty
2713 string. Exim wasn't checking for this, and was changing the host name to an
2714 empty string, assuming it had been canonicized.
2716 5. Although rare, it is permitted to have more than one PTR record for a given
2717 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2718 all the names associated with an address, because they do in Solaris.
2719 However, it seems that they do not in Linux for data that comes from the
2720 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2721 I found this out when I moved to a new Linux workstation and tried to run
2722 the Exim test suite.
2724 To get round this problem I have changed the code so that it now does its
2725 own call to the DNS to look up PTR records when searching for a host name.
2726 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2727 addresses that are only in /etc/hosts are still found.
2729 This behaviour is, however, controlled by an option called host_lookup_
2730 order, which defaults to "bydns:byaddr". If people want to use the other
2731 order, or indeed, just use one or the other means of lookup, they can
2732 specify it in this variable.
2734 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2735 some operating systems, this comes back from gethostbyaddr() as an empty
2736 string, and this is what Exim used to test for. However, it seems that in
2737 other systems, "." is yielded. Exim now tests for this case too.
2739 7. The values of check_spool_space and check_log_space are now held internally
2740 as a number of kilobytes instead of an absolute number of bytes. If a
2741 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2742 kilobyte. This means that much larger values can be stored.
2744 8. Exim monitor: an attempt to get the action menu when not actually pointing
2745 at a message produces an empty menu entitled "No message selected". This
2746 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2747 no entries in it ("Shell widget menu has zero width and/or height"). So I
2748 have added a single, blank menu entry in this case.
2750 9. Added ${quote_local_part.
2752 10. MIME decoding is now applied to the contents of Subject: header lines when
2755 11. Now that a reference to $sender_host_address automatically causes a reverse
2756 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2757 host lookup before query-style lookups in lists that might use this
2758 variable. This has therefore been abolished, and the "net-" prefix is no
2759 longer necessary for query-style lookups.
2761 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2762 have been a typo for LFLAGS, so it has been changed.
2764 13. The install script calls Exim with "-C /dev/null" in order to find the
2765 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2766 to be output. Howeve, since Exim outputs its version number before the
2767 error, it didn't break the script. It just looked ugly. I fixed this by
2768 always allowing "-C /dev/null" if the caller is root.
2770 14. Ignore overlarge ACL variable number when reading spool file - insurance
2771 against a later release with more variables having written the file.
2773 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2774 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2775 this, as well as the form without the "IPv6" on the front.
2777 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2778 OS/Makefile-Darwin file.
2780 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2781 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2783 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2784 variable. However, some OS use ints for inodes. I've added cast to long int
2785 to get rid of the compiler warning.
2787 19. I had forgotten to lock out "/../" in configuration file names when
2788 ALT_CONFIG_PREFIX was set.
2790 20. Routers used for verification do not need to specify transports. However,
2791 if such a router generated a host list, and callout was configured, Exim
2792 crashed, because it could not find a port number from the (non-existent)
2793 transport. It now assumes port 25 in this circumstance.
2795 21. Added the -t option to exigrep.
2797 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2798 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2799 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2800 them is set, LOOKUP_LSEARCH is forced.
2802 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2804 24. Added sender and host information to the "rejected by local_scan()" log
2805 line; previously there was no indication of these.
2807 25. Added .include_if_exists.
2809 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2810 Linux. It turns out that not all file systems support this. Apparently some
2811 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2812 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2813 not supported on the file descriptor, is now ignored when Exim is trying to
2814 sync a directory. This applies only to Linux.
2816 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2818 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2819 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2822 29. Added support for maildirsize files from supplied patch (modified a bit).
2824 30. The use of :fail: followed by an empty string could lead Exim to respond to
2825 sender verification failures with (e.g.):
2827 550 Verification failed for <xxx>
2828 550 Sender verify failed
2830 where the first response line was missing the '-' that indicates it is not
2831 the final line of the response.
2833 31. The loop for finding the name of the user that called Exim had a hardwired
2834 limit of 10; it now uses the value of finduser_retries, which is used for
2835 all other user lookups.
2837 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2840 33. Exim was neglecting to zero errno before one call of strtol() when
2841 expanding a string and expecting an integer value. On some systems this
2842 resulted in spurious "integer overflow" errors. Also, it was casting the
2843 result into an int without checking.
2845 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2846 did not work. The code looks as if it has *never* worked, though it appears
2847 to have been documented since at least releast 1.62. I have made it work.
2849 35. The "timeout_DNS" error in retry rules, also documented since at least
2850 1.62, also never worked. As it isn't clear exactly what this means, and
2851 clearly it isn't a major issue, I have abolished the feature by treating it
2852 as "timeout", and writing a warning to the main and panic logs.
2854 36. The display of retry rules for -brt wasn't always showing the error code
2857 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2858 timeout_connect_A, timeout_connect_MX.
2860 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2861 to the empty sender.
2863 39. The daemon was not analysing the content of -oX till after it had closed
2864 stderr and disconnected from the controlling terminal. This meant that any
2865 syntax errors were only noted on the panic log, and the return code from
2866 the command was 0. By re-arranging the code a little, I've made the
2867 decoding happen first, so such errors now appear on stderr, and the return
2868 code is 1. However, the actual setting up of the sockets still happens in
2869 the disconnected process, so errors there are still only recorded on the
2872 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2873 connections (as happens on some IP stacks) was logged at start up time as
2874 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2875 it from "IPv6 and IPv4", which means that two separate sockets are being
2878 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2879 says whether AF_INET or AF_INET6 was passed as an argument.
2881 42. Exiwhat output was messed up when time zones were included in log
2884 43. Exiwhat now gives more information about the daemon's listening ports,
2885 and whether -tls-on-connect was used.
2887 44. The "port" option of the smtp transport is now expanded.
2889 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2890 silently ignored. Now an error message is written to the main and panic
2893 46. There's a new ACL modifier called "logwrite" which writes to a log file
2894 as soon as it is encountered.
2896 47. Added $local_user_uid and $local_user_gid at routing time.
2898 48. Exim crashed when trying to verify a sender address that was being
2901 49. Exim was recognizing only a space character after ".include". It now also
2902 recognizes a tab character.
2904 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2905 extracting the relevant information from the specification. The man page no
2906 longer contains scrambled data for the -d option, and I've added a section
2907 at the front about calling Exim under different names.
2909 51. Added "extra_headers" argument to the "mail" command in filter files.
2911 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2914 53. Installed eximstats 1.29.
2916 54. Added transport_filter_timeout as a generic transport option.
2918 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2919 Cc: header lines. This was required by RFC 822, but it not required by RFC
2922 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2923 incoming messages that did not have them. Now it does so only if the
2924 message originates locally, that is, if there is no associated remote host
2925 address. When Resent- header lines are present, this applies to the Resent-
2926 lines rather than the non-Resent- lines.
2928 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2929 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2931 58. Messages for configuration errors now include the name of the main
2932 configuration file - useful now that there may be more than one file in a
2933 list (.included file names were always shown).
2935 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2936 for those rare installations that do not start the daemon as root or run it
2937 setuid root. I've cut out the call to initgroups() if the daemon is not
2940 60. The Exim user and group can now be bound into the binary as text strings
2941 that are looked up at the start of Exim's processing.
2943 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2945 62. Added $mailstore_basename variable.
2947 63. Installed patch to sieve.c from Michael Haardt.
2949 64. When Exim failed to open the panic log after failing to open the main log,
2950 the original message it was trying to log was written to stderr and debug
2951 output, but if they were not available (the usual case in production), it
2952 was lost. Now it is written to syslog before the two lines that record the
2953 failures to open the logs.
2955 65. Users' Exim filters run in subprocesses under the user's uid. It is
2956 possible for a "deliver" command or an alias in a "personal" command to
2957 provoke an address rewrite. If logging of address rewriting is configured,
2958 this fails because the process is not running as root or exim. There may be
2959 a better way of dealing with this, but for the moment (because 4.30 needs
2960 to be released), I have disabled address rewrite logging when running a
2961 filter in a non-root, non-exim process.
2967 1. The buildconfig auxiliary program wasn't quoting the value set for
2968 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2969 not defined. This bug was masked in 4.22 by the effect that was fixed in
2972 2. Some messages that were rejected after a message id was allocated were
2973 shown as "incomplete" by exigrep. It no longer does this for messages that
2974 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2976 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2977 have allow_domain_literals set, the ID did not get logged in the <= line.
2978 Domain literals are now always recognized in Message-ID: header lines.
2980 4. The first argument for a ${extract expansion item is the key name or field
2981 number. Leading and trailing spaces in this item were not being ignored,
2982 causing some misleading effects.
2984 5. When deliver_drop_privilege was set, single queue runner processes started
2985 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2986 same command in the process it spins off) were not dropping privilege.
2988 6. When the daemon running as "exim" started a queue runner, it always
2989 re-executed Exim in the spun-off process. This is a waste of effort when
2990 deliver_drop_privilege is set. The new process now just calls the
2991 queue-runner function directly.
2997 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3000 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3001 ignored. Though the use of -oP was forcing the writing of a pid file, it
3002 was always written to the default place.
3004 3. If the message "no IP address found for host xxxx" is generated during
3005 incoming verification, it is now followed by identification of the incoming
3006 connection (so you can more easily find what provoked it).
3008 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3010 5. Added some features to "harden" Exim a bit more against certain attacks:
3012 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3013 be put in Local/Makefile. This is like the never_users runtime option,
3014 but it cannot be overridden. The default setting is "root".
3016 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3017 prefix string with which any file named in a -C command line option
3020 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3021 is retained for -C and -D only if the caller of Exim is root. Without
3022 it, the exim user may also use -C and -D and retain privilege.
3024 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3025 command line option is disabled.
3027 6. Macro names set by the -D option must start with an upper case letter, just
3028 like macro names defined in the configuration file.
3030 7. Added "dereference=" facility to LDAP.
3032 8. Two instances of the typo "uknown" in the source files are fixed.
3034 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3035 the Configure-Makefile script screwed up while processing it.
3037 10. Incorporated PCRE 4.4.
3039 11. The SMTP synchronization check was not operating right at the start of an
3040 SMTP session. For example, it could not catch a HELO sent before the client
3041 waited for the greeting. There is now a check for outstanding input at the
3042 point when the greeting is written. Because of the duplex, asynchronous
3043 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3044 way, but not yet received, when the check is performed.
3046 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3047 on TCP/IP sockets, because this apparently causes some broken clients to
3050 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3051 unchanged) from the Cygwin maintainer.
3053 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3054 was supported instead of testing for mbx. Effectively a typo.
3056 15. The spa authenticator server code was not checking that the input it
3057 received was valid base64.
3059 16. The debug output line for the "set" modifier in ACLs was not showing the
3060 name of the variable that was being set.
3062 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3064 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3065 lookup on its own. Something else was needed to trigger the lookup. For
3066 example, a match in host_lookup or the need for a host name in a host list.
3067 Now, if $sender_host_name is referenced and the host name has not yet been
3068 looked up, a lookup is performed. If the lookup fails, the variable remains
3069 empty, and $host_lookup_failed is set to "1".
3071 19. Added "eqi" as a case-independent comparison operator.
3073 20. The saslauthd authentication condition could segfault if neither service
3074 nor realm was specified.
3076 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3077 error message that was logged was misleading, and incoming SMTP
3078 connections were dropped. The message is now more accurate, and temporary
3079 errors are given to SMTP connections.
3081 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3082 (except RSET) with a 421 error, until QUIT is received. However, it was
3083 failing to send a response to QUIT.
3085 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3086 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3087 ACL is now run after the helo_try_verify_hosts code.
3089 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3090 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3091 case-independent, so other case variants are also recognized). Apparently
3092 some systems use these upper case variants.
3094 25. If more than two messages were waiting for the same host, and a transport
3095 filter was specified for the transport, Exim sent two messages over the
3096 same TCP/IP connection, and then failed with "socket operation on non-
3097 socket" when it tried to send the third.
3099 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3101 27. The extern definition of crypt16() in expand.c was not being excluded when
3102 the OS had its own crypt16() function.
3104 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3105 as a preferred synonym for return_size_limit, both as an option and as an
3108 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3110 30. Changed the default configuration ACL to relax the local part checking rule
3111 for addresses that are not in any local domains. For these addresses,
3112 slashes and pipe symbols are allowed within local parts, but the sequence
3113 /../ is explicitly forbidden.
3115 31. SPA server authentication was not clearing the challenge buffer before
3118 32. log_message in a "warn" ACL statement was writing to the reject log as
3119 well as to the main log, which contradicts the documentation and doesn't
3120 seem right (because no rejection is happening). So I have stopped it.
3122 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3123 However, I am unable to do any testing of this.
3125 34. Fixed an infelicity in the appendfile transport. When checking directories
3126 for a mailbox, to see if any needed to be created, it was accidentally
3127 using path names with one or more superfluous leading slashes; tracing
3128 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3130 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3131 contents are added to the log line that is written for every discarded
3132 recipient. (Previously a log_message setting was ignored.)
3134 36. The ${quote: operator now quotes the string if it is empty.
3136 37. The install script runs exim in order to find its version number. If for
3137 some reason other than non-existence or emptiness, which it checks, it
3138 could not run './exim', it was installing it with an empty version number,
3139 i.e. as "exim-". This error state is now caught, and the installation is
3142 38. An argument was missing from the function that creates an error message
3143 when Exim fails to connect to the socket for saslauthd authentication.
3144 This could cause Exim to crash, or give a corrupted message.
3146 39. Added isip, isip4, and isip6 to ${if conditions.
3148 40. The ACL variables $acl_xx are now saved with the message, and can be
3149 accessed later in routers, transports, and filters.
3151 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3152 strings in the file are not string-expanded.
3154 42. If a MAIL command specified a SIZE value that was too large to fit into an
3155 int variable, the check against message_size_limit failed. Such values are
3156 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3157 day this will have to be increased, but I don't think I want to be around
3158 when emails are that large.
3165 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3166 iconv() is not standard in FreeBSD.
3168 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3169 IPv6 enabled. The observed symptom was a segmentation fault on return from
3170 the function os_common_find_running_interfaces() in src/os.c.
3172 3. In the check_special_case() function in daemon.c I had used "errno" as an
3173 argument name, which causes warnings on some systems. This was basically a
3174 typo, since it was named "eno" in the comments!
3176 4. The code that waits for the clock to tick (at a resolution of some fraction
3177 of a second) so as to ensure message-id uniqueness was always waiting for
3178 at least one whole tick, when it could have waited for less. [This is
3179 almost certainly not relevant at current processor speeds, where it is
3180 unlikely to ever wait at all. But we try to future-proof.]
3182 5. The function that sleeps for a time interval that includes fractions of a
3183 second contained a race. It did not block SIGALRM between setting the
3184 timer, and suspending (a couple of lines later). If the interval was short
3185 and the sigsuspend() was delayed until after it had expired, the suspension
3186 never ended. On busy systems this could lead to processes getting stuck for
3189 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3190 process, before it forks any delivery processes. The open lookup caching
3191 mechanism meant that the open file or database connection was passed into
3192 the delivery process. The problem was that delivery processes always tidy
3193 up cached lookup data. This could cause a problem for the next delivery
3194 process started by the queue runner, because the external queue runner
3195 process does not know about the closure. So the next delivery process
3196 still has data in the lookup cache. In the case of a file lookup, there was
3197 no problem because closing a file descriptor in a subprocess doesn't affect
3198 the parent. However, if the lookup was caching a connection to a database,
3199 the connection was closed, and the second delivery process was likely to
3200 see errors such as "PGSQL: query failed: server closed the connection
3201 unexpectedly". The problem has been fixed by closing all cached lookups
3202 in a queue runner before running a delivery process.
3204 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3205 seem to have the "const" qualifier which it has on other OS. I've
3208 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3209 *of the same type* (client or server) with the same public name that an
3210 error should be diagnosed.
3212 9. When Exim looked up a host name for an IP address, but failed to find the
3213 original IP address when looking up the host name (a safety check), it
3214 output the message "<ip address> does not match any IP for NULL", which was
3215 confusing, to say the least. The bug was that the host name should have
3216 appeared instead of "NULL".
3218 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3219 user that is built into the binary, and the -C or -D options is used, root
3220 privilege is dropped before the configuration file is read. In addition,
3221 logging is switched to stderr instead of the normal log files. If the
3222 configuration then re-defines the Exim user, the unprivileged environment
3223 is probably not what is expected, so Exim logs a panic warning message (but
3226 However, if deliver_drop_privilege is set, the unprivileged state may well
3227 be exactly what is intended, so the warning has been cut out in that case,
3228 and Exim is allowed to try to write to its normal log files.
3234 1. smtp_return_error_details was not giving details for temporary sender
3235 or receiver verification errors.
3237 2. Diagnose a configuration error if two authenticators have the same public
3240 3. Exim used not to create the message log file for a message until the first
3241 delivery attempt. This could be confusing when incoming messages were held
3242 for policy or load reasons. The message log file is now created at the time
3243 the message is received, and an initial "Received" line is written to it.
3245 4. The automatically generated man page for command line options had a minor
3246 bug that caused no ill effects; however, a more serious problem was that
3247 the procedure for building the man page automatically didn't always
3248 operate. Consequently, release 4.20 contains an out-of-date version. This
3249 shouldn't happen again.
3251 5. When building Exim with embedded Perl support, the script that builds the
3252 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3253 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3255 6. The freeze_tell option was not being used for messages that were frozen on
3256 arrival, either by an ACL or by local_scan().
3258 7. Added the smtp_incomplete_transaction log selector.
3260 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3261 was accepting AUTH without a new EHLO.
3263 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3264 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3267 10. Logging of TCP/IP connections (when configured) now happens in the main
3268 daemon process instead of the child process, so that the TCP/IP connection
3269 count is more accurate (but it can never be perfect).
3271 11. The use of "drop" in a nested ACL was not being handled correctly in the
3272 outer ACL. Now, if condition failure induced by the nested "drop" causes
3273 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3274 or "require"), the connection is dropped.
3276 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3277 that yield "discard" can now be used with an "accept" or a "discard" verb,
3278 but an error is generated for any others (because I can't see a useful way
3279 to define what should happen).
3281 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3282 are now processed in the same way as lines in the Exim configuration file.
3283 In particular, continuation lines are supported.
3285 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3287 15. Added -ti meaning -t -i.
3289 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3290 domains, and warn by logging if others are found.
3292 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3293 returned from GIFCONF when Exim is trying to find the list of interfaces on
3294 a host. The code in os.c has been modified to copy each ifreq to an aligned
3295 structure in all cases.
3297 Also, in some cases, the returned ifreq's were being copied to a 'struct
3298 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3299 means the last couple of bytes of an IPv6 address could be chopped if the
3300 ifreq contained only a normal sockaddr (14 bytes storage).
3302 18. Named domain lists were not supported in the hosts_treat_as_local option.
3303 An entry such as +xxxx was not recognized, and was treated as a literal
3306 19. Ensure that header lines added by a DATA ACL are included in the reject log
3307 if the ACL subsequently rejects the message.
3309 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3310 MD5 (which is deprecated).
3312 21. When testing a filter file using -bf, Exim was writing a message when it
3313 took the sender from a "From " line in the message, but it was not doing so
3314 when it took $return_path from a Return-Path: header line. It now does.
3316 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3317 with a valid header line field name (a series of printing characters
3318 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3320 23. Changed "disc" in the source to "disk" to conform to the documentation and
3321 the book and for uniformity.
3323 24. Ignore Sendmail's -Ooption=value command line item.
3325 25. When execve() failed while trying to run a command in a pipe transport,
3326 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3327 could be confused with a return value of 69 from the command itself. This
3328 has been changed to 127, the value the shell returns if it is asked to run
3329 a non-existent command. The wording for the related log line suggests a
3330 non-existent command as the problem.
3332 26. If received_header_text expands to an empty string, do not add a Received:
3333 header line to the message. (Well, it adds a token one on the spool, but
3334 marks it "old" so that it doesn't get used or transmitted.)
3336 27. Installed eximstats 1.28 (addition of -nt option).
3338 28. There was no check for failure on the call to getsockname() in the daemon
3339 code. This can fail if there is a shortage of resources on the system, with
3340 ENOMEM, for example. A temporary error is now given on failure.
3342 29. Contrary to the C standard, it seems that in some environments, the
3343 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3344 program. Exim now does this explicitly; it affects the formatting of
3345 timestamps using strftime().
3347 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3348 complaints. I've now initialized all the variables, to avoid this.
3350 32. Header lines added by a system filter were not being "seen" during
3351 transport-time rewrites.
3353 33. The info_callback() function passed to OpenSSL is set up with type void
3354 (*)(SSL *, int, int), as described somewhere. However, when calling the
3355 function (actually a macro) that sets it up, the type void(*)() is
3356 expected. I've put in a cast to prevent warnings from picky compilers.
3358 34. If a DNS black list lookup found a CNAME record, but there were no A
3359 records associated with the domain it pointed at, Exim crashed.
3361 35. If a DNS black list lookup returned more than one A record, Exim ignored
3362 all but the first. It now scans all returned addresses if a particular IP
3363 value is being sought. In this situation, the contents of the
3364 $dnslist_value variable are a list of all the addresses, separated by a
3367 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3368 to accept a host name and all its aliases if the forward lookup for any of
3369 them yielded the IP address of the incoming connection. Now it accepts only
3370 those names whose forward lookup yields the correct IP address. Any other
3371 names are discarded. This closes a loophole whereby a rogue DNS
3372 administrator could create reverse DNS records to break through a
3373 wildcarded host restriction in an ACL.
3375 37. If a user filter or a system filter that ran in a subprocess used any of
3376 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3377 the wrong values were passed to the pipe command ($thisaddress had the
3378 value of $0, $0 had the value of $1, etc). This bug was introduced by
3379 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3381 38. Improved the line breaking for long SMTP error messages from ACLs.
3382 Previously, if there was no break point between 40 and 75 characters, Exim
3383 left the rest of the message alone. Two changes have been made: (a) I've
3384 reduced the minimum length to 35 characters; (b) if it can't find a break
3385 point between 35 and 75 characters, it looks ahead and uses the first one
3386 that it finds. This may give the occasional overlong line, but at least the
3387 remaining text gets split now.
3389 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3390 file descriptors might be low, and that setting 1000 would always raise it.
3391 It turns out that in some environments, the limit is already over 1000 and
3392 that lowering it causes trouble. So now Exim takes care not to decrease it.
3394 40. When delivering a message, the value of $return_path is set to $sender_
3395 address at the start of routing (routers may change the value). By an
3396 oversight, this default was not being set up when an address was tested by
3397 -bt or -bv, which affected the outcome if any router or filter referred to
3400 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3401 log only when "message" or "log_message" are set. However, if one of the
3402 conditions was an address verification, or a call to a nested ACL, the
3403 messages generated by the underlying test were being passed through. This
3404 no longer happens. The underlying message is available in $acl_verify_
3405 message for both "message" and "log_message" expansions, so it can be
3406 passed through if needed.
3408 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3409 new expansion $bh_ to give the encoded byte string without charset
3410 translation. Translation happens only if iconv() is available; HAVE_ICONV
3411 indicates this at build time. HEADERS_CHARSET gives the charset to
3412 translate to; headers_charset can change it in the configuration, and
3413 "headers charset" can change it in an individual filter file.
3415 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3416 that creates RFC 2047 encoded "words" labels them as that charset instead
3417 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3418 expansion operator; (ii) when Exim creates a From: line for a local
3419 message; (iii) when a header line is rewritten to include a "phrase" part.
3421 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3422 buggy, causing it to skip the first lines of messages whose message ID
3423 ended in 'D'. This would not have bitten before Exim release 4.14, because
3424 message IDs were unlikely to end in 'D' before then. The effect was to have
3425 incorrect size information for certain domains.
3427 45. #include "config.h" was missing at the start of the crypt16.c module. This
3428 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3431 46. If there was a timeout during a "random" callout check, Exim treated it as
3432 a failure of the random address, and carried on sending RSET and the real
3433 address. If the delay was just some slowness somewhere, the response to the
3434 original RCPT would be taken as a response to RSET and so on, causing
3435 mayhem of various kinds.
3437 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3438 when I implemented it. It didn't allow for the fact that some option values
3439 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3440 the right test for positive values.
3442 48. Domain names in DNS records are case-independent. Exim always looks them up
3443 in lower case. Some resolvers return domain names in exactly the case they
3444 appear in the zone file, that is, they may contain uppercase letters. Not
3445 all resolvers do this - some return always lower case. Exim was treating a
3446 change of case by a resolver as a change of domain, similar to a widening
3447 of a domain abbreviation. This triggered its re-routing code and so it was
3448 trying to route what was effectively the same domain again. This normally
3449 caused routing to fail (because the router wouldn't handle the domain
3450 twice). Now Exim checks for this case specially, and just changes the
3451 casing of the domain that it ultimately uses when it transmits the message
3454 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3457 50. If a filter generated a file delivery with a non-absolute name (possible if
3458 no home directory exists for the router), the forbid_file option was not
3461 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3462 the existing equality matching.
3464 52. Exim was using ints instead of ino_t variables in some places where it was
3465 dealing with inode numbers.
3467 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3468 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3469 called TMPDIR, and if it finds it is different, it changes its value.
3471 54. The smtp_printf() function is now made available to local_scan() so
3472 additional output lines can be written before returning. There is also an
3473 smtp_fflush() function to enable the detection of a dropped connection.
3474 The variables smtp_input and smtp_batched_input are exported to
3477 55. Changed the default runtime configuration: the message "Unknown user"
3478 has been removed from the ACL, and instead placed on the localuser router,
3479 using the cannot_route_message feature. This means that any verification
3480 failures that generate their own messages won't get overridden. Similarly,
3481 the "Unrouteable address" message that was in the ACL for unverifiable
3482 relay addresses has also been removed.
3484 56. Added hosts_avoid_esmtp to the smtp transport.
3486 57. The exicyclog script was not checking for the esoteric option
3487 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3488 will work only if exicyclog is run under the appropriate euid.
3490 58. Following a discussion on the list, the rules by which Exim recognises line
3491 endings on incoming messages have been changed. The -dropcr and drop_cr
3492 options are now no-ops, retained only for backwards compatibility. The
3493 following line terminators are recognized: LF CRLF CR. However, special
3494 processing applies to CR:
3496 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3497 nor a local message in the state where . is a terminator.
3499 (ii) If a bare CR is encountered in a header line, an extra space is added
3500 after the line terminator so as not to end the header. The reasoning
3501 behind this is that bare CRs in header lines are most likely either
3502 to be mistakes, or people trying to play silly games.
3504 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3505 was being incorrectly given as 18 bytes larger than it should have been.
3506 This is a VOB (very old bug).
3508 60. This may never have affected anything current, but just in case it has:
3509 When the local host is found other than at the start of a list of hosts,
3510 the local host, those with the same MX, and any that follow, are discarded.
3511 When the list in question was part of a longer list of hosts, the following
3512 hosts (not currently being processed) were also being discarded. This no
3513 longer happens. I'm not sure if this situation could ever has previously
3516 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3519 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3520 header. This is something that is recommended in a new Internet Draft, and
3521 is something that is documented as being done by Sendmail. There are two
3522 possible values. For messages generated by the autoreply transport, Exim
3525 Auto-Submitted: auto-replied
3527 whereas for all other generated messages (e.g. bounces) it adds
3529 Auto-Submitted: auto-generated
3531 63. The "personal" condition in filters now includes a test for the
3532 Auto-Submitted: header. If it contains the string "auto-" the message it
3533 not considered personal.
3535 64. Added rcpt_include_affixes as a generic transport option.
3537 65. Added queue_only_override (default true).
3539 66. Added the syslog_duplication option.
3541 67. If what should have been the first header line of a message consisted of
3542 a space followed by a colon, Exim was mis-interpreting it as a header line.
3543 It isn't of course - it is syntactically invalid and should therefore be
3544 treated as the start of the message body. The misbehaviour could have
3545 caused a number of strange effects, including loss of data in subsequent
3546 header lines, and spool format errors.
3548 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3549 client host had authenticated. This control can now be exercised by an ACL
3550 for more flexibility.
3552 69. By default, callouts do not happen when testing with -bh. There is now a
3553 variant, -bhc, which does actually run the callout code, including
3554 consulting and updating the callout cache.
3556 70. Added support for saslauthd authentication, courtesy of Alexander
3559 71. If statvfs() failed on the spool or log directories while checking their
3560 size for availability, Exim confusingly gave the error "space shortage".
3561 Furthermore, in debugging mode it crashed with a floating point exception.
3562 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3563 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3564 serious problem, Exim now writes to the main and panic logs when this
3565 happens, with details of the failure. It then refuses to accept the
3566 incoming message, giving the message "spool directory problem" or "log
3567 directory problem" with a 421 code for SMTP messages.
3569 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3570 0, 1, and 2 exist, because some OS complain for execs without them (see
3571 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3572 descriptors. However, the code omitted to check that the open succeeded,
3573 causing mysterious errors if for some reason the permissions on /dev/null
3574 got screwed. Now Exim writes a message to the main and panic logs, and
3575 bombs out if it can't open /dev/null.
3577 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3578 interact so that it is all more flexible. It is supposed to remain
3579 backwards compatible. Also added extra_local_interfaces.
3581 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3582 to bomb out with an assertion failure - to the client this appears as a
3583 connection drop. This problem occurs in the part of the code that was taken
3584 from the Samba project. Fortunately, the assertion is in a very simple
3585 function, so I have fixed this by reproducing the function inline in the
3586 one place where it is called, and arranging for authentication to fail
3587 instead of killing the process with assert().
3589 75. The SPA client code was not working when the server requested OEM rather
3590 than Unicode encoding.
3592 76. Added code to make require_files with a specific uid setting more usable in
3593 the case where statting the file as root fails - usually a non-root-mounted
3594 NFS file system. When this happens and the failure is EACCES, Exim now
3595 forks a subprocess and does the per-uid checking as the relevant uid.
3597 77. Added process_log_path.
3599 78. If log_file_path was not explicitly set, a setting of check_log_space or
3600 check_log_inodes was ignored.
3602 79. If a space check for the spool or log partitions fails, the incident is now
3603 logged. Of course, in the latter case the data may get lost...
3605 80. Added the %p formatting code to string_format() so that it can be used to
3606 print addresses in debug_print(). Adjusted all the address printing in the
3607 debugging in store.c to use %p rather than %d.
3609 81. There was a concern that a line of code in smtp_in.c could overflow a
3610 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3611 initially expressed, the concern was not well-founded, because trailing
3612 spaces are removed early. However, if the trailing spaces were followed by
3613 a NULL, they did not get removed, so the overflow was possible. Two fixes
3616 (a) I re-wrote the offending code in a cleaner fashion.
3617 (b) If an incoming SMTP command contains a NULL character, it is rejected
3620 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3621 runs initgroups(), so that if the Exim user is in any additional groups,
3622 they will be used during message reception.
3628 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob