1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.281 2005/12/22 11:47:24 ph10 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
55 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
63 PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
71 PH/01 Two changes to the default runtime configuration:
73 (1) Move the checks for relay_from_hosts and authenticated clients from
74 after to before the (commented out) DNS black list checks.
76 (2) Add control=submission to the relay_from_hosts and authenticated
77 clients checks, on the grounds that messages accepted by these
78 statements are most likely to be submissions.
80 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
82 (1) Generate an error if the third argument for the ${prvs expansion is
85 (2) Treat a missing third argument of ${prvscheck as if it were an empty
88 (3) Reset the variables that are obtained from the first argument of
89 ${prvscheck and used in the second argument before leaving the code,
90 because their memory is reclaimed, so using them afterwards may do
93 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
94 one (it's much easier than Tom thought :-).
96 (5) Because of (4), we can now allow for the use of $prvscheck_result
97 inside the third argument.
99 PH/03 For some reason, the default setting of PATH when running a command from
100 a pipe transport was just "/usr/bin". I have changed it to
103 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
104 anything to be listed in the output from -bV.
106 PH/05 When a filter generated an autoreply, the entire To: header line was
107 quoted in the delivery log line, like this:
109 => >A.N.Other <ano@some.domain> <original@ddress> ...
111 This has been changed so that it extracts the operative address. There
112 may be more than one such address. If so, they are comma-separated, like
115 => >ano@some.domain,ona@other.domain <original@ddress> ...
117 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
118 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
119 not being looked up in the rDNS to get a host name, Exim was showing the
120 IP address twice in Received: lines, even though the IP addresses were
121 identical. For example:
123 Received: from [1.2.3.4] (helo=[1.2.3.4])
125 However, if the real host name was known, it was omitting the HELO data
126 if it matched the actual IP address. This has been tidied up so that it
127 doesn't show the same IP address twice.
129 PH/07 When both +timestamp and +memory debugging was on, the value given by
130 $tod_xxx expansions could be wrong, because the tod_stamp() function was
131 called by the debug printing, thereby overwriting the timestamp buffer.
132 Debugging no longer uses the tod_stamp() function when +timestamp is set.
134 PH/08 When the original message was included in an autoreply transport, it
135 always said "this is a copy of the message, including all the headers",
136 even if body_only or headers_only was set. It now gives an appropriate
139 PH/09 Applied a patch from the Sieve maintainer which:
141 o fixes some comments
142 o adds the (disabled) notify extension core
143 o adds some debug output for the result of if/elsif tests
144 o points to the current vacation draft in the documentation
145 and documents the missing references header update
149 o fixes a bug in processing the envelope test (when testing
150 multiple envelope elements, the last element determinted the
153 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
154 Electronic Mail") by including:
156 Auto-submitted: auto-generated
158 in the messages that it generates (bounce messages and others, such as
159 warnings). In the case of bounce messages for non-SMTP mesages, there was
160 also a typo: it was using "Auto_submitted" (underscore instead of
161 hyphen). Since every message generated by Exim is necessarily in response
162 to another message, thes have all been changed to:
164 Auto-Submitted: auto-replied
166 in accordance with these statements in the RFC:
168 The auto-replied keyword:
170 - SHOULD be used on messages sent in direct response to another
171 message by an automatic process,
173 - MUST NOT be used on manually-generated messages,
175 - MAY be used on Delivery Status Notifications (DSNs) and Message
176 Disposition Notifications (MDNs),
178 - MUST NOT be used on messages generated by automatic or periodic
179 processes, except for messages which are automatic responses to
182 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
183 to the default Received: header definition.
185 PH/12 Added log selector acl_warn_skipped (default on).
187 PH/13 After a successful wildlsearch lookup, discard the values of numeric
188 variables because (a) they are in the wrong storage pool and (b) even if
189 they were copied, it wouldn't work properly because of the caching.
191 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
192 checking when decoding. Apparently there are clients that generate
193 overlong encoded strings. Why am I not surprised?
195 PH/15 If the first argument of "${if match_address" was not empty, but did not
196 contain an "@" character, Exim crashed. Now it writes a panic log message
197 and treats the condition as false.
199 PH/16 In autoreply, treat an empty string for "once" the same as unset.
201 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
202 extension "envelope-auth". The code is finished and in agreement with
203 other implementations, but there is no documentation so far and in fact,
204 nobody wrote the draft yet. This extension is currently #undef'ed, thus
205 not changing the active code.
207 Print executed "if" and "elsif" statements when debugging is used. This
208 helps a great deal to understand what a filter does.
210 Document more things not specified clearly in RFC3028. I had all this
211 sorted out, when out of a sudden new issues came to my mind. Oops."
213 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
216 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
217 canonical form (as documented). However, after a host name lookup from
218 the IP address, check_host() was doing a simple string comparison with
219 addresses acquired from the DNS when checking that the found name did
220 have the original IP as one of its addresses. Since any found IPv6
221 addresses are likely to be in abbreviated form, the comparison could
222 fail. Luckily, there already exists a function for doing the comparison
223 by converting both addresses to binary, so now that is used instead of
226 PH/20 There was another similar case to PH/19, when a complete host name was
227 given in a host list; looking up its IP address could give an abbreviated
228 form, whereas the current host's name might or might not be abbreviated.
229 The same fix has been applied.
235 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
236 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
239 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
240 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
242 PH/03 Typo: missing ".o" in src/pcre/Makefile.
244 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
245 header line, restrict the check to what is listed in RFCs 2369 and 2929.
246 Also, for "Auto-Submitted", treat anything other than "no" as
247 non-personal, in accordance with RFC 3834. (Previously it treated
248 anything starting "auto-" as non-personal.)
250 TF/01 The control=submission/name=... option had a problem with syntax
251 errors if the name included a slash character. The /name= option
252 now slurps the rest of the string, so it can include any characters
253 but it must come last in the list of options (after /sender_retain
256 PH/05 Some modifications to the interface to the fake nameserver for the new
264 TK/01 Added the "success_on_redirect" address verification option. See
265 NewStuff for rationale and an example.
267 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
269 PH/02 Patch to exigrep to allow it to work on syslog lines.
271 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
272 fread() to skip over the body file's header line, because in Cygwin the
273 header line is locked and is inaccessible.
275 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
276 co-exist for some time) to make it clear that it is the Exim ID that is
277 referenced, not the Message-ID: header line.
279 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
280 string_format() function, because snprintf() does not exist on all
283 PH/06 The use of forbid_filter_existstest now also locks out the use of the
284 ${stat: expansion item.
286 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
287 protocol synchronization error", to keep the pedants happy.
289 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
290 well as for IRIX systems, when gcc is being used. See the host.c source
293 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
295 PH/10 Named domain lists were not working if used in a queue_smtp_domains
298 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
299 transport and to the smtp transport in LMTP mode.
301 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
303 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
304 run a filter in a subprocess. This could lead to confusion in subsequent
305 lookups in the parent process. There should also be a search_tidyup() at
306 the end of the subprocess.
308 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
309 only if the host matched helo_try_verify_hosts, which caused the
310 verification to occur when the EHLO/HELO command was issued. The ACL just
311 tested the remembered result. Now, if a previous verification attempt has
312 not happened, "verify = helo" does it there and then.
314 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
316 TK/03 Fix log output including CR from clamd.
318 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
319 exist provoked a memory error which could cause a segfault.
321 PH/15 Installed PCRE 6.2
323 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
325 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
326 of the problem. Specifically, suggested +O2 rather than +O1 for the
329 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
331 PH/20 If a delivery was routed to a non-standard port by means of an SRV
332 record, the port was not correctly logged when the outgoing_port log
333 selector was set (it logged the transort's default port).
335 PH/21 Added support for host-specific ports to manualroute, queryprogram,
336 fallback_hosts, and "hosts" in the smtp transport.
338 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
339 host errors such as "Connection refused".
341 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
342 authentication with radiusclient 0.4.9:
344 - Error returned from rc_read_config was caught wrongly
345 - Username/password not passed on to radius server due to wrong length.
347 The presumption is that some radiusclient API changes for 4.51/PH/17
348 were not taken care of correctly. The code is still untested by me (my
349 Linux distribution still has 0.3.2 of radiusclient), but it was
350 contributed by a Radius user.
352 PH/24 When doing a callout, the value of $domain wasn't set correctly when
353 expanding the "port" option of the smtp transport.
355 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
356 while reading a MIME header. Thanks to Tom Hughes for a patch.
358 PH/24 Include config.h inside local_scan.h so that configuration settings are
361 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
362 that in an ACL for RCPT (for example), you can examine exactly what was
365 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
366 commands, but it was not correctly comparing the address with the actual
367 client host address. Thus, it would show the EHLO address in Received:
368 header lines when this was not necessary.
370 PH/27 Added the % operator to ${eval:}.
372 PH/28 Exim tries to create and chdir to its spool directory when it starts;
373 it should be ignoring failures (because with -C, for example, it has lost
374 privilege). It wasn't ignoring creation failures other than "already
377 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
380 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
381 because an input line was too long, either on its own, or by virtue of
382 too many continuations, the temporary file was not being removed, and the
383 return code was incorrect.
385 PH/31 Missing "BOOL" in function definition in filtertest.c.
387 PH/32 Applied Sieve patches from the maintainer.
389 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
391 PH/33 Added "verify = not_blind".
393 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
394 Local/Makefile (with some defaults set). These are used in built scripts
395 such as exicyclog, but they have never been used in the exim_install
396 script (though there are many overriding facilities there). I have
397 arranged that the exim_install script now takes note of these two
400 PH/35 Installed configuration files for Dragonfly.
402 PH/36 When a locally submitted message by a trusted user did not contain a
403 From: header, and the sender address was obtained from -f or from an SMTP
404 MAIL command, and the trusted user did not use -F to supply a sender
405 name, $originator_name was incorrectly used when constructing a From:
406 header. Furthermore, $originator_name was used for submission mode
407 messages from external hosts without From: headers in a similar way,
408 which is clearly wrong.
410 PH/37 Added control=suppress_local_fixups.
412 PH/38 When log_selector = +received_sender was set, and the addition of the
413 sender made the log line's construction buffer exactly full, or one byte
414 less than full, an overflow happened when the terminating "\n" was
417 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
418 when the result of a list match is failure because a DNS lookup failed.
420 PH/40 RM_COMMAND is now used in the building process.
422 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
423 the "build-* directories that it finds.
425 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
426 address in a domain literal was a prefix of an interface address.
428 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
429 when verifying a sender address, unless rewrite_headers is false.
431 PH/44 Wrote a long comment about why errors_to addresses are verified as
432 recipients, not senders.
434 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
435 the ratelimit ACL was added.
437 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
439 PH/46 Added extra information about PostgreSQL errors to the error string.
441 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
442 avoiding the need to install special zones in a real server. This is
443 backwards compatible; if it can't find the fake resolver, it drops back.
444 Thus, both old and new test suites can be run.
446 TF/02 Added util/ratelimit.pl
448 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
449 clock is set back in time.
451 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
452 Candler <B.Candler@pobox.com>.
454 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
455 OK for addresses that are the result of redirections.
457 PH/48 A number of further additions for the benefit of the new test suite,
458 including a fake gethostbyname() that interfaces to the fake DNS resolver
461 TF/06 The fix for widen_domains has also been applied to qualify_single and
462 search_parents which are the other dnslookup options that can cause
465 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
468 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
470 TF/07 Exim produced the error message "an SRV record indicated no SMTP
471 service" if it encountered an MX record with an empty target hostname.
472 The message is now "an MX or SRV record indicated no SMTP service".
474 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
475 if the DNS of the sending site is misconfigured. This is quite a
476 common situation. This change restores the behaviour of treating a
477 helo verification defer as a failure.
479 PH/51 If self=fail was set on a router, the bounce message did not include the
480 actual error message.
486 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
488 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
489 command itself ended in error, the underlying message about the transport
490 filter timeout was being overwritten with the pipe command error. Now the
491 underlying error message should be appended to the second error message.
493 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
495 PH/02 Reduce the amount of output that "make" produces by default. Full output
496 can still be requested.
498 PH/03 The warning log line about a condition test deferring for a "warn" verb
499 was being output only once per connection, rather than after each
500 occurrence (because it was using the same function as for successful
501 "warn" verbs). This seems wrong, so I have changed it.
503 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
504 it should not have, which might have caused a crash in the right
505 circumstances, but probably never did.
507 PH/04 Installed a modified version of Tony Finch's patch to make submission
508 mode fix the return path as well as the Sender: header line, and to
509 add a /name= option so that you can make the user's friendly name appear
512 TF/03 Added the control = fakedefer ACL modifier.
514 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
515 Mark Lowes for thorough testing.
517 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
519 TK/03 Merged latest SRS patch from Miles Wilton.
521 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
522 with the definition in sysexits.h (which is #included earlier).
523 Fortunately, Exim does not actually use EX_OK. The code used to try to
524 preserve the sysexits.h value, by assumimg that macro definitions were
525 scanned for macro replacements. I have been disabused of this notion,
526 so now the code just undefines EX_OK before #including unistd.h.
528 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
529 in the smtp transport. When a block could not be written in a single
530 write() function, the timeout was being re-applied to each part-write.
531 This seems wrong - if the receiver was accepting one byte at a time it
532 would take for ever. The timeout is now adjusted when this happens. It
533 doesn't have to be particularly precise.
535 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
536 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
538 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
539 without a domain if the check to <postmaster@domain> fails.
541 SC/01 Eximstats: added -xls and the ability to specify output files
542 (patch written by Frank Heydlauf).
544 SC/02 Eximstats: use FileHandles for outputing results.
546 SC/03 Eximstats: allow any combination of xls, txt, and html output.
548 SC/04 Eximstats: fixed display of large numbers with -nvr option
550 SC/05 Eximstats: fixed merging of reports with empty tables.
552 SC/06 Eximstats: added the -include_original_destination flag
554 SC/07 Eximstats: removed tabs and trailing whitespace.
556 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
558 TK/06 MBOX spool code: Add real "From " MBOX separator line
559 so the .eml file is really in mbox format (even though
560 most programs do not really care). Patch from Alex Miller.
562 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
563 The latter is generated from $received_to and is only set if the
564 message has one envelope recipient. SA can use these headers,
565 obviously out-of-the-box. Patch from Alex Miller.
567 PH/08 The ${def test on a variable was returning false if the variable's
568 value was "0", contrary to what the specification has always said!
569 The result should be true unless the variable is empty.
571 PH/09 The syntax error of a character other than { following "${if
572 def:variable_name" (after optional whitespace) was not being diagnosed.
573 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
574 accidental colon was present, for example, could give incorrect results.
576 PH/10 Tidied the code in a number of places where the st_size field of a stat()
577 result is used (not including appendfile, where other changes are about
580 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
581 This involved changing a lot of size variables from int to off_t. It
582 should work with maildirs and everything.
584 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
585 spamd dying while we are connected to it.
587 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
590 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
591 function for BDB 4.3.
593 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
595 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
596 the sources for PCRE have been re-arranged and more files are now
599 PH/15 The code I had for printing potentially long long variables in PH/11
600 above was not the best (it lost precision). The length of off_t variables
601 is now inspected at build time, and an appropriate printing format (%ld
602 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
603 to be "long long int" or "long int". This is needed for the internal
604 formatting function string_vformat().
606 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
607 the configuration file to be ":syslog", then the script "guesses" where
608 the logs files are, rather than using the compiled in default. In our
609 case the guess is not the same as the compiled default, so the script
610 suddenly stopped working when I started to use syslog. The patch checks
611 to see if log_file_path is "". If so, it attempts to read it from exim
612 with no configuration file to get the compiled in version, before it
613 falls back to the previous guessing code."
615 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
616 implementing BATV in an Exim configuration. See NewStuff for the gory
619 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
620 Makefile that are specific to HP-UX.
622 PH/18 If the "use_postmaster" option was set for a recipient callout together
623 with the "random" option, the postmaster address was used as the MAIL
624 FROM address for the random test, but not for the subsequent recipient
625 test. It is now used for both.
627 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
628 patch removes a few documentation additions to RFC 3028, because the
629 latest draft now contains them. It adds the new en;ascii-case comparator
630 and a new error check for 8bit text in MIME parts. Comparator and
631 require names are now matched exactly. I enabled the subaddress
632 extension, but it is not well tested yet (read: it works for me)."
634 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
635 rework some of the code of TK/09 above to avoid the hardwired use of
636 "%lld" and "long long". Replaced the call to snprintf() with a call to
639 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
640 records point to non-existent hosts", "retry timeout exceeded", and
641 "retry time not reached for any host after a long failure period".
643 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
644 experimental DomainKeys support:
646 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
647 (2) On an error such as an illegally used "control", the wrong name for
648 the control was given.
650 These problems did NOT occur unless DomainKeys support was compiled.
652 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
654 PH/24 Added ${if match_ip condition.
656 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
657 databases so that it will be absolutely obvious if a crash occurs in the
658 DB library. This is a regular occurrence (often caused by mis-matched
661 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
662 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
663 picked up on a user's system that detects such things. There doesn't seem
664 to be a gcc warning option for this - only an attribute that has to be
665 put on the function's prototype. It seems that in Fedora Core 4 they have
666 set this on a number of new functions. No doubt there will be more in due
669 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
670 specify a transport. However, if an address that was verified by such a
671 router was the subject of a callout, Exim crashed because it tried to
672 read the rcpt_include_affixes from the non-existent transport. Now it
673 just assumes that the setting of that option is false. This bug was
674 introduced by 4.51/PH/31.
676 PH/28 Changed -d+all to exclude +memory, because that information is very
677 rarely of interest, but it makes the output a lot bigger. People tend to
678 do -d+all out of habit.
680 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
681 code in os-type was giving problems when libc.so lives in lib64, like on
684 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
685 aren't the modern standard, and it seems that some systems' include files
686 don't always have them. Exim was already checking for some of the newer
687 ones like T_AAAA, and defining it itself. I've added checks for all the
688 record types that Exim uses.
690 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
691 not automatically generating a new one, as it is supposed to. This
692 prevented TLS from working. If the file did exist, but contained invalid
693 data, a new version was generated, as expected. It was only the case of a
694 non-existent file that was broken.
696 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
697 with a change in libdomainkeys > 0.64.
699 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
700 from DNS. If the selector record carries the flag, it now has
701 precedence over the domain-wide flag.
703 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
705 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
706 the use of a port name that isn't defined in /etc/services) occurred, the
707 message was deferred as in a normal delivery, and thus remained on the
708 spool, instead of being failed because of the mua_wrapper setting. This
709 is now fixed, and I tidied up some of the mua_wrapper messages at the
712 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
713 the messages in a hash of arrays rather than using individual hashes.
714 This is a bit cleaner and results in dramatic memory savings, albeit
715 at a slight CPU cost.
717 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
718 as requested by Marc Sherman.
720 SC/10 Eximstats: added histograms for user specified patterns as requested
723 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
725 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
726 fopen() in the content-scanning modules that did not already have it.
732 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
733 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
735 TK/02 Fix ACL "control" statement not being available in MIME ACL.
737 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
739 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
740 to test Sieve filters that use "vacation".
742 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
743 that changes the way the GnuTLS parameters are stored in the cache file.
744 The new format can be generated externally. For backward compatibility,
745 if the data in the cache doesn't make sense, Exim assumes it has read an
746 old-format file, and it generates new data and writes a new file. This
747 means that you can't go back to an older release without removing the
750 PH/03 A redirect router that has both "unseen" and "one_time" set does not
751 work if there are any delivery delays because "one_time" forces the
752 parent to be marked "delivered", so its unseen clone is never tried
753 again. For this reason, Exim now forbids the simultaneous setting of
756 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
757 redirected to themselves ("homonym" addresses). Read the long ChangeLog
758 entry if you want to know the details. The fix, however, neglected to
759 consider the case when local delivery batching is involved. The test for
760 "previously delivered" was not happening when checking to see if an
761 address could be batched with a previous (undelivered) one; under
762 certain circumstances this could lead to multiple deliveries to the same
765 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
766 in its include files, and this causes problems building Exim.
768 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
769 header_syntax) but Exim was just ignoring anything given after a slash.
770 In particular, this caused confusion with an attempt to use "verify =
771 reverse_host_lookup/defer_ok". An error is now given when options are
772 supplied for verify items that do not have them. (Maybe reverse_host_
773 lookup should have a defer_ok option, but that's a different point.)
775 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
776 defined by RFC 821) to 2048, because there were problems with some AUTH
777 commands, and RFC 1869 says the size should be increased for extended
778 SMTP commands that take arguments.
780 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
783 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
784 "unknown" error; now it says that the functionality isn't in the binary.
786 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
787 an address' error message when a string expansion fails (syntax or
788 whatever). Otherwise the password may appear in the log. Following change
789 PH/42 below, there is no longer a chance of it appearing in a bounce
792 PH/11 Installed exipick version 20050225.0 from John Jetmore.
794 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
795 of its addresses was ever tried. (Bugzilla bug #2.)
797 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
798 the result incorrectly in the debug output. (It correctly added a newline
799 to what was transported.)
801 TF/01 Added $received_time.
803 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
804 commented out examples of how to interface to a virus scanner and to
805 SpamAssassin. Also added commented examples of av_scanner and
806 spamd_address settings.
808 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
809 and controls are allowed in which ACLs. There were a couple of minor
810 errors. Some of the entries in the conditions table (which is a table of
811 where they are NOT allowed) were getting very unwieldy; rewrote them as a
812 negation of where the condition IS allowed.
814 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
816 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
817 header file does not have a version number, so I've had to invent a new
818 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
819 API. The code is untested by me (my Linux distribution still has 0.3.2 of
820 radiusclient), but it was contributed by a Radius user.
822 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
823 files or directories, for OpenSSL.
825 PH/19 When an Exim process that is running as root has to create an Exim log
826 file, it does so in a subprocess that runs as exim:exim so as to get the
827 ownership right at creation (otherwise, other Exim processes might see
828 the file with the wrong ownership). There was no test for failure of this
829 fork() call, which would lead to the process getting stuck as it waited
830 for a non-existent subprocess. Forks do occasionally fail when resources
831 run out. I reviewed all the other calls to fork(); they all seem to check
834 PH/20 When checking for unexpected SMTP input at connect time (before writing
835 the banner), Exim was not dealing correctly with a non-positive return
836 from the read() function. If the client had disconnected by this time,
837 the result was a log entry for a synchronization error with an empty
838 string after "input=" when read() returned zero. If read() returned -1
839 (an event I could not check), uninitialized data bytes were printed.
840 There were reports of junk text (parts of files, etc) appearing after
843 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
845 PH/22 Added support for macro redefinition, and (re)definition in between
846 driver and ACL definitions.
848 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
849 forgetting to use the resulting value; it was using the unexpanded value.
851 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
852 hadn't been configured. The fix is from Juergen Kreileder, who
853 understands it better than I do:
855 "Here's what I see happening with three configured cyrus_sasl
856 authenticators configured (plain, login, cram-md5):
858 On startup auth_cyrus_sasl_init() gets called for each of these.
859 This means three calls to sasl_listmech() without a specified mech_list.
860 => SASL tests which mechs of all available mechs actually work
861 => three warnings about OTP not working
862 => the returned list contains: plain, login, cram-md5, digest-md5, ...
864 With the patch, sasl_listmech() also gets called three times. But now
865 SASL's mech_list option is set to the server_mech specified in the the
866 authenticator. Or in other words, the answer from sasl_listmech()
867 gets limited to just the mech you're testing for (which is different
869 => the return list contains just 'plain' or 'login', 'cram-md5' or
870 nothing depending on the value of ob->server_mech.
872 I've just tested the patch: Authentication still works fine,
873 unavailable mechs specified in the exim configuration are still
874 caught, and the auth.log warnings about OTP are gone."
876 PH/25 When debugging is enabled, the contents of the command line are added
877 to the debugging output, even when log_selector=+arguments is not
880 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
881 answer is "GNU", and only if the return is "GNU/something" is the answer
884 PH/27 $acl_verify_message is now set immediately after the failure of a
885 verification in an ACL, and so is available in subsequent modifiers. In
886 particular, the message can be preserved by coding like this:
888 warn !verify = sender
889 set acl_m0 = $acl_verify_message
891 Previously, $acl_verify_message was set only while expanding "message"
892 and "log_message" when a very denied access.
894 PH/28 Modified OS/os.c-Linux with
896 -#ifndef OS_LOAD_AVERAGE
897 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
899 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
900 nomenclature these days.)
902 PH/29 Installed patch from the Sieve maintainer that adds the options
903 sieve_useraddress and sieve_subaddress to the redirect router.
905 PH/30 In these circumstances:
906 . Two addresses routed to the same list of hosts;
907 . First host does not offer TLS;
908 . First host accepts first address;
909 . First host gives temporary error to second address;
910 . Second host offers TLS and a TLS session is established;
911 . Second host accepts second address.
912 Exim incorrectly logged both deliveries with the TLS parameters (cipher
913 and peerdn, if requested) that were in fact used only for the second
916 PH/31 When doing a callout as part of verifying an address, Exim was not paying
917 attention to any local part prefix or suffix that was matched by the
918 router that accepted the address. It now behaves in the same way as it
919 does for delivery: the affixes are removed from the local part unless
920 rcpt_include_affixes is set on the transport.
922 PH/32 Add the sender address, as F=<...>, to the log line when logging a
923 timeout during the DATA phase of an incoming message.
925 PH/33 Sieve envelope tests were broken for match types other than :is. I have
926 applied a patch sanctioned by the Sieve maintainer.
928 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
929 the uid or gid is negative. A case of a negative gid caused this to be
930 noticed. The fix allows for either to be negative.
932 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
933 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
936 PH/36 The change PH/12 above was broken. Fixed it.
938 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
939 the grounds that routing the same address twice would always produce the
940 same answer. This might have been true once, but it is certainly no
941 longer true now. Routing a child address may depend on the previous
942 routing that produced that child. Some complicated redirection strategies
943 went wrong when messages had multiple recipients, and made Exim's
944 behaviour dependent on the order in which the addresses were given.
946 I have moved the duplicate checking until after the routing is complete.
947 Exim scans the addresses that are assigned to local and remote
948 transports, and removes any duplicates. This means that more work will be
949 done, as duplicates will always all be routed, but duplicates are
950 presumably rare, so I don't expect this is of any significance.
952 For deliveries to pipes, files, and autoreplies, the duplicate checking
953 still happens during the routing process, since they are not going to be
956 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
957 It corrects a timeout issue with spamd. This is Ian's comment: "The
958 background is that sometimes spamd either never reads data from a
959 connection it has accepted, or it never writes response data. The exiscan
960 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
961 blindly assumes that writes won't block so it may never time out."
963 PH/39 Allow G after quota size as well as K and M.
965 PH/40 The value set for $authenticated_id in an authenticator may not contain
966 binary zeroes or newlines because the value is written to log lines and
967 to spool files. There was no check on this. Now the value is run through
968 the string_printing() function so that such characters are converted to
969 printable escape sequences.
971 PH/41 $message_linecount is a new variable that contains the total number of
972 lines in the message. Compare $body_linecount, which is the count for the
975 PH/42 Exim no longer gives details of delivery errors for specific addresses in
976 bounce and delay warning messages, except in certain special cases, which
979 (a) An SMTP error message from a remote host;
980 (b) A message specified in a :fail: redirection;
981 (c) A message specified in a "fail" command in a system filter;
982 (d) A message specified in a FAIL return from the queryprogram router;
983 (e) A message specified by the cannot_route_message router option.
985 In these cases only, Exim does include the error details in bounce and
986 warning messages. There are also a few cases where bland messages such
987 as "unrouteable address" or "local delivery error" are given.
989 PH/43 $value is now also set for the "else" part of a ${run expansion.
991 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
992 being worked on, but at least Exim now implements the latest version to
995 PH/45 In a pipe transport, although a timeout while waiting for the pipe
996 process to complete was treated as a delivery failure, a timeout while
997 writing the message to the pipe was logged, but erroneously treated as a
998 successful delivery. Such timeouts include transport filter timeouts. For
999 consistency with the overall process timeout, these timeouts are now
1000 treated as errors, giving rise to delivery failures by default. However,
1001 there is now a new Boolean option for the pipe transport called
1002 timeout_defer, which, if set TRUE, converts the failures into defers for
1003 both kinds of timeout. A transport filter timeout is now identified in
1006 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1007 systems where "make" and "gmake" are different, calling "gmake" at top
1008 level broke things. I've arranged for the value of $(MAKE) to be passed
1009 from the Makefile to this script so that it can call the same version of
1013 A note about Exim versions 4.44 and 4.50
1014 ----------------------------------------
1016 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1017 changes of various kinds. As a consequence, a big documentation update was
1018 needed. This delayed the release for rather longer than seemed good, especially
1019 in the light of a couple of (minor) security issues. Therefore, the changes
1020 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1021 release. So 4.44 and 4.50 are in effect two different branches that both start
1024 I have left the 4.50 change log unchanged; it contains all the changes since
1025 4.43. The change log for 4.44 is below; many of its items are identical to
1026 those for 4.50. This seems to be the most sensible way to preserve the
1027 historical information.
1033 1. Minor wording change to the doc/README.SIEVE file.
1035 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1036 computation of the current number of files was incorrect.
1038 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1039 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1042 4. Give more explanation in the error message when the command for a transport
1043 filter fails to execute.
1045 5. There are several places where Exim runs a non-Exim command in a
1046 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1047 was being done only for the command run by the queryprogram router. It is
1048 now done for all such subprocesses. The other cases are: ${run, transport
1049 filters, and the commands run by the lmtp and pipe transports.
1051 6. Added CONFIGURE_GROUP build-time option.
1053 7. Some older OS have a limit of 256 on the maximum number of file
1054 descriptors. Exim was using setrlimit() to set 1000 as a large value
1055 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1056 systems. I've change it so that if it can't get 1000, it tries for 256.
1058 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1059 was an oversight, and furthermore, ever since the addition of extra
1060 controls (e.g. 4.43/32), the checks on when to allow different forms of
1061 "control" were broken. There should now be diagnostics for all cases when a
1062 control that does not make sense is encountered.
1064 9. Added the /retain_sender option to "control=submission".
1066 10. $recipients is now available in the predata ACL (oversight).
1068 11. Tidy the search cache before the fork to do a delivery from a message
1069 received from the command line. Otherwise the child will trigger a lookup
1070 failure and thereby defer the delivery if it tries to use (for example) a
1071 cached ldap connection that the parent has called unbind on.
1073 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1074 of $address_data from the recipient verification was clobbered by the
1075 sender verification.
1077 13. The value of address_data from a sender verification is now available in
1078 $sender_address_data in subsequent conditions in the ACL statement.
1080 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1082 15. Added a new option "connect=<time>" to callout options, to set a different
1085 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1086 was its contents. (It was OK if the option was not defined at all.)
1088 17. A "Completed" log line is now written for messages that are removed from
1089 the spool by the -Mrm option.
1091 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1092 information about exactly what failed.
1094 19. Added -dd to debug only the daemon process.
1096 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1097 handles timeouts, both on the server side and network timeouts. Renamed the
1098 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1100 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1101 It is now set to "smtps".
1103 22. $host_address is now set to the target address during the checking of
1104 ignore_target_hosts.
1106 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1107 being passed; this would have caused $sender_host_name to have been used if
1108 matching the list had actually called for a host name (not very likely,
1109 since this list is usually IP addresses). A host name is now passed as
1112 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1113 code that specifies a non-restarting handler (typically sigaction() in
1114 modern systems) in an attempt to fix a rare and obscure crash bug.
1116 25. Narrowed the window for a race in the daemon that could cause it to ignore
1117 SIGCHLD signals. This is not a major problem, because they are used only to
1118 wake it up if nothing else does.
1120 26. A malformed maildirsize file could cause Exim to calculate negative values
1121 for the mailbox size or file count. Odd effects could occur as a result.
1122 The maildirsize information is now recalculated if the size or filecount
1125 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1126 support for a long time. Removed HAVE_SYS_VFS_H.
1128 28. Installed the latest version of exipick from John Jetmore.
1130 29. In an address list, if the pattern was not a regular expression, an empty
1131 subject address (from a bounce message) matched only if the pattern was an
1132 empty string. Non-empty patterns were not even tested. This was the wrong
1133 because it is perfectly reasonable to use an empty address as part of a
1134 database query. An empty address is now tested by patterns that are
1135 lookups. However, all the other forms of pattern expect the subject to
1136 contain a local part and a domain, and therefore, for them, an empty
1137 address still always fails if the pattern is not itself empty.
1139 30. Exim went into a mad DNS loop when attempting to do a callout where the
1140 host was specified on an smtp transport, and looking it up yielded more
1141 than one IP address.
1143 31. Re-factored the code for checking spool and log partition space into a
1144 function that finds that data and another that does the check. The former
1145 is then used to implement four new variables: $spool_space, $log_space,
1146 $spool_inodes, and $log_inodes.
1148 32. The RFC2047 encoding function was originally intended for short strings
1149 such as real names; it was not keeping to the 75-character limit for
1150 encoded words that the RFC imposes. It now respects the limit, and
1151 generates multiple encoded words if necessary. To be on the safe side, I
1152 have increased the buffer size for the ${rfc2047: expansion operator from
1155 33. It is now permitted to omit both strings after an "if" condition; if the
1156 condition is true, the result is "true". As before, when the second string
1157 is omitted, a false condition yields an empty string. This makes it less
1158 cumbersome to write custom ACL and router conditions.
1160 34. Failure to deliver a bounce message always caused it to be frozen, even if
1161 there was an errors_to setting on the router. The errors_to setting is now
1164 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1165 canonical form (fully expanded) before being placed in
1166 $sender_host_address.
1168 36. The table in the code that translates DNS record types into text (T_A to
1169 "A" for instance) was missing entries for NS and CNAME. It is just possible
1170 that this could have caused confusion if both these types were looked up
1171 for the same domain, because the text type is used as part of Exim's
1172 per-process caching. But the chance of anyone hitting this buglet seems
1175 37. The dnsdb lookup has been extended in a number of ways.
1177 (1) There is a new type, "zns", which walks up the domain tree until it
1178 finds some nameserver records. It should be used with care.
1180 (2) There is a new type, "mxh", which is like "mx" except that it returns
1181 just the host names, not the priorities.
1183 (3) It is now possible to give a list of domains (or IP addresses) to be
1184 looked up. The behaviour when one of the lookups defers can be
1185 controlled by a keyword.
1187 (4) It is now possible to specify the separator character for use when
1188 multiple records are returned.
1190 38. The dnslists ACL condition has been extended: it is now possible to supply
1191 a list of IP addresses and/or domains to be looked up in a particular DNS
1194 39. Added log_selector=+queue_time_overall.
1196 40. When running the queue in the test harness, wait just a tad after forking a
1197 delivery process, to get repeatability of debugging output.
1199 41. Include certificate and key file names in error message when GnuTLS fails
1200 to set them up, because the GnuTLS error message doesn't include the name
1201 of the failing file when there is a problem reading it.
1203 42. Allow both -bf and -bF in the same test run.
1205 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1207 44. The "Exiscan patch" is now merged into the mainline Exim source.
1209 45. Sometimes the final signoff response after QUIT could fail to get
1210 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1211 < 0 before doing a fflush(). This bug looks as though it goes back to the
1212 introduction of TLS in release 3.20, but "sometimes" must have been rare
1213 because the tests only now provoked it.
1215 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1216 (this can affect the format of dates).
1218 47. exim_tidydb, when checking for the continued existence of a message for
1219 which it has found a message-specific retry record, was not finding
1220 messages that were in split spool directories. Consequently, it was
1221 deleting retry records that should have stayed in existence.
1223 48. Steve fixed some bugs in eximstats.
1225 49. The SPA authentication driver was not abandoning authentication and moving
1226 on to the next authenticator when an expansion was forced to fail,
1227 contradicting the general specification for all authenticators. Instead it
1228 was generating a temporary error. It now behaves as specified.
1230 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1231 (the order specifies the preference for clients). The order is now AES256,
1232 AES128, 3DES, ARCFOUR128.
1234 51. Small patch to Sieve code - explicitly set From: when generating an
1237 52. Exim crashed if a remote delivery caused a very long error message to be
1238 recorded - for instance if somebody sent an entire SpamAssassin report back
1239 as a large number of 550 error lines. This bug was coincidentally fixed by
1240 increasing the size of one of Exim's internal buffers (big_buffer) that
1241 happened as part of the Exiscan merge. However, to be on the safe side, I
1242 have made the code more robust (and fixed the comments that describe what
1245 53. Now that there can be additional text after "Completed" in log lines (if
1246 the queue_time_overall log selector is set), a one-byte patch to exigrep
1247 was needed to allow it to recognize "Completed" as not the last thing in
1250 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1251 patch that reportedly fixes this has been added. I am not expert enough to
1252 create a test for it. This is what the patch creator wrote:
1254 "I found a little strange behaviour of ldap code when working with
1255 Windows 2003 AD Domain, where users was placed in more than one
1256 Organization Units. When I tried to give exim partial DN, the exit code
1257 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1258 But simultaneously result of request was absolutely normal ldap result,
1259 so I produce this patch..."
1261 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1262 REFERENCE, so I have modified the code to exclude the patch when that macro
1265 55. Some experimental protocols are using DNS PTR records for new purposes. The
1266 keys for these records are domain names, not reversed IP addresses. The
1267 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1268 leaves it alone. Component reversal etc. now happens only for IP addresses.
1271 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1273 57. Double the size of the debug message buffer (to 2048) so that more of very
1274 long debug lines gets shown.
1276 58. The exicyclog utility now does better if the number of log files to keep
1277 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1279 59. Two changes related to the smtp_active_hostname option:
1281 (1) $smtp_active_hostname is now available as a variable.
1282 (2) The default for smtp_banner uses $smtp_active_hostname instead
1283 of $primary_hostname.
1285 60. The host_aton() function is supposed to be passed a string that is known
1286 to be a valid IP address. However, in the case of IPv6 addresses, it was
1287 not checking this. This is a hostage to fortune. Exim now panics and dies
1288 if the condition is not met. A case was found where this could be provoked
1289 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1290 components; fortuitously, this particular loophole had already been fixed
1291 by change 4.50/55 above.
1293 If there are any other similar loopholes, the new check in host_aton()
1294 itself should stop them being exploited. The report I received stated that
1295 data on the command line could provoke the exploit when Exim was running as
1296 exim, but did not say which command line option was involved. All I could
1297 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1298 running as the user.
1301 61. There was a buffer overflow vulnerability in the SPA authentication code
1302 (which came originally from the Samba project). I have added a test to the
1303 spa_base64_to_bits() function which I hope fixes it.
1306 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1307 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1309 63. The daemon start-up calls getloadavg() while still root for those OS that
1310 need the first call to be done as root, but it missed one case: when
1311 deliver_queue_load_max is set with deliver_drop_privilege. This is
1312 necessary for the benefit of the queue runner, because there is no re-exec
1313 when deliver_drop_privilege is set.
1315 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1316 This has been fixed.
1318 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1319 was in use, was not putting the data itself into the right store pool;
1320 consequently, it could be overwritten for a subsequent message in the same
1321 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1324 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1326 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1327 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1328 the function were treating the return as a boolean value, which happened to
1329 work because 0=false and not-0=true, but is not correct code.
1331 68. The host_aton() function was not handling scoped IPv6 addresses (those
1332 with, for example, "%eth0" on the end) correctly.
1334 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1335 negated items (that is, ~something) in unsigned ints. Some compilers
1336 apparently mutter when there is no cast.
1338 70. If an address verification called from an ACL failed, and did not produce a
1339 user-specific message (i.e. there was only a "system" message), nothing was
1340 put in $acl_verify_message. In this situation, it now puts the system
1343 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1344 session; change 4.31/43 added the unwanted input to the log line - except
1345 that it did not do this in the start of session case. It now does.
1347 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1348 This is wrong and can cause the other end to generate a synchronization
1349 error if it is another Exim or anything else that does the synchronization
1350 check. A QUIT command is no longer sent after a timeout.
1352 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1353 during host lookups.
1355 74. The defer_ok option of callout verification was not working if it was used
1356 when verifying addresses in header lines, that is, for this case:
1358 verify = header_sender/callout=defer_ok
1360 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1361 those file descriptors could be used for SMTP connections. If anything
1362 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1363 could be sent to the SMTP client, causing chaos. The daemon now opens
1364 stdin, stdout, and stderr to /dev/null when it puts itself into the
1367 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1368 log by default. The user can override this with suitable Perl magic.
1370 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1371 the log message when discard triggers, was not working for the DATA ACL or
1372 for the non-SMTP ACL.
1374 78. Error message wording change in sieve.c.
1376 79. If smtp_accept_max_per_host was set, the number of connections could be
1377 restricted to fewer than expected, because the daemon was trying to set up
1378 a new connection before checking whether the processes handling previous
1379 connections had finished. The check for completed processes is now done
1380 earlier. On busy systems, this bug wouldn't be noticed because something
1381 else would have woken the daemon, and it would have reaped the completed
1384 80. If a message was submitted locally by a user whose login name contained one
1385 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1386 It caused a spool format error. I have fixed the spool reading code. A
1387 related problem was that the "from" clause in the Received: line became
1388 illegal because of the space(s). It is now covered by ${quote_local_part.
1390 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1393 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1395 83. Patch from Sieve maintainer to fix unterminated string problem in
1396 "vacation" handling.
1398 84. Some minor changes to the Linux configuration files to help with other
1399 OS variants using glibc.
1401 85. One more patch for Sieve to update vacation handling to latest spec.
1404 ----------------------------------------------------
1405 See the note above about the 4.44 and 4.50 releases.
1406 ----------------------------------------------------
1412 1. Change 4.43/35 introduced a bug that caused file counts to be
1413 incorrectly computed when quota_filecount was set in an appendfile
1416 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1417 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1420 3. Give more explanation in the error message when the command for a transport
1421 filter fails to execute.
1423 4. There are several places where Exim runs a non-Exim command in a
1424 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1425 was being done only for the command run by the queryprogram router. It is
1426 now done for all such subprocesses. The other cases are: ${run, transport
1427 filters, and the commands run by the lmtp and pipe transports.
1429 5. Some older OS have a limit of 256 on the maximum number of file
1430 descriptors. Exim was using setrlimit() to set 1000 as a large value
1431 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1432 systems. I've change it so that if it can't get 1000, it tries for 256.
1434 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1435 was an oversight, and furthermore, ever since the addition of extra
1436 controls (e.g. 4.43/32), the checks on when to allow different forms of
1437 "control" were broken. There should now be diagnostics for all cases when a
1438 control that does not make sense is encountered.
1440 7. $recipients is now available in the predata ACL (oversight).
1442 8. Tidy the search cache before the fork to do a delivery from a message
1443 received from the command line. Otherwise the child will trigger a lookup
1444 failure and thereby defer the delivery if it tries to use (for example) a
1445 cached ldap connection that the parent has called unbind on.
1447 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1448 of $address_data from the recipient verification was clobbered by the
1449 sender verification.
1451 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1452 was its contents. (It was OK if the option was not defined at all.)
1454 11. A "Completed" log line is now written for messages that are removed from
1455 the spool by the -Mrm option.
1457 12. $host_address is now set to the target address during the checking of
1458 ignore_target_hosts.
1460 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1461 being passed; this would have caused $sender_host_name to have been used if
1462 matching the list had actually called for a host name (not very likely,
1463 since this list is usually IP addresses). A host name is now passed as
1466 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1467 code that specifies a non-restarting handler (typically sigaction() in
1468 modern systems) in an attempt to fix a rare and obscure crash bug.
1470 15. Narrowed the window for a race in the daemon that could cause it to ignore
1471 SIGCHLD signals. This is not a major problem, because they are used only to
1472 wake it up if nothing else does.
1474 16. A malformed maildirsize file could cause Exim to calculate negative values
1475 for the mailbox size or file count. Odd effects could occur as a result.
1476 The maildirsize information is now recalculated if the size or filecount
1479 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1480 support for a long time. Removed HAVE_SYS_VFS_H.
1482 18. Updated exipick to current release from John Jetmore.
1484 19. Allow an empty sender to be matched against a lookup in an address list.
1485 Previously the only cases considered were a regular expression, or an
1488 20. Exim went into a mad DNS lookup loop when doing a callout where the
1489 host was specified on the transport, if the DNS lookup yielded more than
1492 21. The RFC2047 encoding function was originally intended for short strings
1493 such as real names; it was not keeping to the 75-character limit for
1494 encoded words that the RFC imposes. It now respects the limit, and
1495 generates multiple encoded words if necessary. To be on the safe side, I
1496 have increased the buffer size for the ${rfc2047: expansion operator from
1499 22. Failure to deliver a bounce message always caused it to be frozen, even if
1500 there was an errors_to setting on the router. The errors_to setting is now
1503 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1504 canonical form (fully expanded) before being placed in
1505 $sender_host_address.
1507 24. Updated eximstats to version 1.33
1509 25. Include certificate and key file names in error message when GnuTLS fails
1510 to set them up, because the GnuTLS error message doesn't include the name
1511 of the failing file when there is a problem reading it.
1513 26. Expand error message when OpenSSL has problems setting up cert/key files.
1516 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1517 (this can affect the format of dates).
1519 28. exim_tidydb, when checking for the continued existence of a message for
1520 which it has found a message-specific retry record, was not finding
1521 messages that were in split spool directories. Consequently, it was
1522 deleting retry records that should have stayed in existence.
1524 29. eximstats updated to version 1.35
1525 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1526 1.35 - bugfix such that pie charts by volume are generated correctly
1528 30. The SPA authentication driver was not abandoning authentication and moving
1529 on to the next authenticator when an expansion was forced to fail,
1530 contradicting the general specification for all authenticators. Instead it
1531 was generating a temporary error. It now behaves as specified.
1533 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1534 (the order specifies the preference for clients). The order is now AES256,
1535 AES128, 3DES, ARCFOUR128.
1537 31. Small patch to Sieve code - explicitly set From: when generating an
1540 32. Exim crashed if a remote delivery caused a very long error message to be
1541 recorded - for instance if somebody sent an entire SpamAssassin report back
1542 as a large number of 550 error lines. This bug was coincidentally fixed by
1543 increasing the size of one of Exim's internal buffers (big_buffer) that
1544 happened as part of the Exiscan merge. However, to be on the safe side, I
1545 have made the code more robust (and fixed the comments that describe what
1548 33. Some experimental protocols are using DNS PTR records for new purposes. The
1549 keys for these records are domain names, not reversed IP addresses. The
1550 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1551 leaves it alone. Component reversal etc. now happens only for IP addresses.
1554 34. The host_aton() function is supposed to be passed a string that is known
1555 to be a valid IP address. However, in the case of IPv6 addresses, it was
1556 not checking this. This is a hostage to fortune. Exim now panics and dies
1557 if the condition is not met. A case was found where this could be provoked
1558 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1559 components; fortuitously, this particular loophole had already been fixed
1560 by change 4.50/55 or 4.44/33 above.
1562 If there are any other similar loopholes, the new check in host_aton()
1563 itself should stop them being exploited. The report I received stated that
1564 data on the command line could provoke the exploit when Exim was running as
1565 exim, but did not say which command line option was involved. All I could
1566 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1567 running as the user.
1570 35. There was a buffer overflow vulnerability in the SPA authentication code
1571 (which came originally from the Samba project). I have added a test to the
1572 spa_base64_to_bits() function which I hope fixes it.
1575 36. The daemon start-up calls getloadavg() while still root for those OS that
1576 need the first call to be done as root, but it missed one case: when
1577 deliver_queue_load_max is set with deliver_drop_privilege. This is
1578 necessary for the benefit of the queue runner, because there is no re-exec
1579 when deliver_drop_privilege is set.
1581 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1582 was in use, was not putting the data itself into the right store pool;
1583 consequently, it could be overwritten for a subsequent message in the same
1584 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1587 38. Sometimes the final signoff response after QUIT could fail to get
1588 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1589 < 0 before doing a fflush(). This bug looks as though it goes back to the
1590 introduction of TLS in release 3.20, but "sometimes" must have been rare
1591 because the tests only now provoked it.
1597 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1598 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1599 Now it returns an integer. A number of calls were still expecting a T/F
1600 return. Fortuitously, in all cases, the tests worked in OK situations,
1601 which is the norm. However, things would have gone wrong on any write
1602 failures on the smtp file descriptor. This function is used when sending
1603 messages over SMTP and also when doing verify callouts.
1605 2. When Exim is called to do synchronous delivery of a locally submitted
1606 message (the -odf or -odi options), it no longer closes stderr before doing
1609 3. Implemented the mua_wrapper option.
1611 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1613 5. Implemented the functions header_remove(), header_testname(),
1614 header_add_at_position(), and receive_remove_recipient(), and exported them
1617 6. If an ACL "warn" statement specified the addition of headers, Exim already
1618 inserted X-ACL-Warn: at the start if there was no header name. However, it
1619 was not making this test for the second and subsequent header lines if
1620 there were newlines in the string. This meant that an invalid header could
1621 be inserted if Exim was badly configured.
1623 7. Allow an ACL "warn" statement to add header lines at the start or after all
1624 the Received: headers, as well as at the end.
1626 8. Added the rcpt_4xx retry error code.
1628 9. Added postmaster_mailfrom=xxx to callout verification option.
1630 10. Added mailfrom=xxxx to the callout verification option, for verify=
1633 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1634 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1636 12. Inserted some casts to stop certain compilers warning when using pointer
1637 differences as field lengths or precisions in printf-type calls (mostly
1638 affecting debugging statements).
1640 13. Added optional readline() support for -be (dynamically loaded).
1642 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1643 same clock tick as a message's arrival, so that its received time was the
1644 same as the "first fail" time on the retry record, and that message
1645 remained on the queue past the ultimate address timeout, every queue runner
1646 would try a delivery (because it was past the ultimate address timeout) but
1647 after another failure, the ultimate address timeout, which should have then
1648 bounced the address, did not kick in. This was a "< instead of <=" error;
1649 in most cases the first failure would have been in the next clock tick
1650 after the received time, and all would be well.
1652 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1653 being recognized when the domain list was tested by the match_domain
1654 condition in an expansion string.
1656 16. Added the ${str2b64: operator.
1658 17. Exim was always calling setrlimit() to set a large limit for the number of
1659 processes, without checking whether the existing limit was already
1660 adequate. (It did check for the limit on file descriptors.) Furthermore,
1661 errors from getrlimit() and setrlimit() were being ignored. Now they are
1662 logged to the main and panic logs, but Exim does carry on, to try to do its
1663 job under whatever limits there are.
1665 18. Imported PCRE 5.0.
1667 19. Trivial typo in log message " temporarily refused connection" (the leading
1670 20. If the log selector return_path_on_delivery was set and an address was
1671 redirected to /dev/null, the delivery process crashed because it assumed
1672 that a return path would always be set for a "successful" delivery. In this
1673 case, the whole delivery is bypassed as an optimization, and therefore no
1676 21. Internal re-arrangement: the function for sending a challenge and reading
1677 a response while authentication was assuming a zero-terminated challenge
1678 string. It's now changed to take a pointer and a length, to allow for
1679 binary data in such strings.
1681 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1683 23. Exim was not respecting finduser_retries when seeking the login of the
1684 uid under which it was called; it was always trying 10 times. (The default
1685 setting of finduser_retries is zero.) Also, it was sleeping after the final
1686 failure, which is pointless.
1688 24. Implemented tls_on_connect_ports.
1690 25. Implemented acl_smtp_predata.
1692 26. If the domain in control=submission is set empty, Exim assumes that the
1693 authenticated id is a complete email address when it generates From: or
1694 Sender: header lines.
1696 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1697 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1698 chown and chgrp in /bin and hostname in /usr/bin.
1700 28. Exim was keeping the "process log" file open after each use, just as it
1701 does for the main log. This opens the possibility of it remaining open for
1702 long periods when the USR1 signal hits a daemon. Occasional processlog
1703 errors were reported, that could have been caused by this. Anyway, it seems
1704 much more sensible not to leave this file open at all, so that is what now
1707 29. The long-running daemon process does not normally write to the log once it
1708 has entered its main loop, and it closes the log before doing so. This is
1709 so that log files can straightforwardly be renamed and moved. However,
1710 there are a couple of unusual error situations where the daemon does write
1711 log entries, and I had neglected to close the log afterwards.
1713 30. The text of an SMTP error response that was received during a remote
1714 delivery was being truncated at 512 bytes. This is too short for some of
1715 the long messages that one sometimes sees. I've increased the limit to
1718 31. It is now possible to make retry rules that apply only when a message has a
1719 specific sender, in particular, an empty sender.
1721 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1722 it possible to be selective about when SMTP synchronization is enforced.
1724 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1726 32. Implemented hosts_connection_nolog.
1728 33. Added an ACL for QUIT.
1730 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1733 35. Added mailbox_size and mailbox_filecount to appendfile.
1735 36. Added control = no_multiline_responses to ACLs.
1737 37. There was a bug in the logic of the code that waits for the clock to tick
1738 in the case where the clock went backwards by a substantial amount such
1739 that the microsecond fraction of "now" was more than the microsecond
1740 fraction of "then" (but the whole seconds number was less).
1742 38. Added support for the libradius Radius client library this is found on
1743 FreeBSD (previously only the radiusclient library was supported).
1749 1. When certain lookups returned multiple values in the form name=value, the
1750 quoting of the values was not always being done properly. Specifically:
1751 (a) If the value started with a double quote, but contained no whitespace,
1753 (b) If the value contained whitespace other than a space character (i.e.
1754 tabs or newlines or carriage returns) it was not quoted.
1755 This fix has been applied to the mysql and pgsql lookups by writing a
1756 separate quoting function and calling it from the lookup code. The fix
1757 should probably also be applied to nisplus, ibase and oracle lookups, but
1758 since I cannot test any of those, I have not disturbed their existing code.
1760 2. A hit in the callout cache for a specific address caused a log line with no
1761 reason for rejecting RCPT. Now it says "Previous (cached) callout
1762 verification failure".
1764 3. There was an off-by-one bug in the queryprogram router. An over-long
1765 return line was truncated at 256 instead of 255 characters, thereby
1766 overflowing its buffer with the terminating zero. As well as fixing this, I
1767 have increased the buffer size to 1024 (and made a note to document this).
1769 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1770 when Exim is waiting for an SMTP response from a remote server, Exim
1771 restarts its select() call on the socket, thereby resetting its timeout.
1772 This is not a problem when such interrupts are rare. Somebody set up a cron
1773 job to run exiwhat every 2 minutes, which is less than the normal select()
1774 timeout (5 or 10 minutes). This meant that the select() timeout never
1775 kicked in because it was always reset. I have fixed this by comparing the
1776 time when an interrupt arrives with the time at the start of the first call
1777 to select(). If more time than the timeout has elapsed, the interrupt is
1778 treated as a timeout.
1780 5. Some internal re-factoring in preparation for the addition of Sieve
1781 extensions (by MH). In particular, the "personal" test is moved to a
1782 separate function, and given an option for scanning Cc: and Bcc: (which is
1783 not set for Exim filters).
1785 6. When Exim created an email address using the login of the caller as the
1786 local part (e.g. when creating a From: or Sender: header line), it was not
1787 quoting the local part when it contained special characters such as @.
1789 7. Installed new OpenBSD configuration files.
1791 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1792 try to make them clearer.
1794 9. Callout options, other than the timeout value, were being ignored when
1795 verifying sender addresses in header lines. For example, when using
1797 verify = header_sender/callout=no_cache
1799 the cache was (incorrectly) being used.
1801 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1802 only the Cygwin environment.
1804 11. When return_path_on_delivery was set as a log selector, if different remote
1805 addresses in the same message used different return paths and parallel
1806 remote delivery occurred, the wrong values would sometimes be logged.
1807 (Whenever a remote delivery process finished, the return path value from
1808 the most recently started remote delivery process was logged.)
1810 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1811 lines when AUTH and/or TLS are in use. This is the "received protocol"
1812 field. Exim used to use "asmtp" for authenticated SMTP, without any
1813 indication (in the protocol name) for TLS use. Now it follows the RFC and
1814 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1815 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1816 names appear in log lines as well as in Received: header lines.
1818 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1819 extensions, and comparison tests, and to fix some bugs.
1821 14. Changes to the "personal" filter test:
1823 (1) The test was buggy in that it was just doing the equivalent of
1824 "contains" tests on header lines. For example, if a user's address was
1825 anne@some.where, the "personal" test would incorrectly be true for
1827 To: susanne@some.where
1829 This test is now done by extracting each address from the header in turn,
1830 and checking the entire address. Other tests that are part of "personal"
1831 are now done using regular expressions (for example, to check local parts
1832 of addresses in From: header lines).
1834 (2) The list of non-personal local parts in From: addresses has been
1835 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1836 taken from the Sieve specification recommendations.
1838 (3) If the message contains any header line starting with "List-" it is
1839 treated as non-personal.
1841 (4) The test for "circular" in the Subject: header line has been removed
1842 because it now seems ill-conceived.
1844 15. Minor typos in src/EDITME comments corrected.
1846 16. Installed latest exipick from John Jetmore.
1848 17. If headers_add on a router specified a text string that was too long for
1849 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1850 of string_sprintf() is now avoided.
1852 18. $message_body_size was not set (it was always zero) when running the DATA
1853 ACL and the local_scan() function.
1855 19. For the "mail" command in an Exim filter, no default was being set for
1856 the once_repeat time, causing a random time value to be used if "once" was
1857 specified. (If the value happened to be <= 0, no repeat happened.) The
1858 default is now 0s, meaning "never repeat". The "vacation" command was OK
1859 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1860 (I found it when inspecting the code).
1862 20. There is now an overall timeout for performing a callout verification. It
1863 defaults to 4 times the callout timeout, which applies to individual SMTP
1864 commands during the callout. The overall timeout applies when there is more
1865 than one host that can be tried. The timeout is checked before trying the
1866 next host. This prevents very long delays if there are a large number of
1867 hosts and all are timing out (e.g. when the network connections are timing
1868 out). The value of the overall timeout can be changed by specifying an
1869 additional sub-option for "callout", called "maxwait". For example:
1871 verify = sender/callout=5s,maxwait=20s
1873 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1874 to the end before writing, but this should make it even safer).
1876 22. Exim was forgetting that it had advertised PIPELINING for the second and
1877 subsequent messages on an SMTP connection. It was also not resetting its
1878 memory on STARTTLS and an internal HELO.
1880 23. When Exim logs an SMTP synchronization error within a session, it now
1881 records whether PIPELINING has been advertised or not.
1883 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1884 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1885 rather than long int.
1887 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1889 26. Added the never_mail option to autoreply.
1895 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1896 crash if the getsockname() call failed; this can happen if a connection is
1897 closed very soon after it is established. The problem was simply in the
1898 order in which certain operations were done, causing Exim to try to write
1899 to the SMTP stream before it had set up the file descriptor. The bug has
1900 been fixed by making things happen in the correct order.
1906 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1907 before the connection was closed, thus losing the rejection response.
1909 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1910 some early Solaris releases, but causes trouble in current releases where
1911 socklen_t is defined.
1913 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1916 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1919 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1920 that releasing the top of it at the end releases what was used for sub-
1921 expansions (unless the block got too big). However, discard this block if
1922 the first thing is a variable or header, so that we can use its block when
1923 it is dynamic (useful for very large $message_headers, for example).
1925 6. Lookups now cache *every* query, not just the most recent. A new, separate
1926 store pool is used for this. It can be recovered when all lookup caches are
1927 flushed. Lookups now release memory at the end of their result strings.
1928 This has involved some general refactoring of the lookup sources.
1930 7. Some code has been added to the store_xxx() functions to reduce the amount
1931 of flapping under certain conditions.
1933 8. log_incoming_interface used to affect only the <= reception log lines. Now
1934 it causes the local interface and port to be added to several more SMTP log
1935 lines, for example "SMTP connection from", and rejection lines.
1937 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1939 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1941 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1942 could be overwritten at the end of the current message (or the start of a
1943 new message if it was set in a HELO ACL). The value is now preserved for
1944 the duration of the SMTP connection.
1946 12. If a transport had a headers_rewrite setting, and a matching header line
1947 contained an unqualified address, that address was qualified, even if it
1948 did not match any rewriting rules. The underlying bug was that the values
1949 of the flags that permit the existence of unqualified sender and recipient
1950 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1951 non-local messages, and by -bnq for local messages) were not being
1952 preserved with the message after it was received.
1954 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1955 "next input=" as part of the text comprising the host identity instead of
1956 the correct text. The code was using the same buffer for two different
1957 strings. However, depending on which order the printing function evaluated
1958 its arguments, the bug did not always show up. Under Linux, for example, my
1959 test suite worked just fine.
1961 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1962 doesn't work with some older versions of Perl. It has been changed to "my",
1963 which in any case is probably the better facility to use.
1965 15. A really picky compiler found some instances of statements for creating
1966 error messages that either had too many or two few arguments for the format
1969 16. The size of the buffer for calls to the DNS resolver has been increased
1970 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1971 for addresses that have a lot of PTR records. This alleviates a problem; it
1972 does not fully solve it.
1974 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1975 buffer now truncates the list and logs the incident, which is the same
1976 action as happens when Exim is looking up a host name and its aliases.
1977 Previously in this situation something unpredictable would happen;
1978 sometimes it was "internal error: store_reset failed".
1980 18. If a server dropped the connection unexpectedly when an Exim client was
1981 using GnuTLS and trying to read a response, the client delivery process
1982 crashed while trying to generate an error log message.
1984 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1985 string, for example:
1987 warn message = H1: something\nH2: something
1989 the text was added as a single header line from Exim's point of view
1990 though it ended up OK in the delivered message. However, searching for the
1991 second and subsequent header lines using $h_h2: did not work. This has been
1992 fixed. Similarly, if a system filter added multiple headers in this way,
1993 the routers could not see them.
1995 20. Expanded the error message when iplsearch is called with an invalid key to
1996 suggest using net-iplsearch in a host list.
1998 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1999 ACLs are no longer actually imposed (and a message to that effect is
2002 22. If a "gecos" field in a passwd entry contained escaped characters, in
2003 particular, if it contained a \" sequence, Exim got it wrong when building
2004 a From: or a Sender: header from that name. A second bug also caused
2005 incorrect handling when an unquoted " was present following a character
2006 that needed quoting.
2008 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2009 was not being matched caselessly.
2011 24. Arranged for all hyphens in the exim.8 source to be escaped with
2014 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2015 properly. Recipient callout cache records were still being keyed to include
2016 the sender, even when use_sender was set false. This led to far more
2017 callouts that were necessary. The sender is no longer included in the key
2018 when use_sender is false.
2020 26. Added "control = submission" modifier to ACLs.
2022 27. Added the ${base62d: operator to decode base 62 numbers.
2024 28. dnsdb lookups can now access SRV records.
2026 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2027 the configuration file.
2029 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2030 (-v) mode. This makes the output for a verbose queue run more intelligible.
2032 31. Added a use_postmaster feature to recipient callouts.
2034 32. Added the $body_zerocount variable, containing the number of binary zero
2035 bytes in the message body.
2037 33. The time of last modification of the "new" subdirectory is now used as the
2038 "mailbox time last read" when there is a quota error for a maildir
2041 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2043 35. Added +ignore_unknown as a special item in host lists.
2045 36. Code for decoding IPv6 addresses in host lists is now included, even if
2046 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2047 address was recognized as an IP address, but was then not correctly decoded
2048 into binary, causing unexpected and incorrect effects when compared with
2055 1. Very minor rewording of debugging text in manualroute to say "list of
2056 hosts" instead of "hostlist".
2058 2. If verify=header_syntax was set, and a header line with an unqualified
2059 address (no domain) and a large number of spaces between the end of the
2060 name and the colon was received, the reception process suffered a buffer
2061 overflow, and (when I tested it) crashed. This was caused by some obsolete
2062 code that should have been removed. The fix is to remove it!
2064 3. When running in the test harness, delay a bit after writing a bounce
2065 message to get a bit more predictability in the log output.
2067 4. Added a call to search_tidyup() just before forking a reception process. In
2068 theory, someone could use a lookup in the expansion of smtp_accept_max_
2069 per_host which, without the tidyup, could leave open a database connection.
2071 5. Added the variables $recipient_data and $sender_data which get set from a
2072 lookup success in an ACL "recipients" or "senders" condition, or a router
2073 "senders" option, similar to $domain_data and $local_part_data.
2075 6. Moved the writing of debug_print from before to after the "senders" test
2078 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2079 problems for message scanning, either using a data ACL, or using
2080 local_scan() because the Received: header was not generated till after they
2081 were called (in order to set the time as the time of reception completion).
2082 I have revised the way this works. The header is now generated after the
2083 body is received, but before the ACL or local_scan() are called. After they
2084 are run, the timestamp in the header is updated.
2090 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2091 before starting a queue runner without re-exec. This happened only when
2092 deliver_drop_privilege was set or when the Exim user was set to root. The
2093 effect of the bug was that timeouts during subsequent deliveries caused
2094 crashes instead of being properly handled. The handler is now left at its
2095 default (and expected) setting.
2097 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2098 message, again when deliver_drop_privilege is set or Exim is run as root.
2099 The bug described in (1) was not present in this case, but the tidying up
2100 of the other signals was missing. I have made the two cases consistent.
2102 3. The ignore_target_hosts setting on a manualroute router was being ignored
2103 for hosts that were looked up using the /MX notation.
2105 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2108 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2109 operated on the sender address. After changing the $sender_address to <>
2110 for the sender address verify, Exim was re-instated it as the original
2111 (before rewriting) address, but remembering that it had rewritten it, so it
2112 wasn't rewriting it again. This bug also had the effect of breaking the
2113 sender address verification caching when the sender address was rewritten.
2115 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2116 This has been changed so that if the ip literal address matches
2117 ignore_target_hosts, the router declines.
2119 7. Added expansion conditions match_domain, match_address, and match_local_
2120 part (NOT match_host).
2122 8. The placeholder for the Received: header didn't have a length field set.
2124 9. Added code to Exim itself and to exim_lock to test for a specific race
2125 condition that could lead to file corruption when using MBX delivery. The
2126 issue is with the lockfile that is created in /tmp. If this file is removed
2127 after a process has opened it but before that process has acquired a lock,
2128 there is the potential for a second process to recreate the file and also
2129 acquire a lock. This could lead to two Exim processes writing to the file
2130 at the same time. The added code performs the same test as UW imapd; it
2131 checks after acquiring the lock that its file descriptor still refers to
2132 the same named file.
2134 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2135 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2138 11. Added the smtp_active_hostname option. If used, this will typically be made
2139 to depend on the incoming interface address. Because $interface_address is
2140 not set up until the daemon has forked a reception process, error responses
2141 that can happen earlier (such as "too many connections") no longer contain
2144 12. If an expansion in a condition on a "warn" statement fails because a lookup
2145 defers, the "warn" statement is abandoned, and the next ACL statement is
2146 processed. Previously this caused the whole ACL to be aborted.
2148 13. Added the iplsearch lookup type.
2150 14. Added ident_timeout as a log selector.
2152 15. Added tls_certificate_verified as a log selector.
2154 16. Added a global option tls_require_ciphers (compare the smtp transport
2155 option of the same name). This controls incoming TLS connections.
2157 17. I finally figured out how to make tls_require_ciphers do a similar thing
2158 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2159 before starting the TLS session.
2161 18. Tabs are now shown as \t in -bP output.
2163 19. If the log selector return_path_on_delivery was set, Exim crashed when
2164 bouncing a message because it had too many Received: header lines.
2166 20. If two routers both had headers_remove settings, and the first one included
2167 a superfluous trailing colon, the final name in the first list and the
2168 first name in the second list were incorrectly joined into one item (with a
2169 colon in the middle).
2175 1. Added -C and -D options to the exinext utility, mainly to make it easier
2176 to include in the automated testing, but these could be helpful when
2177 multiple configurations are in use.
2179 2. The exinext utility was not formatting the output nicely when there was
2180 an alternate port involved in the retry record key, nor when there was a
2181 message id as well (for retries that were specific to a specific message
2182 and a specific host). It was also confused by IPv6 addresses, because of
2183 the additional colons they contain. I have fixed the IPv4 problem, and
2184 patched it up to do a reasonable job for IPv6.
2186 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2187 delivery, the log line now contains "pipelined" if PIPELINING was used.
2189 4. An SMTP transport process used to panic and die if the bind() call to set
2190 an explicit outgoing interface failed. This has been changed; it is now
2191 treated in the same way as a connect() failure.
2193 5. A reference to $sender_host_name in the part of a conditional expansion
2194 that was being skipped was still causing a DNS lookup. This no longer
2197 6. The def: expansion condition was not recognizing references to header lines
2198 that used bh_ and bheader_.
2200 7. Added the _cache feature to named lists.
2202 8. The code for checking quota_filecount in the appendfile transport was
2203 allowing one more file than it should have been.
2205 9. For compatibility with Sendmail, the command line option
2213 and sets the incoming protocol and host name (for trusted callers). The
2214 host name and its colon can be omitted when only the protocol is to be set.
2215 Note the Exim already has two private options, -pd and -ps, that refer to
2216 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2217 "s", but I don't think that's a major issue.
2219 10. A number of refactoring changes to the code, none of which should affect
2222 (a) The number of logging options was getting close to filling up the
2223 32-bit word that was used as a bit map. I have split them into two classes:
2224 those that are passed in the argument to log_write(), and those that are
2225 only ever tested independently outside of that function. These are now in
2226 separate 32-bit words, so there is plenty of room for expansion again.
2227 There is no change in the user interface or the logging behaviour.
2229 (b) When building, for example, log lines, the code previously used a
2230 macro that called string_cat() twice, in order to add two strings. This is
2231 not really sufficiently general. Furthermore, there was one instance where
2232 it was actually wrong because one of the argument was used twice, and in
2233 one call a function was used. (As it happened, calling the function twice
2234 did not affect the overall behaviour.) The macro has been replaced by a
2235 function that can join an arbitrary number of extra strings onto a growing
2238 (c) The code for expansion conditions now uses a table and a binary chop
2239 instead of a serial search (which was left over from when there were very
2240 few conditions). Also, it now recognizes conditions like "pam" even when
2241 the relevant support is not compiled in: a suitably worded error message is
2242 given if an attempt is made to use such a condition.
2244 11. Added ${time_interval:xxxxx}.
2246 12. A bug was causing one of the ddress fields not to be passed back correctly
2247 from remote delivery subprocesses. The field in question was not being
2248 subsequently used, so this caused to problems in practice.
2250 13. Added new log selectors queue_time and deliver_time.
2252 14. Might have fixed a bug in maildirsizefile handling that threw up
2253 "unexpected character" debug warnings, and recalculated the data
2254 unnecessarily. In any case, I expanded the warning message to give more
2257 15. Added the message "Restricted characters in address" to the statements in
2258 the default ACL that block characters like @ and % in local parts.
2260 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2261 Three changes have been made:
2263 (a) There was a serious bug; a negative response to MAIL caused the whole
2264 recipient domain to be cached as invalid, thereby blocking all messages
2265 to all local parts at the same domain, from all senders. This bug has
2266 been fixed. The domain is no longer cached after a negative response to
2267 MAIL if the sender used is not empty.
2269 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2272 (c) A new callout option, "use_sender" has been added for people who want
2273 the modified behaviour.
2279 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2282 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2283 indeed breaks things for older releases.
2285 3. Added additional logging to the case where there is a problem reading data
2286 from a filter that is running in a subprocess using a pipe, in order to
2287 try to track down a specific problem.
2289 4. Testing facility fudge: when running in the test harness and attempting
2290 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2291 getting "No route to host". Convert this to a timeout.
2293 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2296 6. Some OS don't have socklen_t but use size_t instead. This affects the
2297 fifth argument of getsockopt() amongst other things. This is now
2298 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2299 can be set for individual OS. I have set it for SunOS5, OSF1, and
2300 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2301 some earlier ones do not.
2303 7. Change 4.30/15 was not doing the test caselessly.
2305 8. The standard form for an IPv6 address literal was being rejected by address
2306 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2307 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2308 this, as well as the form without the "IPv6" on the front (but only when
2309 address literals are enabled, of course).
2311 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2313 10. Exim crashed if a message with an empty sender address specified by -f
2314 encountered a router with an errors_to setting. This could be provoked only
2315 by a command such as
2319 where an empty string was supplied; "<>" did not hit this bug.
2321 11. Installed PCRE release 4.5.
2323 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2324 remained set. It is now erased.
2326 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2327 times from message ids (which are base 36 rather than the normal 62).
2329 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2330 were being counted as actual protocol errors, and logged if the log
2331 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2332 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2333 and DATA following a set of rejected RCPTs do not count as protocol errors.
2334 In other words, Exim assumes they were pipelined, though this may not
2335 actually be the case. Of course, in all cases the client gets an
2336 appropriate error code.
2338 15. If a lookup fails in an ACL condition, a message about the failure may
2339 be available; it is used if testing the ACL cannot continue, because most
2340 such messages specify what the cause of the deferral is. However, some
2341 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2342 that caused an old message to be retained and used if a later statement
2343 caused a defer, replacing the real cause of the deferral.
2345 16. If an IP address had so many PTR records that the DNS lookup buffer
2346 was not large enough to hold them, Exim could crash while trying to process
2347 the truncated data. It now detects and logs this case.
2349 17. Further to 4.21/58, another change has been made: if (and only if) the
2350 first line of a message (the first header line) ends with CRLF, a bare LF
2351 in a subsequent header line has a space inserted after it, so as not to
2352 terminate the header.
2354 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2355 unnecessarily, used atoi() instead of strtol(), and didn't check the
2356 termination when getting file sizes from file names by regex.
2358 19. Completely re-implemented the support for maildirsize files, in the light
2359 of a number of problems with the previous contributed implementation
2360 (4.30/29). In particular:
2362 . If the quota is zero, the maildirsize file is maintained, but no quota is
2365 . If the maildir directory does not exist, it is created before any attempt
2366 to write a maildirsize file.
2368 . The quota value in the file is just a cache; if the quota is changed in
2369 the transport, the new value overrides.
2371 . A regular expression is available for excluding directories from the
2374 20. The autoreply transport checks the characters in options that define the
2375 message's headers; it allows continued headers, but it was checking with
2376 isspace() after an embedded newline instead of explicitly looking for a
2379 21. If all the "regular" hosts to which an address was routed had passed their
2380 expiry times, and had not reached their retry times, the address was
2381 bounced, even if fallback hosts were defined. Now Exim should go on to try
2384 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2385 equivalent code in the SMTP transport. Some hosts send humungous responses
2386 to HELO/EHLO, more than 1024 it seems.
2388 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2389 gives compiler warnings in some environments. I've now done it "properly",
2392 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2393 (because of problems with the built-in one) was declared to return uschar *
2394 instead of char *, causing compiler failure.
2396 25. Fixed a file descriptor leak when processing alias/forward files.
2398 26. Fixed a minor format string issue in dbfn.c.
2400 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2402 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2403 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2406 29. When a sender address is verified, it is cached, to save repeating the test
2407 when there is more than one recipient in a message. However, when the
2408 verification involves a callout, it is possible for different callout
2409 options to be set for different recipients. It is too complicated to keep
2410 track of this in the cache, so now Exim always runs a verification when a
2411 callout is required, relying on the callout cache for the optimization.
2412 The overhead is duplication of the address routing, but this should not be
2415 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2416 to be verified with callout=postmaster, and the main callout worked but the
2417 postmaster check failed, the verification correctly failed. However, if a
2418 subsequent RCPT command asked for sender verification *without* the
2419 postmaster check, incorrect caching caused this verification also to fail,
2422 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2423 it was not caching the DNS options (qualify_single, search_parents) that
2424 were used when the lookup failed. A subsequent lookup with different
2425 options therefore always gave the same answer, though there were cases
2426 where it should not have. (Example: a "domains = !$mx_any" option on a
2427 dnslookup router: the "domains" option is always processed without any
2428 widening, but the router might have qualify_single set.) Now Exim uses the
2429 cached value only when the same options are set.
2431 32. Added John Jetmore's "exipick" utility to the distribution.
2433 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2434 than a timeout (e.g. a certificate is required, and is not provided), an
2435 Exim server now closes the connection immediately. Previously it waited for
2436 the client to close - but if the client is SSL, it seems that they each
2437 wait for each other, leading to a delay before one of them times out.
2439 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2440 maintained 0.8.x compatibility because I don't think many are using it, and
2441 it is clearly obsolete.
2443 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2446 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2447 client certificate was expired. A simple patch fixes this, though I don't
2448 understand the full logic of why the verify callback is called multiple
2451 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2452 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2453 which causes problems with some clients (such as the Certicom SSL Plus
2454 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2455 disables the coutermeasure allowing Eudora to connect."
2457 38. Exim was not checking that a write() to a log file succeeded. This could
2458 lead to Bad Things if a log got too big, in particular if it hit a file
2459 size limit. Exim now panics and dies if it cannot write to a log file, just
2460 as it does if it cannot open a log file.
2462 39. Modified OS/Makefile-Linux so that it now contains
2464 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2466 The two -D definitions ensure that Exim is compiled with large file
2467 support, which makes it possible to handle log files that are bigger than
2470 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2471 instance) a domain was checked against a named list that involved a lookup,
2472 causing $domain_data to be set, then another domain was checked against the
2473 same list, then the first domain was re-checked, the value of $domain_data
2474 after the final check could be wrong. In particular, if the second check
2475 failed, it could be set empty. This bug probably also applied to
2478 41. The strip_trailing_dot option was not being applied to the address given
2479 with the -f command-line option.
2481 42. The code for reading a message's header from the spool was incrementing
2482 $received_count, but never initializing it. This meant that the value was
2483 incorrect (doubled) while delivering a message in the same process in which
2484 it was received. In the most common configuration of Exim, this never
2485 happens - a fresh exec is done - but it can happen when
2486 deliver_drop_privilege is set.
2488 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2489 it now includes up to 150 characters of the unexpected data in the log
2492 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2493 and building data strings. The size of both of these buffers was 10 000
2494 bytes - far larger than anybody would *ever* want, thought I. Needless to
2495 say, somebody hit the limit. I have increased the maximum line length to
2496 20 000 and the maximum data length of concatenated lines to 100 000. I have
2497 also fixed two bugs, because there was no checking on these buffers. Tsk,
2498 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2499 buffer is too small.
2501 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2502 lsearch lookups. Now it does.
2504 46. When parsing a route_list item in a manualroute router, a fixed-length
2505 buffer was used for the list of hosts. I made this 1024 bytes long,
2506 thinking that nobody would ever have a list of hosts that long. Wrong.
2507 Somebody had a whole pile of complicated expansion conditions, and the
2508 string was silently truncated, leading to an expansion error. It turns out
2509 that it is easier to change to an unlimited length (owing to other changes
2510 that have happened since this code was originally written) than to build
2511 structure for giving a limitation error. The length of the item that
2512 expands into the list of hosts is now unlimited.
2514 47. The lsearch lookup could not handle data where the length of text line was
2515 more than 4095 characters. Such lines were truncated, leading to shortened
2516 data being returned. It should now handle lines of any length.
2518 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2519 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2522 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2523 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2524 no longer happens in comment lines. A list of replacements is now placed
2525 at the head of all of the source files, except those whose only change is
2526 to replace PERL_COMMAND in the very first #! line.
2528 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2529 messages on the queue, with a bottom-up merge sort, using code contributed
2530 by Michael Haardt. This should make operations like -bp somewhat faster on
2531 large queues. It won't affect queue runners, except when queue_run_in_order
2534 51. Installed eximstats 1.31 in the distribution.
2536 52. Added support for SRV lookups to the dnslookup router.
2538 53. If an ACL referred to $message_body or $message_body_end, the value was not
2539 reset for any messages that followed in the same SMTP session.
2541 54. The store-handling optimization for building very long strings was not
2542 differentiating between the different store pools. I don't think this
2543 actually made any difference in practice, but I've tidied it.
2545 55. While running the routers to verify a sender address, $sender_address
2546 was still set to the sender address. This is wrong, because when routing to
2547 send a bounce to the sender, it would be empty. Therefore, I have changed
2548 it so that, while verifying a sender address, $sender_address is set to <>.
2549 (There is no change to what happens when verifying a recipient address.)
2551 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2552 target A or AAAA records (if not already returned) without resetting the
2553 qualify_single or search_parents options of the DNS resolver. These are
2554 inappropriate in this case because the targets of MX and SRV records must
2555 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2556 target that, when qualified, matched something in the local domain. These
2557 two options are now turned off when doing these lookups.
2559 57. It seems that at least some releases of Reiserfs (which does not have the
2560 concept of a fixed number of inodes) returns zero and not -1 for the
2561 number of available inodes. This interacted badly with check_spool_inodes,
2562 which assumed that -1 was the "no such thing" setting. What I have done is
2563 to check that the total number of inodes is greater than zero before doing
2564 the test of how many are available.
2566 58. When a "warn" ACL statement has a log_message modifier, the message is
2567 remembered, and not repeated. This is to avoid a lot of repetition when a
2568 message has many recipients that cause the same warning to be written.
2569 Howewer, Exim was preserving the list of already written lines for an
2570 entire SMTP session, which doesn't seem right. The memory is now reset if a
2571 new message is started.
2573 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2574 debugging output unless log_rewrite was also set.
2576 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2577 of (int)(handle) when we know that handle contains (void *)(-1).
2579 61. The Exim daemon panic-logs an error return when it closes the incoming
2580 connection. However "connection reset by peer" seems to be common, and
2581 isn't really an error worthy of noting specially, so that particular error
2584 62. When Exim is trying to find all the local interfaces, it used to panic and
2585 die if the ioctl to get the interface flags failed. However, it seems that
2586 on at least one OS (Solaris 9) it is possible to have an interface that is
2587 included in the list of interfaces, but for which you get a failure error
2588 for this call. This happens when the interface is not "plumbed" into a
2589 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2590 failure of the "get flags" call assumes that the interface is down.
2592 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2593 makes life easier for people who are doing arithmetic on fields extracted
2594 from dates, where you often get leading zeros that should not be
2595 interpreted as octal.
2597 64. Added qualify_domain to the redirect router, to override the global
2600 65. If a pathologically long header line contained very many addresses (the
2601 report of this problem mentioned 10 000) and each of them was rewritten,
2602 Exim could use up a very large amount of memory. (It kept on making new
2603 copies of the header line as it rewrote, and never released the old ones.)
2604 At the expense of a bit more processing, the header rewriting function has
2605 been changed so that it no longer eats memory in this way.
2607 66. The generation of the Received: header has been moved from the time that a
2608 message starts to be received, to the time that it finishes. The timestamp
2609 in the Received: header should now be very close to that of the <= log
2610 line. There are two side-effects of this change:
2612 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2613 logged header lines no longer include the local Received: line, because
2614 it has not yet been created. The same applies to a copy of the message
2615 that is returned to a non-SMTP sender when a message is rejected.
2617 (b) When a filter file is tested using -bf, no additional Received: header
2618 is added to the test message. After some thought, I decided that this
2621 This change does not affect the value of $received_for. It is still set
2622 after address rewriting, but before local_scan() is called.
2624 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2626 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2627 gave an unhelpful panic error message, and a defer error. I have managed to
2628 change this behaviour so that it now rejects any supplied certificate,
2629 which seems right, as the list of acceptable certificates is empty.
2631 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2632 gave an unhelpful defer error. I have not managed to make this reject any
2633 supplied certificates, but the error message it gives is "no certificate
2634 supplied", which is not helpful.
2636 70. exigrep's output now also includes lines that are not associated with any
2637 message, but which match the given pattern. Implemented by a patch from
2638 Martin Sluka, which also tidied up the Perl a bit.
2640 71. Recipient callout verification, like sender verification, was using <> in
2641 the MAIL FROM command. This isn't really the right thing, since the actual
2642 sender may affect whether the remote host accepts the recipient or not. I
2643 have changed it to use the actual sender in the callout; this means that
2644 the cache record is now keyed on a recipient/sender pair, not just the
2645 recipient address. There doesn't seem to be a real danger of callout loops,
2646 since a callout by the remote host to check the sender would use <>.
2647 [SEE ABOVE: changed after hitting problems.]
2649 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2650 temporary errors. However, in the case of such a code being given after
2651 the end of a data transmission (i.e. after ".") Exim was failing to write
2652 a retry record for the message. (Yes, there was some broken host that was
2653 actually sending 8xx at this point.)
2655 73. An unknown lookup type in a host list could cause Exim to panic-die when
2656 the list was checked. (An example that provoked this was putting <; in the
2657 middle of a list instead of at the start.) If this happened during a DATA
2658 ACL check, a -D file could be left lying around. This kind of configuration
2659 error no longer causes Exim to die; instead it causes a defer errror. The
2660 incident is still logged to the main and panic logs.
2662 74. Buglet left over from Exim 3 conversion. The message "too many messages
2663 in one connection" was written to the rejectlog but not the mainlog, except
2664 when address rewriting (yes!) was being logged.
2666 75. Added write_rejectlog option.
2668 76. When a system filter was run not as root (that is, when system_filter_user
2669 was set), the values of the $n variables were not being returned to the
2670 main process; thus, they were not subsequently available in the $sn
2673 77. Added +return_path_on_delivery log selector.
2675 78. A connection timeout was being treated differently from recipients deferred
2676 when testing hosts_max_try with a message that was older than the host's
2677 retry timeout. (The host should not be counted, thus allowing all hosts to
2678 be tried at least once before bouncing.) This may have been the cause of an
2679 occasionally reported bug whereby a message would remain on the queue
2680 longer than the retry timeout, but would be bounced if a delivery was
2681 forced. I say "may" because I never totally pinned down the problem;
2682 setting up timeout/retry tests is difficult. See also the next item.
2684 79. The ultimate address timeout was not being applied to errors that involved
2685 a combination of host plus message (for example, a timeout on a MAIL
2686 command). When an address resolved to a number of possible hosts, and they
2687 were not all tried for each delivery (e.g. because of hosts_max_try), a
2688 message could remain on the queue longer than the retry timeout.
2690 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2693 81. Fixed an obscure SMTP outgoing bug which required at least the following
2694 conditions: (a) there was another message waiting for the same server;
2695 (b) the server returned 5xx to all RCPT commands in the first message so
2696 that the message was not completed; (c) the server dropped the connection
2697 or gave a negative response to the RSET that Exim sends to abort the
2698 transaction. The observed case was a dropped connection after DATA that had
2699 been sent in pipelining mode. That is, the server had advertised PIPELINING
2700 but was not implementing it correctly. The effect of the bug was incorrect
2701 behaviour, such as trying another host, and this could lead to a crash.
2707 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2708 and daemon.c were passed as pointers to ints; they should have been
2709 pointers to socklen_t variables (which are typically unsigned ints).
2711 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2714 3. Fixed a really odd bug that affected only the testing scheme; patching a
2715 certain fixed string in the binary changed the value of another string that
2716 happened to be identical to the end of the original first string.
2718 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2719 name", it returns that address as the IP address. On some operating
2720 systems (e.g. Solaris), it also passes back the IP address string as the
2721 "host name". However, on others (e.g. Linux), it passes back an empty
2722 string. Exim wasn't checking for this, and was changing the host name to an
2723 empty string, assuming it had been canonicized.
2725 5. Although rare, it is permitted to have more than one PTR record for a given
2726 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2727 all the names associated with an address, because they do in Solaris.
2728 However, it seems that they do not in Linux for data that comes from the
2729 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2730 I found this out when I moved to a new Linux workstation and tried to run
2731 the Exim test suite.
2733 To get round this problem I have changed the code so that it now does its
2734 own call to the DNS to look up PTR records when searching for a host name.
2735 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2736 addresses that are only in /etc/hosts are still found.
2738 This behaviour is, however, controlled by an option called host_lookup_
2739 order, which defaults to "bydns:byaddr". If people want to use the other
2740 order, or indeed, just use one or the other means of lookup, they can
2741 specify it in this variable.
2743 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2744 some operating systems, this comes back from gethostbyaddr() as an empty
2745 string, and this is what Exim used to test for. However, it seems that in
2746 other systems, "." is yielded. Exim now tests for this case too.
2748 7. The values of check_spool_space and check_log_space are now held internally
2749 as a number of kilobytes instead of an absolute number of bytes. If a
2750 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2751 kilobyte. This means that much larger values can be stored.
2753 8. Exim monitor: an attempt to get the action menu when not actually pointing
2754 at a message produces an empty menu entitled "No message selected". This
2755 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2756 no entries in it ("Shell widget menu has zero width and/or height"). So I
2757 have added a single, blank menu entry in this case.
2759 9. Added ${quote_local_part.
2761 10. MIME decoding is now applied to the contents of Subject: header lines when
2764 11. Now that a reference to $sender_host_address automatically causes a reverse
2765 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2766 host lookup before query-style lookups in lists that might use this
2767 variable. This has therefore been abolished, and the "net-" prefix is no
2768 longer necessary for query-style lookups.
2770 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2771 have been a typo for LFLAGS, so it has been changed.
2773 13. The install script calls Exim with "-C /dev/null" in order to find the
2774 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2775 to be output. Howeve, since Exim outputs its version number before the
2776 error, it didn't break the script. It just looked ugly. I fixed this by
2777 always allowing "-C /dev/null" if the caller is root.
2779 14. Ignore overlarge ACL variable number when reading spool file - insurance
2780 against a later release with more variables having written the file.
2782 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2783 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2784 this, as well as the form without the "IPv6" on the front.
2786 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2787 OS/Makefile-Darwin file.
2789 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2790 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2792 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2793 variable. However, some OS use ints for inodes. I've added cast to long int
2794 to get rid of the compiler warning.
2796 19. I had forgotten to lock out "/../" in configuration file names when
2797 ALT_CONFIG_PREFIX was set.
2799 20. Routers used for verification do not need to specify transports. However,
2800 if such a router generated a host list, and callout was configured, Exim
2801 crashed, because it could not find a port number from the (non-existent)
2802 transport. It now assumes port 25 in this circumstance.
2804 21. Added the -t option to exigrep.
2806 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2807 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2808 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2809 them is set, LOOKUP_LSEARCH is forced.
2811 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2813 24. Added sender and host information to the "rejected by local_scan()" log
2814 line; previously there was no indication of these.
2816 25. Added .include_if_exists.
2818 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2819 Linux. It turns out that not all file systems support this. Apparently some
2820 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2821 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2822 not supported on the file descriptor, is now ignored when Exim is trying to
2823 sync a directory. This applies only to Linux.
2825 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2827 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2828 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2831 29. Added support for maildirsize files from supplied patch (modified a bit).
2833 30. The use of :fail: followed by an empty string could lead Exim to respond to
2834 sender verification failures with (e.g.):
2836 550 Verification failed for <xxx>
2837 550 Sender verify failed
2839 where the first response line was missing the '-' that indicates it is not
2840 the final line of the response.
2842 31. The loop for finding the name of the user that called Exim had a hardwired
2843 limit of 10; it now uses the value of finduser_retries, which is used for
2844 all other user lookups.
2846 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2849 33. Exim was neglecting to zero errno before one call of strtol() when
2850 expanding a string and expecting an integer value. On some systems this
2851 resulted in spurious "integer overflow" errors. Also, it was casting the
2852 result into an int without checking.
2854 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2855 did not work. The code looks as if it has *never* worked, though it appears
2856 to have been documented since at least releast 1.62. I have made it work.
2858 35. The "timeout_DNS" error in retry rules, also documented since at least
2859 1.62, also never worked. As it isn't clear exactly what this means, and
2860 clearly it isn't a major issue, I have abolished the feature by treating it
2861 as "timeout", and writing a warning to the main and panic logs.
2863 36. The display of retry rules for -brt wasn't always showing the error code
2866 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2867 timeout_connect_A, timeout_connect_MX.
2869 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2870 to the empty sender.
2872 39. The daemon was not analysing the content of -oX till after it had closed
2873 stderr and disconnected from the controlling terminal. This meant that any
2874 syntax errors were only noted on the panic log, and the return code from
2875 the command was 0. By re-arranging the code a little, I've made the
2876 decoding happen first, so such errors now appear on stderr, and the return
2877 code is 1. However, the actual setting up of the sockets still happens in
2878 the disconnected process, so errors there are still only recorded on the
2881 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2882 connections (as happens on some IP stacks) was logged at start up time as
2883 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2884 it from "IPv6 and IPv4", which means that two separate sockets are being
2887 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2888 says whether AF_INET or AF_INET6 was passed as an argument.
2890 42. Exiwhat output was messed up when time zones were included in log
2893 43. Exiwhat now gives more information about the daemon's listening ports,
2894 and whether -tls-on-connect was used.
2896 44. The "port" option of the smtp transport is now expanded.
2898 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2899 silently ignored. Now an error message is written to the main and panic
2902 46. There's a new ACL modifier called "logwrite" which writes to a log file
2903 as soon as it is encountered.
2905 47. Added $local_user_uid and $local_user_gid at routing time.
2907 48. Exim crashed when trying to verify a sender address that was being
2910 49. Exim was recognizing only a space character after ".include". It now also
2911 recognizes a tab character.
2913 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2914 extracting the relevant information from the specification. The man page no
2915 longer contains scrambled data for the -d option, and I've added a section
2916 at the front about calling Exim under different names.
2918 51. Added "extra_headers" argument to the "mail" command in filter files.
2920 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2923 53. Installed eximstats 1.29.
2925 54. Added transport_filter_timeout as a generic transport option.
2927 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2928 Cc: header lines. This was required by RFC 822, but it not required by RFC
2931 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2932 incoming messages that did not have them. Now it does so only if the
2933 message originates locally, that is, if there is no associated remote host
2934 address. When Resent- header lines are present, this applies to the Resent-
2935 lines rather than the non-Resent- lines.
2937 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2938 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2940 58. Messages for configuration errors now include the name of the main
2941 configuration file - useful now that there may be more than one file in a
2942 list (.included file names were always shown).
2944 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2945 for those rare installations that do not start the daemon as root or run it
2946 setuid root. I've cut out the call to initgroups() if the daemon is not
2949 60. The Exim user and group can now be bound into the binary as text strings
2950 that are looked up at the start of Exim's processing.
2952 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2954 62. Added $mailstore_basename variable.
2956 63. Installed patch to sieve.c from Michael Haardt.
2958 64. When Exim failed to open the panic log after failing to open the main log,
2959 the original message it was trying to log was written to stderr and debug
2960 output, but if they were not available (the usual case in production), it
2961 was lost. Now it is written to syslog before the two lines that record the
2962 failures to open the logs.
2964 65. Users' Exim filters run in subprocesses under the user's uid. It is
2965 possible for a "deliver" command or an alias in a "personal" command to
2966 provoke an address rewrite. If logging of address rewriting is configured,
2967 this fails because the process is not running as root or exim. There may be
2968 a better way of dealing with this, but for the moment (because 4.30 needs
2969 to be released), I have disabled address rewrite logging when running a
2970 filter in a non-root, non-exim process.
2976 1. The buildconfig auxiliary program wasn't quoting the value set for
2977 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2978 not defined. This bug was masked in 4.22 by the effect that was fixed in
2981 2. Some messages that were rejected after a message id was allocated were
2982 shown as "incomplete" by exigrep. It no longer does this for messages that
2983 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2985 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2986 have allow_domain_literals set, the ID did not get logged in the <= line.
2987 Domain literals are now always recognized in Message-ID: header lines.
2989 4. The first argument for a ${extract expansion item is the key name or field
2990 number. Leading and trailing spaces in this item were not being ignored,
2991 causing some misleading effects.
2993 5. When deliver_drop_privilege was set, single queue runner processes started
2994 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2995 same command in the process it spins off) were not dropping privilege.
2997 6. When the daemon running as "exim" started a queue runner, it always
2998 re-executed Exim in the spun-off process. This is a waste of effort when
2999 deliver_drop_privilege is set. The new process now just calls the
3000 queue-runner function directly.
3006 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3009 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3010 ignored. Though the use of -oP was forcing the writing of a pid file, it
3011 was always written to the default place.
3013 3. If the message "no IP address found for host xxxx" is generated during
3014 incoming verification, it is now followed by identification of the incoming
3015 connection (so you can more easily find what provoked it).
3017 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3019 5. Added some features to "harden" Exim a bit more against certain attacks:
3021 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3022 be put in Local/Makefile. This is like the never_users runtime option,
3023 but it cannot be overridden. The default setting is "root".
3025 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3026 prefix string with which any file named in a -C command line option
3029 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3030 is retained for -C and -D only if the caller of Exim is root. Without
3031 it, the exim user may also use -C and -D and retain privilege.
3033 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3034 command line option is disabled.
3036 6. Macro names set by the -D option must start with an upper case letter, just
3037 like macro names defined in the configuration file.
3039 7. Added "dereference=" facility to LDAP.
3041 8. Two instances of the typo "uknown" in the source files are fixed.
3043 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3044 the Configure-Makefile script screwed up while processing it.
3046 10. Incorporated PCRE 4.4.
3048 11. The SMTP synchronization check was not operating right at the start of an
3049 SMTP session. For example, it could not catch a HELO sent before the client
3050 waited for the greeting. There is now a check for outstanding input at the
3051 point when the greeting is written. Because of the duplex, asynchronous
3052 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3053 way, but not yet received, when the check is performed.
3055 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3056 on TCP/IP sockets, because this apparently causes some broken clients to
3059 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3060 unchanged) from the Cygwin maintainer.
3062 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3063 was supported instead of testing for mbx. Effectively a typo.
3065 15. The spa authenticator server code was not checking that the input it
3066 received was valid base64.
3068 16. The debug output line for the "set" modifier in ACLs was not showing the
3069 name of the variable that was being set.
3071 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3073 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3074 lookup on its own. Something else was needed to trigger the lookup. For
3075 example, a match in host_lookup or the need for a host name in a host list.
3076 Now, if $sender_host_name is referenced and the host name has not yet been
3077 looked up, a lookup is performed. If the lookup fails, the variable remains
3078 empty, and $host_lookup_failed is set to "1".
3080 19. Added "eqi" as a case-independent comparison operator.
3082 20. The saslauthd authentication condition could segfault if neither service
3083 nor realm was specified.
3085 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3086 error message that was logged was misleading, and incoming SMTP
3087 connections were dropped. The message is now more accurate, and temporary
3088 errors are given to SMTP connections.
3090 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3091 (except RSET) with a 421 error, until QUIT is received. However, it was
3092 failing to send a response to QUIT.
3094 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3095 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3096 ACL is now run after the helo_try_verify_hosts code.
3098 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3099 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3100 case-independent, so other case variants are also recognized). Apparently
3101 some systems use these upper case variants.
3103 25. If more than two messages were waiting for the same host, and a transport
3104 filter was specified for the transport, Exim sent two messages over the
3105 same TCP/IP connection, and then failed with "socket operation on non-
3106 socket" when it tried to send the third.
3108 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3110 27. The extern definition of crypt16() in expand.c was not being excluded when
3111 the OS had its own crypt16() function.
3113 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3114 as a preferred synonym for return_size_limit, both as an option and as an
3117 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3119 30. Changed the default configuration ACL to relax the local part checking rule
3120 for addresses that are not in any local domains. For these addresses,
3121 slashes and pipe symbols are allowed within local parts, but the sequence
3122 /../ is explicitly forbidden.
3124 31. SPA server authentication was not clearing the challenge buffer before
3127 32. log_message in a "warn" ACL statement was writing to the reject log as
3128 well as to the main log, which contradicts the documentation and doesn't
3129 seem right (because no rejection is happening). So I have stopped it.
3131 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3132 However, I am unable to do any testing of this.
3134 34. Fixed an infelicity in the appendfile transport. When checking directories
3135 for a mailbox, to see if any needed to be created, it was accidentally
3136 using path names with one or more superfluous leading slashes; tracing
3137 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3139 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3140 contents are added to the log line that is written for every discarded
3141 recipient. (Previously a log_message setting was ignored.)
3143 36. The ${quote: operator now quotes the string if it is empty.
3145 37. The install script runs exim in order to find its version number. If for
3146 some reason other than non-existence or emptiness, which it checks, it
3147 could not run './exim', it was installing it with an empty version number,
3148 i.e. as "exim-". This error state is now caught, and the installation is
3151 38. An argument was missing from the function that creates an error message
3152 when Exim fails to connect to the socket for saslauthd authentication.
3153 This could cause Exim to crash, or give a corrupted message.
3155 39. Added isip, isip4, and isip6 to ${if conditions.
3157 40. The ACL variables $acl_xx are now saved with the message, and can be
3158 accessed later in routers, transports, and filters.
3160 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3161 strings in the file are not string-expanded.
3163 42. If a MAIL command specified a SIZE value that was too large to fit into an
3164 int variable, the check against message_size_limit failed. Such values are
3165 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3166 day this will have to be increased, but I don't think I want to be around
3167 when emails are that large.
3174 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3175 iconv() is not standard in FreeBSD.
3177 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3178 IPv6 enabled. The observed symptom was a segmentation fault on return from
3179 the function os_common_find_running_interfaces() in src/os.c.
3181 3. In the check_special_case() function in daemon.c I had used "errno" as an
3182 argument name, which causes warnings on some systems. This was basically a
3183 typo, since it was named "eno" in the comments!
3185 4. The code that waits for the clock to tick (at a resolution of some fraction
3186 of a second) so as to ensure message-id uniqueness was always waiting for
3187 at least one whole tick, when it could have waited for less. [This is
3188 almost certainly not relevant at current processor speeds, where it is
3189 unlikely to ever wait at all. But we try to future-proof.]
3191 5. The function that sleeps for a time interval that includes fractions of a
3192 second contained a race. It did not block SIGALRM between setting the
3193 timer, and suspending (a couple of lines later). If the interval was short
3194 and the sigsuspend() was delayed until after it had expired, the suspension
3195 never ended. On busy systems this could lead to processes getting stuck for
3198 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3199 process, before it forks any delivery processes. The open lookup caching
3200 mechanism meant that the open file or database connection was passed into
3201 the delivery process. The problem was that delivery processes always tidy
3202 up cached lookup data. This could cause a problem for the next delivery
3203 process started by the queue runner, because the external queue runner
3204 process does not know about the closure. So the next delivery process
3205 still has data in the lookup cache. In the case of a file lookup, there was
3206 no problem because closing a file descriptor in a subprocess doesn't affect
3207 the parent. However, if the lookup was caching a connection to a database,
3208 the connection was closed, and the second delivery process was likely to
3209 see errors such as "PGSQL: query failed: server closed the connection
3210 unexpectedly". The problem has been fixed by closing all cached lookups
3211 in a queue runner before running a delivery process.
3213 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3214 seem to have the "const" qualifier which it has on other OS. I've
3217 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3218 *of the same type* (client or server) with the same public name that an
3219 error should be diagnosed.
3221 9. When Exim looked up a host name for an IP address, but failed to find the
3222 original IP address when looking up the host name (a safety check), it
3223 output the message "<ip address> does not match any IP for NULL", which was
3224 confusing, to say the least. The bug was that the host name should have
3225 appeared instead of "NULL".
3227 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3228 user that is built into the binary, and the -C or -D options is used, root
3229 privilege is dropped before the configuration file is read. In addition,
3230 logging is switched to stderr instead of the normal log files. If the
3231 configuration then re-defines the Exim user, the unprivileged environment
3232 is probably not what is expected, so Exim logs a panic warning message (but
3235 However, if deliver_drop_privilege is set, the unprivileged state may well
3236 be exactly what is intended, so the warning has been cut out in that case,
3237 and Exim is allowed to try to write to its normal log files.
3243 1. smtp_return_error_details was not giving details for temporary sender
3244 or receiver verification errors.
3246 2. Diagnose a configuration error if two authenticators have the same public
3249 3. Exim used not to create the message log file for a message until the first
3250 delivery attempt. This could be confusing when incoming messages were held
3251 for policy or load reasons. The message log file is now created at the time
3252 the message is received, and an initial "Received" line is written to it.
3254 4. The automatically generated man page for command line options had a minor
3255 bug that caused no ill effects; however, a more serious problem was that
3256 the procedure for building the man page automatically didn't always
3257 operate. Consequently, release 4.20 contains an out-of-date version. This
3258 shouldn't happen again.
3260 5. When building Exim with embedded Perl support, the script that builds the
3261 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3262 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3264 6. The freeze_tell option was not being used for messages that were frozen on
3265 arrival, either by an ACL or by local_scan().
3267 7. Added the smtp_incomplete_transaction log selector.
3269 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3270 was accepting AUTH without a new EHLO.
3272 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3273 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3276 10. Logging of TCP/IP connections (when configured) now happens in the main
3277 daemon process instead of the child process, so that the TCP/IP connection
3278 count is more accurate (but it can never be perfect).
3280 11. The use of "drop" in a nested ACL was not being handled correctly in the
3281 outer ACL. Now, if condition failure induced by the nested "drop" causes
3282 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3283 or "require"), the connection is dropped.
3285 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3286 that yield "discard" can now be used with an "accept" or a "discard" verb,
3287 but an error is generated for any others (because I can't see a useful way
3288 to define what should happen).
3290 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3291 are now processed in the same way as lines in the Exim configuration file.
3292 In particular, continuation lines are supported.
3294 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3296 15. Added -ti meaning -t -i.
3298 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3299 domains, and warn by logging if others are found.
3301 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3302 returned from GIFCONF when Exim is trying to find the list of interfaces on
3303 a host. The code in os.c has been modified to copy each ifreq to an aligned
3304 structure in all cases.
3306 Also, in some cases, the returned ifreq's were being copied to a 'struct
3307 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3308 means the last couple of bytes of an IPv6 address could be chopped if the
3309 ifreq contained only a normal sockaddr (14 bytes storage).
3311 18. Named domain lists were not supported in the hosts_treat_as_local option.
3312 An entry such as +xxxx was not recognized, and was treated as a literal
3315 19. Ensure that header lines added by a DATA ACL are included in the reject log
3316 if the ACL subsequently rejects the message.
3318 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3319 MD5 (which is deprecated).
3321 21. When testing a filter file using -bf, Exim was writing a message when it
3322 took the sender from a "From " line in the message, but it was not doing so
3323 when it took $return_path from a Return-Path: header line. It now does.
3325 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3326 with a valid header line field name (a series of printing characters
3327 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3329 23. Changed "disc" in the source to "disk" to conform to the documentation and
3330 the book and for uniformity.
3332 24. Ignore Sendmail's -Ooption=value command line item.
3334 25. When execve() failed while trying to run a command in a pipe transport,
3335 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3336 could be confused with a return value of 69 from the command itself. This
3337 has been changed to 127, the value the shell returns if it is asked to run
3338 a non-existent command. The wording for the related log line suggests a
3339 non-existent command as the problem.
3341 26. If received_header_text expands to an empty string, do not add a Received:
3342 header line to the message. (Well, it adds a token one on the spool, but
3343 marks it "old" so that it doesn't get used or transmitted.)
3345 27. Installed eximstats 1.28 (addition of -nt option).
3347 28. There was no check for failure on the call to getsockname() in the daemon
3348 code. This can fail if there is a shortage of resources on the system, with
3349 ENOMEM, for example. A temporary error is now given on failure.
3351 29. Contrary to the C standard, it seems that in some environments, the
3352 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3353 program. Exim now does this explicitly; it affects the formatting of
3354 timestamps using strftime().
3356 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3357 complaints. I've now initialized all the variables, to avoid this.
3359 32. Header lines added by a system filter were not being "seen" during
3360 transport-time rewrites.
3362 33. The info_callback() function passed to OpenSSL is set up with type void
3363 (*)(SSL *, int, int), as described somewhere. However, when calling the
3364 function (actually a macro) that sets it up, the type void(*)() is
3365 expected. I've put in a cast to prevent warnings from picky compilers.
3367 34. If a DNS black list lookup found a CNAME record, but there were no A
3368 records associated with the domain it pointed at, Exim crashed.
3370 35. If a DNS black list lookup returned more than one A record, Exim ignored
3371 all but the first. It now scans all returned addresses if a particular IP
3372 value is being sought. In this situation, the contents of the
3373 $dnslist_value variable are a list of all the addresses, separated by a
3376 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3377 to accept a host name and all its aliases if the forward lookup for any of
3378 them yielded the IP address of the incoming connection. Now it accepts only
3379 those names whose forward lookup yields the correct IP address. Any other
3380 names are discarded. This closes a loophole whereby a rogue DNS
3381 administrator could create reverse DNS records to break through a
3382 wildcarded host restriction in an ACL.
3384 37. If a user filter or a system filter that ran in a subprocess used any of
3385 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3386 the wrong values were passed to the pipe command ($thisaddress had the
3387 value of $0, $0 had the value of $1, etc). This bug was introduced by
3388 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3390 38. Improved the line breaking for long SMTP error messages from ACLs.
3391 Previously, if there was no break point between 40 and 75 characters, Exim
3392 left the rest of the message alone. Two changes have been made: (a) I've
3393 reduced the minimum length to 35 characters; (b) if it can't find a break
3394 point between 35 and 75 characters, it looks ahead and uses the first one
3395 that it finds. This may give the occasional overlong line, but at least the
3396 remaining text gets split now.
3398 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3399 file descriptors might be low, and that setting 1000 would always raise it.
3400 It turns out that in some environments, the limit is already over 1000 and
3401 that lowering it causes trouble. So now Exim takes care not to decrease it.
3403 40. When delivering a message, the value of $return_path is set to $sender_
3404 address at the start of routing (routers may change the value). By an
3405 oversight, this default was not being set up when an address was tested by
3406 -bt or -bv, which affected the outcome if any router or filter referred to
3409 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3410 log only when "message" or "log_message" are set. However, if one of the
3411 conditions was an address verification, or a call to a nested ACL, the
3412 messages generated by the underlying test were being passed through. This
3413 no longer happens. The underlying message is available in $acl_verify_
3414 message for both "message" and "log_message" expansions, so it can be
3415 passed through if needed.
3417 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3418 new expansion $bh_ to give the encoded byte string without charset
3419 translation. Translation happens only if iconv() is available; HAVE_ICONV
3420 indicates this at build time. HEADERS_CHARSET gives the charset to
3421 translate to; headers_charset can change it in the configuration, and
3422 "headers charset" can change it in an individual filter file.
3424 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3425 that creates RFC 2047 encoded "words" labels them as that charset instead
3426 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3427 expansion operator; (ii) when Exim creates a From: line for a local
3428 message; (iii) when a header line is rewritten to include a "phrase" part.
3430 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3431 buggy, causing it to skip the first lines of messages whose message ID
3432 ended in 'D'. This would not have bitten before Exim release 4.14, because
3433 message IDs were unlikely to end in 'D' before then. The effect was to have
3434 incorrect size information for certain domains.
3436 45. #include "config.h" was missing at the start of the crypt16.c module. This
3437 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3440 46. If there was a timeout during a "random" callout check, Exim treated it as
3441 a failure of the random address, and carried on sending RSET and the real
3442 address. If the delay was just some slowness somewhere, the response to the
3443 original RCPT would be taken as a response to RSET and so on, causing
3444 mayhem of various kinds.
3446 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3447 when I implemented it. It didn't allow for the fact that some option values
3448 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3449 the right test for positive values.
3451 48. Domain names in DNS records are case-independent. Exim always looks them up
3452 in lower case. Some resolvers return domain names in exactly the case they
3453 appear in the zone file, that is, they may contain uppercase letters. Not
3454 all resolvers do this - some return always lower case. Exim was treating a
3455 change of case by a resolver as a change of domain, similar to a widening
3456 of a domain abbreviation. This triggered its re-routing code and so it was
3457 trying to route what was effectively the same domain again. This normally
3458 caused routing to fail (because the router wouldn't handle the domain
3459 twice). Now Exim checks for this case specially, and just changes the
3460 casing of the domain that it ultimately uses when it transmits the message
3463 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3466 50. If a filter generated a file delivery with a non-absolute name (possible if
3467 no home directory exists for the router), the forbid_file option was not
3470 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3471 the existing equality matching.
3473 52. Exim was using ints instead of ino_t variables in some places where it was
3474 dealing with inode numbers.
3476 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3477 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3478 called TMPDIR, and if it finds it is different, it changes its value.
3480 54. The smtp_printf() function is now made available to local_scan() so
3481 additional output lines can be written before returning. There is also an
3482 smtp_fflush() function to enable the detection of a dropped connection.
3483 The variables smtp_input and smtp_batched_input are exported to
3486 55. Changed the default runtime configuration: the message "Unknown user"
3487 has been removed from the ACL, and instead placed on the localuser router,
3488 using the cannot_route_message feature. This means that any verification
3489 failures that generate their own messages won't get overridden. Similarly,
3490 the "Unrouteable address" message that was in the ACL for unverifiable
3491 relay addresses has also been removed.
3493 56. Added hosts_avoid_esmtp to the smtp transport.
3495 57. The exicyclog script was not checking for the esoteric option
3496 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3497 will work only if exicyclog is run under the appropriate euid.
3499 58. Following a discussion on the list, the rules by which Exim recognises line
3500 endings on incoming messages have been changed. The -dropcr and drop_cr
3501 options are now no-ops, retained only for backwards compatibility. The
3502 following line terminators are recognized: LF CRLF CR. However, special
3503 processing applies to CR:
3505 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3506 nor a local message in the state where . is a terminator.
3508 (ii) If a bare CR is encountered in a header line, an extra space is added
3509 after the line terminator so as not to end the header. The reasoning
3510 behind this is that bare CRs in header lines are most likely either
3511 to be mistakes, or people trying to play silly games.
3513 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3514 was being incorrectly given as 18 bytes larger than it should have been.
3515 This is a VOB (very old bug).
3517 60. This may never have affected anything current, but just in case it has:
3518 When the local host is found other than at the start of a list of hosts,
3519 the local host, those with the same MX, and any that follow, are discarded.
3520 When the list in question was part of a longer list of hosts, the following
3521 hosts (not currently being processed) were also being discarded. This no
3522 longer happens. I'm not sure if this situation could ever has previously
3525 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3528 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3529 header. This is something that is recommended in a new Internet Draft, and
3530 is something that is documented as being done by Sendmail. There are two
3531 possible values. For messages generated by the autoreply transport, Exim
3534 Auto-Submitted: auto-replied
3536 whereas for all other generated messages (e.g. bounces) it adds
3538 Auto-Submitted: auto-generated
3540 63. The "personal" condition in filters now includes a test for the
3541 Auto-Submitted: header. If it contains the string "auto-" the message it
3542 not considered personal.
3544 64. Added rcpt_include_affixes as a generic transport option.
3546 65. Added queue_only_override (default true).
3548 66. Added the syslog_duplication option.
3550 67. If what should have been the first header line of a message consisted of
3551 a space followed by a colon, Exim was mis-interpreting it as a header line.
3552 It isn't of course - it is syntactically invalid and should therefore be
3553 treated as the start of the message body. The misbehaviour could have
3554 caused a number of strange effects, including loss of data in subsequent
3555 header lines, and spool format errors.
3557 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3558 client host had authenticated. This control can now be exercised by an ACL
3559 for more flexibility.
3561 69. By default, callouts do not happen when testing with -bh. There is now a
3562 variant, -bhc, which does actually run the callout code, including
3563 consulting and updating the callout cache.
3565 70. Added support for saslauthd authentication, courtesy of Alexander
3568 71. If statvfs() failed on the spool or log directories while checking their
3569 size for availability, Exim confusingly gave the error "space shortage".
3570 Furthermore, in debugging mode it crashed with a floating point exception.
3571 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3572 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3573 serious problem, Exim now writes to the main and panic logs when this
3574 happens, with details of the failure. It then refuses to accept the
3575 incoming message, giving the message "spool directory problem" or "log
3576 directory problem" with a 421 code for SMTP messages.
3578 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3579 0, 1, and 2 exist, because some OS complain for execs without them (see
3580 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3581 descriptors. However, the code omitted to check that the open succeeded,
3582 causing mysterious errors if for some reason the permissions on /dev/null
3583 got screwed. Now Exim writes a message to the main and panic logs, and
3584 bombs out if it can't open /dev/null.
3586 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3587 interact so that it is all more flexible. It is supposed to remain
3588 backwards compatible. Also added extra_local_interfaces.
3590 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3591 to bomb out with an assertion failure - to the client this appears as a
3592 connection drop. This problem occurs in the part of the code that was taken
3593 from the Samba project. Fortunately, the assertion is in a very simple
3594 function, so I have fixed this by reproducing the function inline in the
3595 one place where it is called, and arranging for authentication to fail
3596 instead of killing the process with assert().
3598 75. The SPA client code was not working when the server requested OEM rather
3599 than Unicode encoding.
3601 76. Added code to make require_files with a specific uid setting more usable in
3602 the case where statting the file as root fails - usually a non-root-mounted
3603 NFS file system. When this happens and the failure is EACCES, Exim now
3604 forks a subprocess and does the per-uid checking as the relevant uid.
3606 77. Added process_log_path.
3608 78. If log_file_path was not explicitly set, a setting of check_log_space or
3609 check_log_inodes was ignored.
3611 79. If a space check for the spool or log partitions fails, the incident is now
3612 logged. Of course, in the latter case the data may get lost...
3614 80. Added the %p formatting code to string_format() so that it can be used to
3615 print addresses in debug_print(). Adjusted all the address printing in the
3616 debugging in store.c to use %p rather than %d.
3618 81. There was a concern that a line of code in smtp_in.c could overflow a
3619 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3620 initially expressed, the concern was not well-founded, because trailing
3621 spaces are removed early. However, if the trailing spaces were followed by
3622 a NULL, they did not get removed, so the overflow was possible. Two fixes
3625 (a) I re-wrote the offending code in a cleaner fashion.
3626 (b) If an incoming SMTP command contains a NULL character, it is rejected
3629 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3630 runs initgroups(), so that if the Exim user is in any additional groups,
3631 they will be used during message reception.
3637 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob