1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.280 2005/12/19 12:25:21 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
67 PH/01 Two changes to the default runtime configuration:
69 (1) Move the checks for relay_from_hosts and authenticated clients from
70 after to before the (commented out) DNS black list checks.
72 (2) Add control=submission to the relay_from_hosts and authenticated
73 clients checks, on the grounds that messages accepted by these
74 statements are most likely to be submissions.
76 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
78 (1) Generate an error if the third argument for the ${prvs expansion is
81 (2) Treat a missing third argument of ${prvscheck as if it were an empty
84 (3) Reset the variables that are obtained from the first argument of
85 ${prvscheck and used in the second argument before leaving the code,
86 because their memory is reclaimed, so using them afterwards may do
89 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
90 one (it's much easier than Tom thought :-).
92 (5) Because of (4), we can now allow for the use of $prvscheck_result
93 inside the third argument.
95 PH/03 For some reason, the default setting of PATH when running a command from
96 a pipe transport was just "/usr/bin". I have changed it to
99 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
100 anything to be listed in the output from -bV.
102 PH/05 When a filter generated an autoreply, the entire To: header line was
103 quoted in the delivery log line, like this:
105 => >A.N.Other <ano@some.domain> <original@ddress> ...
107 This has been changed so that it extracts the operative address. There
108 may be more than one such address. If so, they are comma-separated, like
111 => >ano@some.domain,ona@other.domain <original@ddress> ...
113 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
114 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
115 not being looked up in the rDNS to get a host name, Exim was showing the
116 IP address twice in Received: lines, even though the IP addresses were
117 identical. For example:
119 Received: from [1.2.3.4] (helo=[1.2.3.4])
121 However, if the real host name was known, it was omitting the HELO data
122 if it matched the actual IP address. This has been tidied up so that it
123 doesn't show the same IP address twice.
125 PH/07 When both +timestamp and +memory debugging was on, the value given by
126 $tod_xxx expansions could be wrong, because the tod_stamp() function was
127 called by the debug printing, thereby overwriting the timestamp buffer.
128 Debugging no longer uses the tod_stamp() function when +timestamp is set.
130 PH/08 When the original message was included in an autoreply transport, it
131 always said "this is a copy of the message, including all the headers",
132 even if body_only or headers_only was set. It now gives an appropriate
135 PH/09 Applied a patch from the Sieve maintainer which:
137 o fixes some comments
138 o adds the (disabled) notify extension core
139 o adds some debug output for the result of if/elsif tests
140 o points to the current vacation draft in the documentation
141 and documents the missing references header update
145 o fixes a bug in processing the envelope test (when testing
146 multiple envelope elements, the last element determinted the
149 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
150 Electronic Mail") by including:
152 Auto-submitted: auto-generated
154 in the messages that it generates (bounce messages and others, such as
155 warnings). In the case of bounce messages for non-SMTP mesages, there was
156 also a typo: it was using "Auto_submitted" (underscore instead of
157 hyphen). Since every message generated by Exim is necessarily in response
158 to another message, thes have all been changed to:
160 Auto-Submitted: auto-replied
162 in accordance with these statements in the RFC:
164 The auto-replied keyword:
166 - SHOULD be used on messages sent in direct response to another
167 message by an automatic process,
169 - MUST NOT be used on manually-generated messages,
171 - MAY be used on Delivery Status Notifications (DSNs) and Message
172 Disposition Notifications (MDNs),
174 - MUST NOT be used on messages generated by automatic or periodic
175 processes, except for messages which are automatic responses to
178 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
179 to the default Received: header definition.
181 PH/12 Added log selector acl_warn_skipped (default on).
183 PH/13 After a successful wildlsearch lookup, discard the values of numeric
184 variables because (a) they are in the wrong storage pool and (b) even if
185 they were copied, it wouldn't work properly because of the caching.
187 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
188 checking when decoding. Apparently there are clients that generate
189 overlong encoded strings. Why am I not surprised?
191 PH/15 If the first argument of "${if match_address" was not empty, but did not
192 contain an "@" character, Exim crashed. Now it writes a panic log message
193 and treats the condition as false.
195 PH/16 In autoreply, treat an empty string for "once" the same as unset.
197 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
198 extension "envelope-auth". The code is finished and in agreement with
199 other implementations, but there is no documentation so far and in fact,
200 nobody wrote the draft yet. This extension is currently #undef'ed, thus
201 not changing the active code.
203 Print executed "if" and "elsif" statements when debugging is used. This
204 helps a great deal to understand what a filter does.
206 Document more things not specified clearly in RFC3028. I had all this
207 sorted out, when out of a sudden new issues came to my mind. Oops."
209 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
212 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
213 canonical form (as documented). However, after a host name lookup from
214 the IP address, check_host() was doing a simple string comparison with
215 addresses acquired from the DNS when checking that the found name did
216 have the original IP as one of its addresses. Since any found IPv6
217 addresses are likely to be in abbreviated form, the comparison could
218 fail. Luckily, there already exists a function for doing the comparison
219 by converting both addresses to binary, so now that is used instead of
222 PH/20 There was another similar case to PH/19, when a complete host name was
223 given in a host list; looking up its IP address could give an abbreviated
224 form, whereas the current host's name might or might not be abbreviated.
225 The same fix has been applied.
231 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
232 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
235 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
236 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
238 PH/03 Typo: missing ".o" in src/pcre/Makefile.
240 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
241 header line, restrict the check to what is listed in RFCs 2369 and 2929.
242 Also, for "Auto-Submitted", treat anything other than "no" as
243 non-personal, in accordance with RFC 3834. (Previously it treated
244 anything starting "auto-" as non-personal.)
246 TF/01 The control=submission/name=... option had a problem with syntax
247 errors if the name included a slash character. The /name= option
248 now slurps the rest of the string, so it can include any characters
249 but it must come last in the list of options (after /sender_retain
252 PH/05 Some modifications to the interface to the fake nameserver for the new
260 TK/01 Added the "success_on_redirect" address verification option. See
261 NewStuff for rationale and an example.
263 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
265 PH/02 Patch to exigrep to allow it to work on syslog lines.
267 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
268 fread() to skip over the body file's header line, because in Cygwin the
269 header line is locked and is inaccessible.
271 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
272 co-exist for some time) to make it clear that it is the Exim ID that is
273 referenced, not the Message-ID: header line.
275 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
276 string_format() function, because snprintf() does not exist on all
279 PH/06 The use of forbid_filter_existstest now also locks out the use of the
280 ${stat: expansion item.
282 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
283 protocol synchronization error", to keep the pedants happy.
285 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
286 well as for IRIX systems, when gcc is being used. See the host.c source
289 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
291 PH/10 Named domain lists were not working if used in a queue_smtp_domains
294 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
295 transport and to the smtp transport in LMTP mode.
297 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
299 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
300 run a filter in a subprocess. This could lead to confusion in subsequent
301 lookups in the parent process. There should also be a search_tidyup() at
302 the end of the subprocess.
304 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
305 only if the host matched helo_try_verify_hosts, which caused the
306 verification to occur when the EHLO/HELO command was issued. The ACL just
307 tested the remembered result. Now, if a previous verification attempt has
308 not happened, "verify = helo" does it there and then.
310 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
312 TK/03 Fix log output including CR from clamd.
314 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
315 exist provoked a memory error which could cause a segfault.
317 PH/15 Installed PCRE 6.2
319 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
321 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
322 of the problem. Specifically, suggested +O2 rather than +O1 for the
325 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
327 PH/20 If a delivery was routed to a non-standard port by means of an SRV
328 record, the port was not correctly logged when the outgoing_port log
329 selector was set (it logged the transort's default port).
331 PH/21 Added support for host-specific ports to manualroute, queryprogram,
332 fallback_hosts, and "hosts" in the smtp transport.
334 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
335 host errors such as "Connection refused".
337 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
338 authentication with radiusclient 0.4.9:
340 - Error returned from rc_read_config was caught wrongly
341 - Username/password not passed on to radius server due to wrong length.
343 The presumption is that some radiusclient API changes for 4.51/PH/17
344 were not taken care of correctly. The code is still untested by me (my
345 Linux distribution still has 0.3.2 of radiusclient), but it was
346 contributed by a Radius user.
348 PH/24 When doing a callout, the value of $domain wasn't set correctly when
349 expanding the "port" option of the smtp transport.
351 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
352 while reading a MIME header. Thanks to Tom Hughes for a patch.
354 PH/24 Include config.h inside local_scan.h so that configuration settings are
357 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
358 that in an ACL for RCPT (for example), you can examine exactly what was
361 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
362 commands, but it was not correctly comparing the address with the actual
363 client host address. Thus, it would show the EHLO address in Received:
364 header lines when this was not necessary.
366 PH/27 Added the % operator to ${eval:}.
368 PH/28 Exim tries to create and chdir to its spool directory when it starts;
369 it should be ignoring failures (because with -C, for example, it has lost
370 privilege). It wasn't ignoring creation failures other than "already
373 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
376 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
377 because an input line was too long, either on its own, or by virtue of
378 too many continuations, the temporary file was not being removed, and the
379 return code was incorrect.
381 PH/31 Missing "BOOL" in function definition in filtertest.c.
383 PH/32 Applied Sieve patches from the maintainer.
385 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
387 PH/33 Added "verify = not_blind".
389 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
390 Local/Makefile (with some defaults set). These are used in built scripts
391 such as exicyclog, but they have never been used in the exim_install
392 script (though there are many overriding facilities there). I have
393 arranged that the exim_install script now takes note of these two
396 PH/35 Installed configuration files for Dragonfly.
398 PH/36 When a locally submitted message by a trusted user did not contain a
399 From: header, and the sender address was obtained from -f or from an SMTP
400 MAIL command, and the trusted user did not use -F to supply a sender
401 name, $originator_name was incorrectly used when constructing a From:
402 header. Furthermore, $originator_name was used for submission mode
403 messages from external hosts without From: headers in a similar way,
404 which is clearly wrong.
406 PH/37 Added control=suppress_local_fixups.
408 PH/38 When log_selector = +received_sender was set, and the addition of the
409 sender made the log line's construction buffer exactly full, or one byte
410 less than full, an overflow happened when the terminating "\n" was
413 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
414 when the result of a list match is failure because a DNS lookup failed.
416 PH/40 RM_COMMAND is now used in the building process.
418 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
419 the "build-* directories that it finds.
421 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
422 address in a domain literal was a prefix of an interface address.
424 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
425 when verifying a sender address, unless rewrite_headers is false.
427 PH/44 Wrote a long comment about why errors_to addresses are verified as
428 recipients, not senders.
430 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
431 the ratelimit ACL was added.
433 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
435 PH/46 Added extra information about PostgreSQL errors to the error string.
437 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
438 avoiding the need to install special zones in a real server. This is
439 backwards compatible; if it can't find the fake resolver, it drops back.
440 Thus, both old and new test suites can be run.
442 TF/02 Added util/ratelimit.pl
444 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
445 clock is set back in time.
447 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
448 Candler <B.Candler@pobox.com>.
450 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
451 OK for addresses that are the result of redirections.
453 PH/48 A number of further additions for the benefit of the new test suite,
454 including a fake gethostbyname() that interfaces to the fake DNS resolver
457 TF/06 The fix for widen_domains has also been applied to qualify_single and
458 search_parents which are the other dnslookup options that can cause
461 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
464 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
466 TF/07 Exim produced the error message "an SRV record indicated no SMTP
467 service" if it encountered an MX record with an empty target hostname.
468 The message is now "an MX or SRV record indicated no SMTP service".
470 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
471 if the DNS of the sending site is misconfigured. This is quite a
472 common situation. This change restores the behaviour of treating a
473 helo verification defer as a failure.
475 PH/51 If self=fail was set on a router, the bounce message did not include the
476 actual error message.
482 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
484 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
485 command itself ended in error, the underlying message about the transport
486 filter timeout was being overwritten with the pipe command error. Now the
487 underlying error message should be appended to the second error message.
489 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
491 PH/02 Reduce the amount of output that "make" produces by default. Full output
492 can still be requested.
494 PH/03 The warning log line about a condition test deferring for a "warn" verb
495 was being output only once per connection, rather than after each
496 occurrence (because it was using the same function as for successful
497 "warn" verbs). This seems wrong, so I have changed it.
499 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
500 it should not have, which might have caused a crash in the right
501 circumstances, but probably never did.
503 PH/04 Installed a modified version of Tony Finch's patch to make submission
504 mode fix the return path as well as the Sender: header line, and to
505 add a /name= option so that you can make the user's friendly name appear
508 TF/03 Added the control = fakedefer ACL modifier.
510 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
511 Mark Lowes for thorough testing.
513 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
515 TK/03 Merged latest SRS patch from Miles Wilton.
517 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
518 with the definition in sysexits.h (which is #included earlier).
519 Fortunately, Exim does not actually use EX_OK. The code used to try to
520 preserve the sysexits.h value, by assumimg that macro definitions were
521 scanned for macro replacements. I have been disabused of this notion,
522 so now the code just undefines EX_OK before #including unistd.h.
524 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
525 in the smtp transport. When a block could not be written in a single
526 write() function, the timeout was being re-applied to each part-write.
527 This seems wrong - if the receiver was accepting one byte at a time it
528 would take for ever. The timeout is now adjusted when this happens. It
529 doesn't have to be particularly precise.
531 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
532 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
534 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
535 without a domain if the check to <postmaster@domain> fails.
537 SC/01 Eximstats: added -xls and the ability to specify output files
538 (patch written by Frank Heydlauf).
540 SC/02 Eximstats: use FileHandles for outputing results.
542 SC/03 Eximstats: allow any combination of xls, txt, and html output.
544 SC/04 Eximstats: fixed display of large numbers with -nvr option
546 SC/05 Eximstats: fixed merging of reports with empty tables.
548 SC/06 Eximstats: added the -include_original_destination flag
550 SC/07 Eximstats: removed tabs and trailing whitespace.
552 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
554 TK/06 MBOX spool code: Add real "From " MBOX separator line
555 so the .eml file is really in mbox format (even though
556 most programs do not really care). Patch from Alex Miller.
558 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
559 The latter is generated from $received_to and is only set if the
560 message has one envelope recipient. SA can use these headers,
561 obviously out-of-the-box. Patch from Alex Miller.
563 PH/08 The ${def test on a variable was returning false if the variable's
564 value was "0", contrary to what the specification has always said!
565 The result should be true unless the variable is empty.
567 PH/09 The syntax error of a character other than { following "${if
568 def:variable_name" (after optional whitespace) was not being diagnosed.
569 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
570 accidental colon was present, for example, could give incorrect results.
572 PH/10 Tidied the code in a number of places where the st_size field of a stat()
573 result is used (not including appendfile, where other changes are about
576 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
577 This involved changing a lot of size variables from int to off_t. It
578 should work with maildirs and everything.
580 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
581 spamd dying while we are connected to it.
583 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
586 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
587 function for BDB 4.3.
589 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
591 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
592 the sources for PCRE have been re-arranged and more files are now
595 PH/15 The code I had for printing potentially long long variables in PH/11
596 above was not the best (it lost precision). The length of off_t variables
597 is now inspected at build time, and an appropriate printing format (%ld
598 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
599 to be "long long int" or "long int". This is needed for the internal
600 formatting function string_vformat().
602 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
603 the configuration file to be ":syslog", then the script "guesses" where
604 the logs files are, rather than using the compiled in default. In our
605 case the guess is not the same as the compiled default, so the script
606 suddenly stopped working when I started to use syslog. The patch checks
607 to see if log_file_path is "". If so, it attempts to read it from exim
608 with no configuration file to get the compiled in version, before it
609 falls back to the previous guessing code."
611 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
612 implementing BATV in an Exim configuration. See NewStuff for the gory
615 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
616 Makefile that are specific to HP-UX.
618 PH/18 If the "use_postmaster" option was set for a recipient callout together
619 with the "random" option, the postmaster address was used as the MAIL
620 FROM address for the random test, but not for the subsequent recipient
621 test. It is now used for both.
623 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
624 patch removes a few documentation additions to RFC 3028, because the
625 latest draft now contains them. It adds the new en;ascii-case comparator
626 and a new error check for 8bit text in MIME parts. Comparator and
627 require names are now matched exactly. I enabled the subaddress
628 extension, but it is not well tested yet (read: it works for me)."
630 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
631 rework some of the code of TK/09 above to avoid the hardwired use of
632 "%lld" and "long long". Replaced the call to snprintf() with a call to
635 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
636 records point to non-existent hosts", "retry timeout exceeded", and
637 "retry time not reached for any host after a long failure period".
639 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
640 experimental DomainKeys support:
642 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
643 (2) On an error such as an illegally used "control", the wrong name for
644 the control was given.
646 These problems did NOT occur unless DomainKeys support was compiled.
648 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
650 PH/24 Added ${if match_ip condition.
652 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
653 databases so that it will be absolutely obvious if a crash occurs in the
654 DB library. This is a regular occurrence (often caused by mis-matched
657 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
658 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
659 picked up on a user's system that detects such things. There doesn't seem
660 to be a gcc warning option for this - only an attribute that has to be
661 put on the function's prototype. It seems that in Fedora Core 4 they have
662 set this on a number of new functions. No doubt there will be more in due
665 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
666 specify a transport. However, if an address that was verified by such a
667 router was the subject of a callout, Exim crashed because it tried to
668 read the rcpt_include_affixes from the non-existent transport. Now it
669 just assumes that the setting of that option is false. This bug was
670 introduced by 4.51/PH/31.
672 PH/28 Changed -d+all to exclude +memory, because that information is very
673 rarely of interest, but it makes the output a lot bigger. People tend to
674 do -d+all out of habit.
676 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
677 code in os-type was giving problems when libc.so lives in lib64, like on
680 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
681 aren't the modern standard, and it seems that some systems' include files
682 don't always have them. Exim was already checking for some of the newer
683 ones like T_AAAA, and defining it itself. I've added checks for all the
684 record types that Exim uses.
686 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
687 not automatically generating a new one, as it is supposed to. This
688 prevented TLS from working. If the file did exist, but contained invalid
689 data, a new version was generated, as expected. It was only the case of a
690 non-existent file that was broken.
692 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
693 with a change in libdomainkeys > 0.64.
695 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
696 from DNS. If the selector record carries the flag, it now has
697 precedence over the domain-wide flag.
699 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
701 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
702 the use of a port name that isn't defined in /etc/services) occurred, the
703 message was deferred as in a normal delivery, and thus remained on the
704 spool, instead of being failed because of the mua_wrapper setting. This
705 is now fixed, and I tidied up some of the mua_wrapper messages at the
708 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
709 the messages in a hash of arrays rather than using individual hashes.
710 This is a bit cleaner and results in dramatic memory savings, albeit
711 at a slight CPU cost.
713 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
714 as requested by Marc Sherman.
716 SC/10 Eximstats: added histograms for user specified patterns as requested
719 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
721 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
722 fopen() in the content-scanning modules that did not already have it.
728 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
729 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
731 TK/02 Fix ACL "control" statement not being available in MIME ACL.
733 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
735 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
736 to test Sieve filters that use "vacation".
738 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
739 that changes the way the GnuTLS parameters are stored in the cache file.
740 The new format can be generated externally. For backward compatibility,
741 if the data in the cache doesn't make sense, Exim assumes it has read an
742 old-format file, and it generates new data and writes a new file. This
743 means that you can't go back to an older release without removing the
746 PH/03 A redirect router that has both "unseen" and "one_time" set does not
747 work if there are any delivery delays because "one_time" forces the
748 parent to be marked "delivered", so its unseen clone is never tried
749 again. For this reason, Exim now forbids the simultaneous setting of
752 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
753 redirected to themselves ("homonym" addresses). Read the long ChangeLog
754 entry if you want to know the details. The fix, however, neglected to
755 consider the case when local delivery batching is involved. The test for
756 "previously delivered" was not happening when checking to see if an
757 address could be batched with a previous (undelivered) one; under
758 certain circumstances this could lead to multiple deliveries to the same
761 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
762 in its include files, and this causes problems building Exim.
764 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
765 header_syntax) but Exim was just ignoring anything given after a slash.
766 In particular, this caused confusion with an attempt to use "verify =
767 reverse_host_lookup/defer_ok". An error is now given when options are
768 supplied for verify items that do not have them. (Maybe reverse_host_
769 lookup should have a defer_ok option, but that's a different point.)
771 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
772 defined by RFC 821) to 2048, because there were problems with some AUTH
773 commands, and RFC 1869 says the size should be increased for extended
774 SMTP commands that take arguments.
776 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
779 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
780 "unknown" error; now it says that the functionality isn't in the binary.
782 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
783 an address' error message when a string expansion fails (syntax or
784 whatever). Otherwise the password may appear in the log. Following change
785 PH/42 below, there is no longer a chance of it appearing in a bounce
788 PH/11 Installed exipick version 20050225.0 from John Jetmore.
790 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
791 of its addresses was ever tried. (Bugzilla bug #2.)
793 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
794 the result incorrectly in the debug output. (It correctly added a newline
795 to what was transported.)
797 TF/01 Added $received_time.
799 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
800 commented out examples of how to interface to a virus scanner and to
801 SpamAssassin. Also added commented examples of av_scanner and
802 spamd_address settings.
804 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
805 and controls are allowed in which ACLs. There were a couple of minor
806 errors. Some of the entries in the conditions table (which is a table of
807 where they are NOT allowed) were getting very unwieldy; rewrote them as a
808 negation of where the condition IS allowed.
810 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
812 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
813 header file does not have a version number, so I've had to invent a new
814 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
815 API. The code is untested by me (my Linux distribution still has 0.3.2 of
816 radiusclient), but it was contributed by a Radius user.
818 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
819 files or directories, for OpenSSL.
821 PH/19 When an Exim process that is running as root has to create an Exim log
822 file, it does so in a subprocess that runs as exim:exim so as to get the
823 ownership right at creation (otherwise, other Exim processes might see
824 the file with the wrong ownership). There was no test for failure of this
825 fork() call, which would lead to the process getting stuck as it waited
826 for a non-existent subprocess. Forks do occasionally fail when resources
827 run out. I reviewed all the other calls to fork(); they all seem to check
830 PH/20 When checking for unexpected SMTP input at connect time (before writing
831 the banner), Exim was not dealing correctly with a non-positive return
832 from the read() function. If the client had disconnected by this time,
833 the result was a log entry for a synchronization error with an empty
834 string after "input=" when read() returned zero. If read() returned -1
835 (an event I could not check), uninitialized data bytes were printed.
836 There were reports of junk text (parts of files, etc) appearing after
839 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
841 PH/22 Added support for macro redefinition, and (re)definition in between
842 driver and ACL definitions.
844 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
845 forgetting to use the resulting value; it was using the unexpanded value.
847 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
848 hadn't been configured. The fix is from Juergen Kreileder, who
849 understands it better than I do:
851 "Here's what I see happening with three configured cyrus_sasl
852 authenticators configured (plain, login, cram-md5):
854 On startup auth_cyrus_sasl_init() gets called for each of these.
855 This means three calls to sasl_listmech() without a specified mech_list.
856 => SASL tests which mechs of all available mechs actually work
857 => three warnings about OTP not working
858 => the returned list contains: plain, login, cram-md5, digest-md5, ...
860 With the patch, sasl_listmech() also gets called three times. But now
861 SASL's mech_list option is set to the server_mech specified in the the
862 authenticator. Or in other words, the answer from sasl_listmech()
863 gets limited to just the mech you're testing for (which is different
865 => the return list contains just 'plain' or 'login', 'cram-md5' or
866 nothing depending on the value of ob->server_mech.
868 I've just tested the patch: Authentication still works fine,
869 unavailable mechs specified in the exim configuration are still
870 caught, and the auth.log warnings about OTP are gone."
872 PH/25 When debugging is enabled, the contents of the command line are added
873 to the debugging output, even when log_selector=+arguments is not
876 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
877 answer is "GNU", and only if the return is "GNU/something" is the answer
880 PH/27 $acl_verify_message is now set immediately after the failure of a
881 verification in an ACL, and so is available in subsequent modifiers. In
882 particular, the message can be preserved by coding like this:
884 warn !verify = sender
885 set acl_m0 = $acl_verify_message
887 Previously, $acl_verify_message was set only while expanding "message"
888 and "log_message" when a very denied access.
890 PH/28 Modified OS/os.c-Linux with
892 -#ifndef OS_LOAD_AVERAGE
893 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
895 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
896 nomenclature these days.)
898 PH/29 Installed patch from the Sieve maintainer that adds the options
899 sieve_useraddress and sieve_subaddress to the redirect router.
901 PH/30 In these circumstances:
902 . Two addresses routed to the same list of hosts;
903 . First host does not offer TLS;
904 . First host accepts first address;
905 . First host gives temporary error to second address;
906 . Second host offers TLS and a TLS session is established;
907 . Second host accepts second address.
908 Exim incorrectly logged both deliveries with the TLS parameters (cipher
909 and peerdn, if requested) that were in fact used only for the second
912 PH/31 When doing a callout as part of verifying an address, Exim was not paying
913 attention to any local part prefix or suffix that was matched by the
914 router that accepted the address. It now behaves in the same way as it
915 does for delivery: the affixes are removed from the local part unless
916 rcpt_include_affixes is set on the transport.
918 PH/32 Add the sender address, as F=<...>, to the log line when logging a
919 timeout during the DATA phase of an incoming message.
921 PH/33 Sieve envelope tests were broken for match types other than :is. I have
922 applied a patch sanctioned by the Sieve maintainer.
924 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
925 the uid or gid is negative. A case of a negative gid caused this to be
926 noticed. The fix allows for either to be negative.
928 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
929 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
932 PH/36 The change PH/12 above was broken. Fixed it.
934 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
935 the grounds that routing the same address twice would always produce the
936 same answer. This might have been true once, but it is certainly no
937 longer true now. Routing a child address may depend on the previous
938 routing that produced that child. Some complicated redirection strategies
939 went wrong when messages had multiple recipients, and made Exim's
940 behaviour dependent on the order in which the addresses were given.
942 I have moved the duplicate checking until after the routing is complete.
943 Exim scans the addresses that are assigned to local and remote
944 transports, and removes any duplicates. This means that more work will be
945 done, as duplicates will always all be routed, but duplicates are
946 presumably rare, so I don't expect this is of any significance.
948 For deliveries to pipes, files, and autoreplies, the duplicate checking
949 still happens during the routing process, since they are not going to be
952 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
953 It corrects a timeout issue with spamd. This is Ian's comment: "The
954 background is that sometimes spamd either never reads data from a
955 connection it has accepted, or it never writes response data. The exiscan
956 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
957 blindly assumes that writes won't block so it may never time out."
959 PH/39 Allow G after quota size as well as K and M.
961 PH/40 The value set for $authenticated_id in an authenticator may not contain
962 binary zeroes or newlines because the value is written to log lines and
963 to spool files. There was no check on this. Now the value is run through
964 the string_printing() function so that such characters are converted to
965 printable escape sequences.
967 PH/41 $message_linecount is a new variable that contains the total number of
968 lines in the message. Compare $body_linecount, which is the count for the
971 PH/42 Exim no longer gives details of delivery errors for specific addresses in
972 bounce and delay warning messages, except in certain special cases, which
975 (a) An SMTP error message from a remote host;
976 (b) A message specified in a :fail: redirection;
977 (c) A message specified in a "fail" command in a system filter;
978 (d) A message specified in a FAIL return from the queryprogram router;
979 (e) A message specified by the cannot_route_message router option.
981 In these cases only, Exim does include the error details in bounce and
982 warning messages. There are also a few cases where bland messages such
983 as "unrouteable address" or "local delivery error" are given.
985 PH/43 $value is now also set for the "else" part of a ${run expansion.
987 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
988 being worked on, but at least Exim now implements the latest version to
991 PH/45 In a pipe transport, although a timeout while waiting for the pipe
992 process to complete was treated as a delivery failure, a timeout while
993 writing the message to the pipe was logged, but erroneously treated as a
994 successful delivery. Such timeouts include transport filter timeouts. For
995 consistency with the overall process timeout, these timeouts are now
996 treated as errors, giving rise to delivery failures by default. However,
997 there is now a new Boolean option for the pipe transport called
998 timeout_defer, which, if set TRUE, converts the failures into defers for
999 both kinds of timeout. A transport filter timeout is now identified in
1002 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1003 systems where "make" and "gmake" are different, calling "gmake" at top
1004 level broke things. I've arranged for the value of $(MAKE) to be passed
1005 from the Makefile to this script so that it can call the same version of
1009 A note about Exim versions 4.44 and 4.50
1010 ----------------------------------------
1012 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1013 changes of various kinds. As a consequence, a big documentation update was
1014 needed. This delayed the release for rather longer than seemed good, especially
1015 in the light of a couple of (minor) security issues. Therefore, the changes
1016 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1017 release. So 4.44 and 4.50 are in effect two different branches that both start
1020 I have left the 4.50 change log unchanged; it contains all the changes since
1021 4.43. The change log for 4.44 is below; many of its items are identical to
1022 those for 4.50. This seems to be the most sensible way to preserve the
1023 historical information.
1029 1. Minor wording change to the doc/README.SIEVE file.
1031 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1032 computation of the current number of files was incorrect.
1034 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1035 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1038 4. Give more explanation in the error message when the command for a transport
1039 filter fails to execute.
1041 5. There are several places where Exim runs a non-Exim command in a
1042 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1043 was being done only for the command run by the queryprogram router. It is
1044 now done for all such subprocesses. The other cases are: ${run, transport
1045 filters, and the commands run by the lmtp and pipe transports.
1047 6. Added CONFIGURE_GROUP build-time option.
1049 7. Some older OS have a limit of 256 on the maximum number of file
1050 descriptors. Exim was using setrlimit() to set 1000 as a large value
1051 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1052 systems. I've change it so that if it can't get 1000, it tries for 256.
1054 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1055 was an oversight, and furthermore, ever since the addition of extra
1056 controls (e.g. 4.43/32), the checks on when to allow different forms of
1057 "control" were broken. There should now be diagnostics for all cases when a
1058 control that does not make sense is encountered.
1060 9. Added the /retain_sender option to "control=submission".
1062 10. $recipients is now available in the predata ACL (oversight).
1064 11. Tidy the search cache before the fork to do a delivery from a message
1065 received from the command line. Otherwise the child will trigger a lookup
1066 failure and thereby defer the delivery if it tries to use (for example) a
1067 cached ldap connection that the parent has called unbind on.
1069 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1070 of $address_data from the recipient verification was clobbered by the
1071 sender verification.
1073 13. The value of address_data from a sender verification is now available in
1074 $sender_address_data in subsequent conditions in the ACL statement.
1076 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1078 15. Added a new option "connect=<time>" to callout options, to set a different
1081 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1082 was its contents. (It was OK if the option was not defined at all.)
1084 17. A "Completed" log line is now written for messages that are removed from
1085 the spool by the -Mrm option.
1087 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1088 information about exactly what failed.
1090 19. Added -dd to debug only the daemon process.
1092 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1093 handles timeouts, both on the server side and network timeouts. Renamed the
1094 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1096 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1097 It is now set to "smtps".
1099 22. $host_address is now set to the target address during the checking of
1100 ignore_target_hosts.
1102 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1103 being passed; this would have caused $sender_host_name to have been used if
1104 matching the list had actually called for a host name (not very likely,
1105 since this list is usually IP addresses). A host name is now passed as
1108 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1109 code that specifies a non-restarting handler (typically sigaction() in
1110 modern systems) in an attempt to fix a rare and obscure crash bug.
1112 25. Narrowed the window for a race in the daemon that could cause it to ignore
1113 SIGCHLD signals. This is not a major problem, because they are used only to
1114 wake it up if nothing else does.
1116 26. A malformed maildirsize file could cause Exim to calculate negative values
1117 for the mailbox size or file count. Odd effects could occur as a result.
1118 The maildirsize information is now recalculated if the size or filecount
1121 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1122 support for a long time. Removed HAVE_SYS_VFS_H.
1124 28. Installed the latest version of exipick from John Jetmore.
1126 29. In an address list, if the pattern was not a regular expression, an empty
1127 subject address (from a bounce message) matched only if the pattern was an
1128 empty string. Non-empty patterns were not even tested. This was the wrong
1129 because it is perfectly reasonable to use an empty address as part of a
1130 database query. An empty address is now tested by patterns that are
1131 lookups. However, all the other forms of pattern expect the subject to
1132 contain a local part and a domain, and therefore, for them, an empty
1133 address still always fails if the pattern is not itself empty.
1135 30. Exim went into a mad DNS loop when attempting to do a callout where the
1136 host was specified on an smtp transport, and looking it up yielded more
1137 than one IP address.
1139 31. Re-factored the code for checking spool and log partition space into a
1140 function that finds that data and another that does the check. The former
1141 is then used to implement four new variables: $spool_space, $log_space,
1142 $spool_inodes, and $log_inodes.
1144 32. The RFC2047 encoding function was originally intended for short strings
1145 such as real names; it was not keeping to the 75-character limit for
1146 encoded words that the RFC imposes. It now respects the limit, and
1147 generates multiple encoded words if necessary. To be on the safe side, I
1148 have increased the buffer size for the ${rfc2047: expansion operator from
1151 33. It is now permitted to omit both strings after an "if" condition; if the
1152 condition is true, the result is "true". As before, when the second string
1153 is omitted, a false condition yields an empty string. This makes it less
1154 cumbersome to write custom ACL and router conditions.
1156 34. Failure to deliver a bounce message always caused it to be frozen, even if
1157 there was an errors_to setting on the router. The errors_to setting is now
1160 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1161 canonical form (fully expanded) before being placed in
1162 $sender_host_address.
1164 36. The table in the code that translates DNS record types into text (T_A to
1165 "A" for instance) was missing entries for NS and CNAME. It is just possible
1166 that this could have caused confusion if both these types were looked up
1167 for the same domain, because the text type is used as part of Exim's
1168 per-process caching. But the chance of anyone hitting this buglet seems
1171 37. The dnsdb lookup has been extended in a number of ways.
1173 (1) There is a new type, "zns", which walks up the domain tree until it
1174 finds some nameserver records. It should be used with care.
1176 (2) There is a new type, "mxh", which is like "mx" except that it returns
1177 just the host names, not the priorities.
1179 (3) It is now possible to give a list of domains (or IP addresses) to be
1180 looked up. The behaviour when one of the lookups defers can be
1181 controlled by a keyword.
1183 (4) It is now possible to specify the separator character for use when
1184 multiple records are returned.
1186 38. The dnslists ACL condition has been extended: it is now possible to supply
1187 a list of IP addresses and/or domains to be looked up in a particular DNS
1190 39. Added log_selector=+queue_time_overall.
1192 40. When running the queue in the test harness, wait just a tad after forking a
1193 delivery process, to get repeatability of debugging output.
1195 41. Include certificate and key file names in error message when GnuTLS fails
1196 to set them up, because the GnuTLS error message doesn't include the name
1197 of the failing file when there is a problem reading it.
1199 42. Allow both -bf and -bF in the same test run.
1201 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1203 44. The "Exiscan patch" is now merged into the mainline Exim source.
1205 45. Sometimes the final signoff response after QUIT could fail to get
1206 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1207 < 0 before doing a fflush(). This bug looks as though it goes back to the
1208 introduction of TLS in release 3.20, but "sometimes" must have been rare
1209 because the tests only now provoked it.
1211 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1212 (this can affect the format of dates).
1214 47. exim_tidydb, when checking for the continued existence of a message for
1215 which it has found a message-specific retry record, was not finding
1216 messages that were in split spool directories. Consequently, it was
1217 deleting retry records that should have stayed in existence.
1219 48. Steve fixed some bugs in eximstats.
1221 49. The SPA authentication driver was not abandoning authentication and moving
1222 on to the next authenticator when an expansion was forced to fail,
1223 contradicting the general specification for all authenticators. Instead it
1224 was generating a temporary error. It now behaves as specified.
1226 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1227 (the order specifies the preference for clients). The order is now AES256,
1228 AES128, 3DES, ARCFOUR128.
1230 51. Small patch to Sieve code - explicitly set From: when generating an
1233 52. Exim crashed if a remote delivery caused a very long error message to be
1234 recorded - for instance if somebody sent an entire SpamAssassin report back
1235 as a large number of 550 error lines. This bug was coincidentally fixed by
1236 increasing the size of one of Exim's internal buffers (big_buffer) that
1237 happened as part of the Exiscan merge. However, to be on the safe side, I
1238 have made the code more robust (and fixed the comments that describe what
1241 53. Now that there can be additional text after "Completed" in log lines (if
1242 the queue_time_overall log selector is set), a one-byte patch to exigrep
1243 was needed to allow it to recognize "Completed" as not the last thing in
1246 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1247 patch that reportedly fixes this has been added. I am not expert enough to
1248 create a test for it. This is what the patch creator wrote:
1250 "I found a little strange behaviour of ldap code when working with
1251 Windows 2003 AD Domain, where users was placed in more than one
1252 Organization Units. When I tried to give exim partial DN, the exit code
1253 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1254 But simultaneously result of request was absolutely normal ldap result,
1255 so I produce this patch..."
1257 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1258 REFERENCE, so I have modified the code to exclude the patch when that macro
1261 55. Some experimental protocols are using DNS PTR records for new purposes. The
1262 keys for these records are domain names, not reversed IP addresses. The
1263 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1264 leaves it alone. Component reversal etc. now happens only for IP addresses.
1267 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1269 57. Double the size of the debug message buffer (to 2048) so that more of very
1270 long debug lines gets shown.
1272 58. The exicyclog utility now does better if the number of log files to keep
1273 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1275 59. Two changes related to the smtp_active_hostname option:
1277 (1) $smtp_active_hostname is now available as a variable.
1278 (2) The default for smtp_banner uses $smtp_active_hostname instead
1279 of $primary_hostname.
1281 60. The host_aton() function is supposed to be passed a string that is known
1282 to be a valid IP address. However, in the case of IPv6 addresses, it was
1283 not checking this. This is a hostage to fortune. Exim now panics and dies
1284 if the condition is not met. A case was found where this could be provoked
1285 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1286 components; fortuitously, this particular loophole had already been fixed
1287 by change 4.50/55 above.
1289 If there are any other similar loopholes, the new check in host_aton()
1290 itself should stop them being exploited. The report I received stated that
1291 data on the command line could provoke the exploit when Exim was running as
1292 exim, but did not say which command line option was involved. All I could
1293 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1294 running as the user.
1297 61. There was a buffer overflow vulnerability in the SPA authentication code
1298 (which came originally from the Samba project). I have added a test to the
1299 spa_base64_to_bits() function which I hope fixes it.
1302 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1303 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1305 63. The daemon start-up calls getloadavg() while still root for those OS that
1306 need the first call to be done as root, but it missed one case: when
1307 deliver_queue_load_max is set with deliver_drop_privilege. This is
1308 necessary for the benefit of the queue runner, because there is no re-exec
1309 when deliver_drop_privilege is set.
1311 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1312 This has been fixed.
1314 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1315 was in use, was not putting the data itself into the right store pool;
1316 consequently, it could be overwritten for a subsequent message in the same
1317 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1320 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1322 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1323 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1324 the function were treating the return as a boolean value, which happened to
1325 work because 0=false and not-0=true, but is not correct code.
1327 68. The host_aton() function was not handling scoped IPv6 addresses (those
1328 with, for example, "%eth0" on the end) correctly.
1330 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1331 negated items (that is, ~something) in unsigned ints. Some compilers
1332 apparently mutter when there is no cast.
1334 70. If an address verification called from an ACL failed, and did not produce a
1335 user-specific message (i.e. there was only a "system" message), nothing was
1336 put in $acl_verify_message. In this situation, it now puts the system
1339 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1340 session; change 4.31/43 added the unwanted input to the log line - except
1341 that it did not do this in the start of session case. It now does.
1343 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1344 This is wrong and can cause the other end to generate a synchronization
1345 error if it is another Exim or anything else that does the synchronization
1346 check. A QUIT command is no longer sent after a timeout.
1348 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1349 during host lookups.
1351 74. The defer_ok option of callout verification was not working if it was used
1352 when verifying addresses in header lines, that is, for this case:
1354 verify = header_sender/callout=defer_ok
1356 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1357 those file descriptors could be used for SMTP connections. If anything
1358 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1359 could be sent to the SMTP client, causing chaos. The daemon now opens
1360 stdin, stdout, and stderr to /dev/null when it puts itself into the
1363 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1364 log by default. The user can override this with suitable Perl magic.
1366 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1367 the log message when discard triggers, was not working for the DATA ACL or
1368 for the non-SMTP ACL.
1370 78. Error message wording change in sieve.c.
1372 79. If smtp_accept_max_per_host was set, the number of connections could be
1373 restricted to fewer than expected, because the daemon was trying to set up
1374 a new connection before checking whether the processes handling previous
1375 connections had finished. The check for completed processes is now done
1376 earlier. On busy systems, this bug wouldn't be noticed because something
1377 else would have woken the daemon, and it would have reaped the completed
1380 80. If a message was submitted locally by a user whose login name contained one
1381 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1382 It caused a spool format error. I have fixed the spool reading code. A
1383 related problem was that the "from" clause in the Received: line became
1384 illegal because of the space(s). It is now covered by ${quote_local_part.
1386 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1389 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1391 83. Patch from Sieve maintainer to fix unterminated string problem in
1392 "vacation" handling.
1394 84. Some minor changes to the Linux configuration files to help with other
1395 OS variants using glibc.
1397 85. One more patch for Sieve to update vacation handling to latest spec.
1400 ----------------------------------------------------
1401 See the note above about the 4.44 and 4.50 releases.
1402 ----------------------------------------------------
1408 1. Change 4.43/35 introduced a bug that caused file counts to be
1409 incorrectly computed when quota_filecount was set in an appendfile
1412 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1413 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1416 3. Give more explanation in the error message when the command for a transport
1417 filter fails to execute.
1419 4. There are several places where Exim runs a non-Exim command in a
1420 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1421 was being done only for the command run by the queryprogram router. It is
1422 now done for all such subprocesses. The other cases are: ${run, transport
1423 filters, and the commands run by the lmtp and pipe transports.
1425 5. Some older OS have a limit of 256 on the maximum number of file
1426 descriptors. Exim was using setrlimit() to set 1000 as a large value
1427 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1428 systems. I've change it so that if it can't get 1000, it tries for 256.
1430 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1431 was an oversight, and furthermore, ever since the addition of extra
1432 controls (e.g. 4.43/32), the checks on when to allow different forms of
1433 "control" were broken. There should now be diagnostics for all cases when a
1434 control that does not make sense is encountered.
1436 7. $recipients is now available in the predata ACL (oversight).
1438 8. Tidy the search cache before the fork to do a delivery from a message
1439 received from the command line. Otherwise the child will trigger a lookup
1440 failure and thereby defer the delivery if it tries to use (for example) a
1441 cached ldap connection that the parent has called unbind on.
1443 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1444 of $address_data from the recipient verification was clobbered by the
1445 sender verification.
1447 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1448 was its contents. (It was OK if the option was not defined at all.)
1450 11. A "Completed" log line is now written for messages that are removed from
1451 the spool by the -Mrm option.
1453 12. $host_address is now set to the target address during the checking of
1454 ignore_target_hosts.
1456 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1457 being passed; this would have caused $sender_host_name to have been used if
1458 matching the list had actually called for a host name (not very likely,
1459 since this list is usually IP addresses). A host name is now passed as
1462 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1463 code that specifies a non-restarting handler (typically sigaction() in
1464 modern systems) in an attempt to fix a rare and obscure crash bug.
1466 15. Narrowed the window for a race in the daemon that could cause it to ignore
1467 SIGCHLD signals. This is not a major problem, because they are used only to
1468 wake it up if nothing else does.
1470 16. A malformed maildirsize file could cause Exim to calculate negative values
1471 for the mailbox size or file count. Odd effects could occur as a result.
1472 The maildirsize information is now recalculated if the size or filecount
1475 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1476 support for a long time. Removed HAVE_SYS_VFS_H.
1478 18. Updated exipick to current release from John Jetmore.
1480 19. Allow an empty sender to be matched against a lookup in an address list.
1481 Previously the only cases considered were a regular expression, or an
1484 20. Exim went into a mad DNS lookup loop when doing a callout where the
1485 host was specified on the transport, if the DNS lookup yielded more than
1488 21. The RFC2047 encoding function was originally intended for short strings
1489 such as real names; it was not keeping to the 75-character limit for
1490 encoded words that the RFC imposes. It now respects the limit, and
1491 generates multiple encoded words if necessary. To be on the safe side, I
1492 have increased the buffer size for the ${rfc2047: expansion operator from
1495 22. Failure to deliver a bounce message always caused it to be frozen, even if
1496 there was an errors_to setting on the router. The errors_to setting is now
1499 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1500 canonical form (fully expanded) before being placed in
1501 $sender_host_address.
1503 24. Updated eximstats to version 1.33
1505 25. Include certificate and key file names in error message when GnuTLS fails
1506 to set them up, because the GnuTLS error message doesn't include the name
1507 of the failing file when there is a problem reading it.
1509 26. Expand error message when OpenSSL has problems setting up cert/key files.
1512 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1513 (this can affect the format of dates).
1515 28. exim_tidydb, when checking for the continued existence of a message for
1516 which it has found a message-specific retry record, was not finding
1517 messages that were in split spool directories. Consequently, it was
1518 deleting retry records that should have stayed in existence.
1520 29. eximstats updated to version 1.35
1521 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1522 1.35 - bugfix such that pie charts by volume are generated correctly
1524 30. The SPA authentication driver was not abandoning authentication and moving
1525 on to the next authenticator when an expansion was forced to fail,
1526 contradicting the general specification for all authenticators. Instead it
1527 was generating a temporary error. It now behaves as specified.
1529 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1530 (the order specifies the preference for clients). The order is now AES256,
1531 AES128, 3DES, ARCFOUR128.
1533 31. Small patch to Sieve code - explicitly set From: when generating an
1536 32. Exim crashed if a remote delivery caused a very long error message to be
1537 recorded - for instance if somebody sent an entire SpamAssassin report back
1538 as a large number of 550 error lines. This bug was coincidentally fixed by
1539 increasing the size of one of Exim's internal buffers (big_buffer) that
1540 happened as part of the Exiscan merge. However, to be on the safe side, I
1541 have made the code more robust (and fixed the comments that describe what
1544 33. Some experimental protocols are using DNS PTR records for new purposes. The
1545 keys for these records are domain names, not reversed IP addresses. The
1546 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1547 leaves it alone. Component reversal etc. now happens only for IP addresses.
1550 34. The host_aton() function is supposed to be passed a string that is known
1551 to be a valid IP address. However, in the case of IPv6 addresses, it was
1552 not checking this. This is a hostage to fortune. Exim now panics and dies
1553 if the condition is not met. A case was found where this could be provoked
1554 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1555 components; fortuitously, this particular loophole had already been fixed
1556 by change 4.50/55 or 4.44/33 above.
1558 If there are any other similar loopholes, the new check in host_aton()
1559 itself should stop them being exploited. The report I received stated that
1560 data on the command line could provoke the exploit when Exim was running as
1561 exim, but did not say which command line option was involved. All I could
1562 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1563 running as the user.
1566 35. There was a buffer overflow vulnerability in the SPA authentication code
1567 (which came originally from the Samba project). I have added a test to the
1568 spa_base64_to_bits() function which I hope fixes it.
1571 36. The daemon start-up calls getloadavg() while still root for those OS that
1572 need the first call to be done as root, but it missed one case: when
1573 deliver_queue_load_max is set with deliver_drop_privilege. This is
1574 necessary for the benefit of the queue runner, because there is no re-exec
1575 when deliver_drop_privilege is set.
1577 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1578 was in use, was not putting the data itself into the right store pool;
1579 consequently, it could be overwritten for a subsequent message in the same
1580 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1583 38. Sometimes the final signoff response after QUIT could fail to get
1584 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1585 < 0 before doing a fflush(). This bug looks as though it goes back to the
1586 introduction of TLS in release 3.20, but "sometimes" must have been rare
1587 because the tests only now provoked it.
1593 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1594 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1595 Now it returns an integer. A number of calls were still expecting a T/F
1596 return. Fortuitously, in all cases, the tests worked in OK situations,
1597 which is the norm. However, things would have gone wrong on any write
1598 failures on the smtp file descriptor. This function is used when sending
1599 messages over SMTP and also when doing verify callouts.
1601 2. When Exim is called to do synchronous delivery of a locally submitted
1602 message (the -odf or -odi options), it no longer closes stderr before doing
1605 3. Implemented the mua_wrapper option.
1607 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1609 5. Implemented the functions header_remove(), header_testname(),
1610 header_add_at_position(), and receive_remove_recipient(), and exported them
1613 6. If an ACL "warn" statement specified the addition of headers, Exim already
1614 inserted X-ACL-Warn: at the start if there was no header name. However, it
1615 was not making this test for the second and subsequent header lines if
1616 there were newlines in the string. This meant that an invalid header could
1617 be inserted if Exim was badly configured.
1619 7. Allow an ACL "warn" statement to add header lines at the start or after all
1620 the Received: headers, as well as at the end.
1622 8. Added the rcpt_4xx retry error code.
1624 9. Added postmaster_mailfrom=xxx to callout verification option.
1626 10. Added mailfrom=xxxx to the callout verification option, for verify=
1629 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1630 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1632 12. Inserted some casts to stop certain compilers warning when using pointer
1633 differences as field lengths or precisions in printf-type calls (mostly
1634 affecting debugging statements).
1636 13. Added optional readline() support for -be (dynamically loaded).
1638 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1639 same clock tick as a message's arrival, so that its received time was the
1640 same as the "first fail" time on the retry record, and that message
1641 remained on the queue past the ultimate address timeout, every queue runner
1642 would try a delivery (because it was past the ultimate address timeout) but
1643 after another failure, the ultimate address timeout, which should have then
1644 bounced the address, did not kick in. This was a "< instead of <=" error;
1645 in most cases the first failure would have been in the next clock tick
1646 after the received time, and all would be well.
1648 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1649 being recognized when the domain list was tested by the match_domain
1650 condition in an expansion string.
1652 16. Added the ${str2b64: operator.
1654 17. Exim was always calling setrlimit() to set a large limit for the number of
1655 processes, without checking whether the existing limit was already
1656 adequate. (It did check for the limit on file descriptors.) Furthermore,
1657 errors from getrlimit() and setrlimit() were being ignored. Now they are
1658 logged to the main and panic logs, but Exim does carry on, to try to do its
1659 job under whatever limits there are.
1661 18. Imported PCRE 5.0.
1663 19. Trivial typo in log message " temporarily refused connection" (the leading
1666 20. If the log selector return_path_on_delivery was set and an address was
1667 redirected to /dev/null, the delivery process crashed because it assumed
1668 that a return path would always be set for a "successful" delivery. In this
1669 case, the whole delivery is bypassed as an optimization, and therefore no
1672 21. Internal re-arrangement: the function for sending a challenge and reading
1673 a response while authentication was assuming a zero-terminated challenge
1674 string. It's now changed to take a pointer and a length, to allow for
1675 binary data in such strings.
1677 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1679 23. Exim was not respecting finduser_retries when seeking the login of the
1680 uid under which it was called; it was always trying 10 times. (The default
1681 setting of finduser_retries is zero.) Also, it was sleeping after the final
1682 failure, which is pointless.
1684 24. Implemented tls_on_connect_ports.
1686 25. Implemented acl_smtp_predata.
1688 26. If the domain in control=submission is set empty, Exim assumes that the
1689 authenticated id is a complete email address when it generates From: or
1690 Sender: header lines.
1692 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1693 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1694 chown and chgrp in /bin and hostname in /usr/bin.
1696 28. Exim was keeping the "process log" file open after each use, just as it
1697 does for the main log. This opens the possibility of it remaining open for
1698 long periods when the USR1 signal hits a daemon. Occasional processlog
1699 errors were reported, that could have been caused by this. Anyway, it seems
1700 much more sensible not to leave this file open at all, so that is what now
1703 29. The long-running daemon process does not normally write to the log once it
1704 has entered its main loop, and it closes the log before doing so. This is
1705 so that log files can straightforwardly be renamed and moved. However,
1706 there are a couple of unusual error situations where the daemon does write
1707 log entries, and I had neglected to close the log afterwards.
1709 30. The text of an SMTP error response that was received during a remote
1710 delivery was being truncated at 512 bytes. This is too short for some of
1711 the long messages that one sometimes sees. I've increased the limit to
1714 31. It is now possible to make retry rules that apply only when a message has a
1715 specific sender, in particular, an empty sender.
1717 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1718 it possible to be selective about when SMTP synchronization is enforced.
1720 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1722 32. Implemented hosts_connection_nolog.
1724 33. Added an ACL for QUIT.
1726 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1729 35. Added mailbox_size and mailbox_filecount to appendfile.
1731 36. Added control = no_multiline_responses to ACLs.
1733 37. There was a bug in the logic of the code that waits for the clock to tick
1734 in the case where the clock went backwards by a substantial amount such
1735 that the microsecond fraction of "now" was more than the microsecond
1736 fraction of "then" (but the whole seconds number was less).
1738 38. Added support for the libradius Radius client library this is found on
1739 FreeBSD (previously only the radiusclient library was supported).
1745 1. When certain lookups returned multiple values in the form name=value, the
1746 quoting of the values was not always being done properly. Specifically:
1747 (a) If the value started with a double quote, but contained no whitespace,
1749 (b) If the value contained whitespace other than a space character (i.e.
1750 tabs or newlines or carriage returns) it was not quoted.
1751 This fix has been applied to the mysql and pgsql lookups by writing a
1752 separate quoting function and calling it from the lookup code. The fix
1753 should probably also be applied to nisplus, ibase and oracle lookups, but
1754 since I cannot test any of those, I have not disturbed their existing code.
1756 2. A hit in the callout cache for a specific address caused a log line with no
1757 reason for rejecting RCPT. Now it says "Previous (cached) callout
1758 verification failure".
1760 3. There was an off-by-one bug in the queryprogram router. An over-long
1761 return line was truncated at 256 instead of 255 characters, thereby
1762 overflowing its buffer with the terminating zero. As well as fixing this, I
1763 have increased the buffer size to 1024 (and made a note to document this).
1765 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1766 when Exim is waiting for an SMTP response from a remote server, Exim
1767 restarts its select() call on the socket, thereby resetting its timeout.
1768 This is not a problem when such interrupts are rare. Somebody set up a cron
1769 job to run exiwhat every 2 minutes, which is less than the normal select()
1770 timeout (5 or 10 minutes). This meant that the select() timeout never
1771 kicked in because it was always reset. I have fixed this by comparing the
1772 time when an interrupt arrives with the time at the start of the first call
1773 to select(). If more time than the timeout has elapsed, the interrupt is
1774 treated as a timeout.
1776 5. Some internal re-factoring in preparation for the addition of Sieve
1777 extensions (by MH). In particular, the "personal" test is moved to a
1778 separate function, and given an option for scanning Cc: and Bcc: (which is
1779 not set for Exim filters).
1781 6. When Exim created an email address using the login of the caller as the
1782 local part (e.g. when creating a From: or Sender: header line), it was not
1783 quoting the local part when it contained special characters such as @.
1785 7. Installed new OpenBSD configuration files.
1787 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1788 try to make them clearer.
1790 9. Callout options, other than the timeout value, were being ignored when
1791 verifying sender addresses in header lines. For example, when using
1793 verify = header_sender/callout=no_cache
1795 the cache was (incorrectly) being used.
1797 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1798 only the Cygwin environment.
1800 11. When return_path_on_delivery was set as a log selector, if different remote
1801 addresses in the same message used different return paths and parallel
1802 remote delivery occurred, the wrong values would sometimes be logged.
1803 (Whenever a remote delivery process finished, the return path value from
1804 the most recently started remote delivery process was logged.)
1806 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1807 lines when AUTH and/or TLS are in use. This is the "received protocol"
1808 field. Exim used to use "asmtp" for authenticated SMTP, without any
1809 indication (in the protocol name) for TLS use. Now it follows the RFC and
1810 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1811 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1812 names appear in log lines as well as in Received: header lines.
1814 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1815 extensions, and comparison tests, and to fix some bugs.
1817 14. Changes to the "personal" filter test:
1819 (1) The test was buggy in that it was just doing the equivalent of
1820 "contains" tests on header lines. For example, if a user's address was
1821 anne@some.where, the "personal" test would incorrectly be true for
1823 To: susanne@some.where
1825 This test is now done by extracting each address from the header in turn,
1826 and checking the entire address. Other tests that are part of "personal"
1827 are now done using regular expressions (for example, to check local parts
1828 of addresses in From: header lines).
1830 (2) The list of non-personal local parts in From: addresses has been
1831 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1832 taken from the Sieve specification recommendations.
1834 (3) If the message contains any header line starting with "List-" it is
1835 treated as non-personal.
1837 (4) The test for "circular" in the Subject: header line has been removed
1838 because it now seems ill-conceived.
1840 15. Minor typos in src/EDITME comments corrected.
1842 16. Installed latest exipick from John Jetmore.
1844 17. If headers_add on a router specified a text string that was too long for
1845 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1846 of string_sprintf() is now avoided.
1848 18. $message_body_size was not set (it was always zero) when running the DATA
1849 ACL and the local_scan() function.
1851 19. For the "mail" command in an Exim filter, no default was being set for
1852 the once_repeat time, causing a random time value to be used if "once" was
1853 specified. (If the value happened to be <= 0, no repeat happened.) The
1854 default is now 0s, meaning "never repeat". The "vacation" command was OK
1855 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1856 (I found it when inspecting the code).
1858 20. There is now an overall timeout for performing a callout verification. It
1859 defaults to 4 times the callout timeout, which applies to individual SMTP
1860 commands during the callout. The overall timeout applies when there is more
1861 than one host that can be tried. The timeout is checked before trying the
1862 next host. This prevents very long delays if there are a large number of
1863 hosts and all are timing out (e.g. when the network connections are timing
1864 out). The value of the overall timeout can be changed by specifying an
1865 additional sub-option for "callout", called "maxwait". For example:
1867 verify = sender/callout=5s,maxwait=20s
1869 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1870 to the end before writing, but this should make it even safer).
1872 22. Exim was forgetting that it had advertised PIPELINING for the second and
1873 subsequent messages on an SMTP connection. It was also not resetting its
1874 memory on STARTTLS and an internal HELO.
1876 23. When Exim logs an SMTP synchronization error within a session, it now
1877 records whether PIPELINING has been advertised or not.
1879 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1880 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1881 rather than long int.
1883 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1885 26. Added the never_mail option to autoreply.
1891 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1892 crash if the getsockname() call failed; this can happen if a connection is
1893 closed very soon after it is established. The problem was simply in the
1894 order in which certain operations were done, causing Exim to try to write
1895 to the SMTP stream before it had set up the file descriptor. The bug has
1896 been fixed by making things happen in the correct order.
1902 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1903 before the connection was closed, thus losing the rejection response.
1905 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1906 some early Solaris releases, but causes trouble in current releases where
1907 socklen_t is defined.
1909 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1912 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1915 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1916 that releasing the top of it at the end releases what was used for sub-
1917 expansions (unless the block got too big). However, discard this block if
1918 the first thing is a variable or header, so that we can use its block when
1919 it is dynamic (useful for very large $message_headers, for example).
1921 6. Lookups now cache *every* query, not just the most recent. A new, separate
1922 store pool is used for this. It can be recovered when all lookup caches are
1923 flushed. Lookups now release memory at the end of their result strings.
1924 This has involved some general refactoring of the lookup sources.
1926 7. Some code has been added to the store_xxx() functions to reduce the amount
1927 of flapping under certain conditions.
1929 8. log_incoming_interface used to affect only the <= reception log lines. Now
1930 it causes the local interface and port to be added to several more SMTP log
1931 lines, for example "SMTP connection from", and rejection lines.
1933 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1935 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1937 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1938 could be overwritten at the end of the current message (or the start of a
1939 new message if it was set in a HELO ACL). The value is now preserved for
1940 the duration of the SMTP connection.
1942 12. If a transport had a headers_rewrite setting, and a matching header line
1943 contained an unqualified address, that address was qualified, even if it
1944 did not match any rewriting rules. The underlying bug was that the values
1945 of the flags that permit the existence of unqualified sender and recipient
1946 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1947 non-local messages, and by -bnq for local messages) were not being
1948 preserved with the message after it was received.
1950 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1951 "next input=" as part of the text comprising the host identity instead of
1952 the correct text. The code was using the same buffer for two different
1953 strings. However, depending on which order the printing function evaluated
1954 its arguments, the bug did not always show up. Under Linux, for example, my
1955 test suite worked just fine.
1957 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1958 doesn't work with some older versions of Perl. It has been changed to "my",
1959 which in any case is probably the better facility to use.
1961 15. A really picky compiler found some instances of statements for creating
1962 error messages that either had too many or two few arguments for the format
1965 16. The size of the buffer for calls to the DNS resolver has been increased
1966 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1967 for addresses that have a lot of PTR records. This alleviates a problem; it
1968 does not fully solve it.
1970 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1971 buffer now truncates the list and logs the incident, which is the same
1972 action as happens when Exim is looking up a host name and its aliases.
1973 Previously in this situation something unpredictable would happen;
1974 sometimes it was "internal error: store_reset failed".
1976 18. If a server dropped the connection unexpectedly when an Exim client was
1977 using GnuTLS and trying to read a response, the client delivery process
1978 crashed while trying to generate an error log message.
1980 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1981 string, for example:
1983 warn message = H1: something\nH2: something
1985 the text was added as a single header line from Exim's point of view
1986 though it ended up OK in the delivered message. However, searching for the
1987 second and subsequent header lines using $h_h2: did not work. This has been
1988 fixed. Similarly, if a system filter added multiple headers in this way,
1989 the routers could not see them.
1991 20. Expanded the error message when iplsearch is called with an invalid key to
1992 suggest using net-iplsearch in a host list.
1994 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1995 ACLs are no longer actually imposed (and a message to that effect is
1998 22. If a "gecos" field in a passwd entry contained escaped characters, in
1999 particular, if it contained a \" sequence, Exim got it wrong when building
2000 a From: or a Sender: header from that name. A second bug also caused
2001 incorrect handling when an unquoted " was present following a character
2002 that needed quoting.
2004 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2005 was not being matched caselessly.
2007 24. Arranged for all hyphens in the exim.8 source to be escaped with
2010 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2011 properly. Recipient callout cache records were still being keyed to include
2012 the sender, even when use_sender was set false. This led to far more
2013 callouts that were necessary. The sender is no longer included in the key
2014 when use_sender is false.
2016 26. Added "control = submission" modifier to ACLs.
2018 27. Added the ${base62d: operator to decode base 62 numbers.
2020 28. dnsdb lookups can now access SRV records.
2022 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2023 the configuration file.
2025 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2026 (-v) mode. This makes the output for a verbose queue run more intelligible.
2028 31. Added a use_postmaster feature to recipient callouts.
2030 32. Added the $body_zerocount variable, containing the number of binary zero
2031 bytes in the message body.
2033 33. The time of last modification of the "new" subdirectory is now used as the
2034 "mailbox time last read" when there is a quota error for a maildir
2037 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2039 35. Added +ignore_unknown as a special item in host lists.
2041 36. Code for decoding IPv6 addresses in host lists is now included, even if
2042 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2043 address was recognized as an IP address, but was then not correctly decoded
2044 into binary, causing unexpected and incorrect effects when compared with
2051 1. Very minor rewording of debugging text in manualroute to say "list of
2052 hosts" instead of "hostlist".
2054 2. If verify=header_syntax was set, and a header line with an unqualified
2055 address (no domain) and a large number of spaces between the end of the
2056 name and the colon was received, the reception process suffered a buffer
2057 overflow, and (when I tested it) crashed. This was caused by some obsolete
2058 code that should have been removed. The fix is to remove it!
2060 3. When running in the test harness, delay a bit after writing a bounce
2061 message to get a bit more predictability in the log output.
2063 4. Added a call to search_tidyup() just before forking a reception process. In
2064 theory, someone could use a lookup in the expansion of smtp_accept_max_
2065 per_host which, without the tidyup, could leave open a database connection.
2067 5. Added the variables $recipient_data and $sender_data which get set from a
2068 lookup success in an ACL "recipients" or "senders" condition, or a router
2069 "senders" option, similar to $domain_data and $local_part_data.
2071 6. Moved the writing of debug_print from before to after the "senders" test
2074 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2075 problems for message scanning, either using a data ACL, or using
2076 local_scan() because the Received: header was not generated till after they
2077 were called (in order to set the time as the time of reception completion).
2078 I have revised the way this works. The header is now generated after the
2079 body is received, but before the ACL or local_scan() are called. After they
2080 are run, the timestamp in the header is updated.
2086 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2087 before starting a queue runner without re-exec. This happened only when
2088 deliver_drop_privilege was set or when the Exim user was set to root. The
2089 effect of the bug was that timeouts during subsequent deliveries caused
2090 crashes instead of being properly handled. The handler is now left at its
2091 default (and expected) setting.
2093 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2094 message, again when deliver_drop_privilege is set or Exim is run as root.
2095 The bug described in (1) was not present in this case, but the tidying up
2096 of the other signals was missing. I have made the two cases consistent.
2098 3. The ignore_target_hosts setting on a manualroute router was being ignored
2099 for hosts that were looked up using the /MX notation.
2101 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2104 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2105 operated on the sender address. After changing the $sender_address to <>
2106 for the sender address verify, Exim was re-instated it as the original
2107 (before rewriting) address, but remembering that it had rewritten it, so it
2108 wasn't rewriting it again. This bug also had the effect of breaking the
2109 sender address verification caching when the sender address was rewritten.
2111 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2112 This has been changed so that if the ip literal address matches
2113 ignore_target_hosts, the router declines.
2115 7. Added expansion conditions match_domain, match_address, and match_local_
2116 part (NOT match_host).
2118 8. The placeholder for the Received: header didn't have a length field set.
2120 9. Added code to Exim itself and to exim_lock to test for a specific race
2121 condition that could lead to file corruption when using MBX delivery. The
2122 issue is with the lockfile that is created in /tmp. If this file is removed
2123 after a process has opened it but before that process has acquired a lock,
2124 there is the potential for a second process to recreate the file and also
2125 acquire a lock. This could lead to two Exim processes writing to the file
2126 at the same time. The added code performs the same test as UW imapd; it
2127 checks after acquiring the lock that its file descriptor still refers to
2128 the same named file.
2130 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2131 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2134 11. Added the smtp_active_hostname option. If used, this will typically be made
2135 to depend on the incoming interface address. Because $interface_address is
2136 not set up until the daemon has forked a reception process, error responses
2137 that can happen earlier (such as "too many connections") no longer contain
2140 12. If an expansion in a condition on a "warn" statement fails because a lookup
2141 defers, the "warn" statement is abandoned, and the next ACL statement is
2142 processed. Previously this caused the whole ACL to be aborted.
2144 13. Added the iplsearch lookup type.
2146 14. Added ident_timeout as a log selector.
2148 15. Added tls_certificate_verified as a log selector.
2150 16. Added a global option tls_require_ciphers (compare the smtp transport
2151 option of the same name). This controls incoming TLS connections.
2153 17. I finally figured out how to make tls_require_ciphers do a similar thing
2154 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2155 before starting the TLS session.
2157 18. Tabs are now shown as \t in -bP output.
2159 19. If the log selector return_path_on_delivery was set, Exim crashed when
2160 bouncing a message because it had too many Received: header lines.
2162 20. If two routers both had headers_remove settings, and the first one included
2163 a superfluous trailing colon, the final name in the first list and the
2164 first name in the second list were incorrectly joined into one item (with a
2165 colon in the middle).
2171 1. Added -C and -D options to the exinext utility, mainly to make it easier
2172 to include in the automated testing, but these could be helpful when
2173 multiple configurations are in use.
2175 2. The exinext utility was not formatting the output nicely when there was
2176 an alternate port involved in the retry record key, nor when there was a
2177 message id as well (for retries that were specific to a specific message
2178 and a specific host). It was also confused by IPv6 addresses, because of
2179 the additional colons they contain. I have fixed the IPv4 problem, and
2180 patched it up to do a reasonable job for IPv6.
2182 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2183 delivery, the log line now contains "pipelined" if PIPELINING was used.
2185 4. An SMTP transport process used to panic and die if the bind() call to set
2186 an explicit outgoing interface failed. This has been changed; it is now
2187 treated in the same way as a connect() failure.
2189 5. A reference to $sender_host_name in the part of a conditional expansion
2190 that was being skipped was still causing a DNS lookup. This no longer
2193 6. The def: expansion condition was not recognizing references to header lines
2194 that used bh_ and bheader_.
2196 7. Added the _cache feature to named lists.
2198 8. The code for checking quota_filecount in the appendfile transport was
2199 allowing one more file than it should have been.
2201 9. For compatibility with Sendmail, the command line option
2209 and sets the incoming protocol and host name (for trusted callers). The
2210 host name and its colon can be omitted when only the protocol is to be set.
2211 Note the Exim already has two private options, -pd and -ps, that refer to
2212 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2213 "s", but I don't think that's a major issue.
2215 10. A number of refactoring changes to the code, none of which should affect
2218 (a) The number of logging options was getting close to filling up the
2219 32-bit word that was used as a bit map. I have split them into two classes:
2220 those that are passed in the argument to log_write(), and those that are
2221 only ever tested independently outside of that function. These are now in
2222 separate 32-bit words, so there is plenty of room for expansion again.
2223 There is no change in the user interface or the logging behaviour.
2225 (b) When building, for example, log lines, the code previously used a
2226 macro that called string_cat() twice, in order to add two strings. This is
2227 not really sufficiently general. Furthermore, there was one instance where
2228 it was actually wrong because one of the argument was used twice, and in
2229 one call a function was used. (As it happened, calling the function twice
2230 did not affect the overall behaviour.) The macro has been replaced by a
2231 function that can join an arbitrary number of extra strings onto a growing
2234 (c) The code for expansion conditions now uses a table and a binary chop
2235 instead of a serial search (which was left over from when there were very
2236 few conditions). Also, it now recognizes conditions like "pam" even when
2237 the relevant support is not compiled in: a suitably worded error message is
2238 given if an attempt is made to use such a condition.
2240 11. Added ${time_interval:xxxxx}.
2242 12. A bug was causing one of the ddress fields not to be passed back correctly
2243 from remote delivery subprocesses. The field in question was not being
2244 subsequently used, so this caused to problems in practice.
2246 13. Added new log selectors queue_time and deliver_time.
2248 14. Might have fixed a bug in maildirsizefile handling that threw up
2249 "unexpected character" debug warnings, and recalculated the data
2250 unnecessarily. In any case, I expanded the warning message to give more
2253 15. Added the message "Restricted characters in address" to the statements in
2254 the default ACL that block characters like @ and % in local parts.
2256 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2257 Three changes have been made:
2259 (a) There was a serious bug; a negative response to MAIL caused the whole
2260 recipient domain to be cached as invalid, thereby blocking all messages
2261 to all local parts at the same domain, from all senders. This bug has
2262 been fixed. The domain is no longer cached after a negative response to
2263 MAIL if the sender used is not empty.
2265 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2268 (c) A new callout option, "use_sender" has been added for people who want
2269 the modified behaviour.
2275 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2278 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2279 indeed breaks things for older releases.
2281 3. Added additional logging to the case where there is a problem reading data
2282 from a filter that is running in a subprocess using a pipe, in order to
2283 try to track down a specific problem.
2285 4. Testing facility fudge: when running in the test harness and attempting
2286 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2287 getting "No route to host". Convert this to a timeout.
2289 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2292 6. Some OS don't have socklen_t but use size_t instead. This affects the
2293 fifth argument of getsockopt() amongst other things. This is now
2294 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2295 can be set for individual OS. I have set it for SunOS5, OSF1, and
2296 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2297 some earlier ones do not.
2299 7. Change 4.30/15 was not doing the test caselessly.
2301 8. The standard form for an IPv6 address literal was being rejected by address
2302 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2303 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2304 this, as well as the form without the "IPv6" on the front (but only when
2305 address literals are enabled, of course).
2307 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2309 10. Exim crashed if a message with an empty sender address specified by -f
2310 encountered a router with an errors_to setting. This could be provoked only
2311 by a command such as
2315 where an empty string was supplied; "<>" did not hit this bug.
2317 11. Installed PCRE release 4.5.
2319 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2320 remained set. It is now erased.
2322 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2323 times from message ids (which are base 36 rather than the normal 62).
2325 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2326 were being counted as actual protocol errors, and logged if the log
2327 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2328 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2329 and DATA following a set of rejected RCPTs do not count as protocol errors.
2330 In other words, Exim assumes they were pipelined, though this may not
2331 actually be the case. Of course, in all cases the client gets an
2332 appropriate error code.
2334 15. If a lookup fails in an ACL condition, a message about the failure may
2335 be available; it is used if testing the ACL cannot continue, because most
2336 such messages specify what the cause of the deferral is. However, some
2337 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2338 that caused an old message to be retained and used if a later statement
2339 caused a defer, replacing the real cause of the deferral.
2341 16. If an IP address had so many PTR records that the DNS lookup buffer
2342 was not large enough to hold them, Exim could crash while trying to process
2343 the truncated data. It now detects and logs this case.
2345 17. Further to 4.21/58, another change has been made: if (and only if) the
2346 first line of a message (the first header line) ends with CRLF, a bare LF
2347 in a subsequent header line has a space inserted after it, so as not to
2348 terminate the header.
2350 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2351 unnecessarily, used atoi() instead of strtol(), and didn't check the
2352 termination when getting file sizes from file names by regex.
2354 19. Completely re-implemented the support for maildirsize files, in the light
2355 of a number of problems with the previous contributed implementation
2356 (4.30/29). In particular:
2358 . If the quota is zero, the maildirsize file is maintained, but no quota is
2361 . If the maildir directory does not exist, it is created before any attempt
2362 to write a maildirsize file.
2364 . The quota value in the file is just a cache; if the quota is changed in
2365 the transport, the new value overrides.
2367 . A regular expression is available for excluding directories from the
2370 20. The autoreply transport checks the characters in options that define the
2371 message's headers; it allows continued headers, but it was checking with
2372 isspace() after an embedded newline instead of explicitly looking for a
2375 21. If all the "regular" hosts to which an address was routed had passed their
2376 expiry times, and had not reached their retry times, the address was
2377 bounced, even if fallback hosts were defined. Now Exim should go on to try
2380 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2381 equivalent code in the SMTP transport. Some hosts send humungous responses
2382 to HELO/EHLO, more than 1024 it seems.
2384 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2385 gives compiler warnings in some environments. I've now done it "properly",
2388 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2389 (because of problems with the built-in one) was declared to return uschar *
2390 instead of char *, causing compiler failure.
2392 25. Fixed a file descriptor leak when processing alias/forward files.
2394 26. Fixed a minor format string issue in dbfn.c.
2396 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2398 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2399 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2402 29. When a sender address is verified, it is cached, to save repeating the test
2403 when there is more than one recipient in a message. However, when the
2404 verification involves a callout, it is possible for different callout
2405 options to be set for different recipients. It is too complicated to keep
2406 track of this in the cache, so now Exim always runs a verification when a
2407 callout is required, relying on the callout cache for the optimization.
2408 The overhead is duplication of the address routing, but this should not be
2411 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2412 to be verified with callout=postmaster, and the main callout worked but the
2413 postmaster check failed, the verification correctly failed. However, if a
2414 subsequent RCPT command asked for sender verification *without* the
2415 postmaster check, incorrect caching caused this verification also to fail,
2418 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2419 it was not caching the DNS options (qualify_single, search_parents) that
2420 were used when the lookup failed. A subsequent lookup with different
2421 options therefore always gave the same answer, though there were cases
2422 where it should not have. (Example: a "domains = !$mx_any" option on a
2423 dnslookup router: the "domains" option is always processed without any
2424 widening, but the router might have qualify_single set.) Now Exim uses the
2425 cached value only when the same options are set.
2427 32. Added John Jetmore's "exipick" utility to the distribution.
2429 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2430 than a timeout (e.g. a certificate is required, and is not provided), an
2431 Exim server now closes the connection immediately. Previously it waited for
2432 the client to close - but if the client is SSL, it seems that they each
2433 wait for each other, leading to a delay before one of them times out.
2435 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2436 maintained 0.8.x compatibility because I don't think many are using it, and
2437 it is clearly obsolete.
2439 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2442 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2443 client certificate was expired. A simple patch fixes this, though I don't
2444 understand the full logic of why the verify callback is called multiple
2447 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2448 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2449 which causes problems with some clients (such as the Certicom SSL Plus
2450 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2451 disables the coutermeasure allowing Eudora to connect."
2453 38. Exim was not checking that a write() to a log file succeeded. This could
2454 lead to Bad Things if a log got too big, in particular if it hit a file
2455 size limit. Exim now panics and dies if it cannot write to a log file, just
2456 as it does if it cannot open a log file.
2458 39. Modified OS/Makefile-Linux so that it now contains
2460 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2462 The two -D definitions ensure that Exim is compiled with large file
2463 support, which makes it possible to handle log files that are bigger than
2466 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2467 instance) a domain was checked against a named list that involved a lookup,
2468 causing $domain_data to be set, then another domain was checked against the
2469 same list, then the first domain was re-checked, the value of $domain_data
2470 after the final check could be wrong. In particular, if the second check
2471 failed, it could be set empty. This bug probably also applied to
2474 41. The strip_trailing_dot option was not being applied to the address given
2475 with the -f command-line option.
2477 42. The code for reading a message's header from the spool was incrementing
2478 $received_count, but never initializing it. This meant that the value was
2479 incorrect (doubled) while delivering a message in the same process in which
2480 it was received. In the most common configuration of Exim, this never
2481 happens - a fresh exec is done - but it can happen when
2482 deliver_drop_privilege is set.
2484 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2485 it now includes up to 150 characters of the unexpected data in the log
2488 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2489 and building data strings. The size of both of these buffers was 10 000
2490 bytes - far larger than anybody would *ever* want, thought I. Needless to
2491 say, somebody hit the limit. I have increased the maximum line length to
2492 20 000 and the maximum data length of concatenated lines to 100 000. I have
2493 also fixed two bugs, because there was no checking on these buffers. Tsk,
2494 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2495 buffer is too small.
2497 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2498 lsearch lookups. Now it does.
2500 46. When parsing a route_list item in a manualroute router, a fixed-length
2501 buffer was used for the list of hosts. I made this 1024 bytes long,
2502 thinking that nobody would ever have a list of hosts that long. Wrong.
2503 Somebody had a whole pile of complicated expansion conditions, and the
2504 string was silently truncated, leading to an expansion error. It turns out
2505 that it is easier to change to an unlimited length (owing to other changes
2506 that have happened since this code was originally written) than to build
2507 structure for giving a limitation error. The length of the item that
2508 expands into the list of hosts is now unlimited.
2510 47. The lsearch lookup could not handle data where the length of text line was
2511 more than 4095 characters. Such lines were truncated, leading to shortened
2512 data being returned. It should now handle lines of any length.
2514 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2515 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2518 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2519 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2520 no longer happens in comment lines. A list of replacements is now placed
2521 at the head of all of the source files, except those whose only change is
2522 to replace PERL_COMMAND in the very first #! line.
2524 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2525 messages on the queue, with a bottom-up merge sort, using code contributed
2526 by Michael Haardt. This should make operations like -bp somewhat faster on
2527 large queues. It won't affect queue runners, except when queue_run_in_order
2530 51. Installed eximstats 1.31 in the distribution.
2532 52. Added support for SRV lookups to the dnslookup router.
2534 53. If an ACL referred to $message_body or $message_body_end, the value was not
2535 reset for any messages that followed in the same SMTP session.
2537 54. The store-handling optimization for building very long strings was not
2538 differentiating between the different store pools. I don't think this
2539 actually made any difference in practice, but I've tidied it.
2541 55. While running the routers to verify a sender address, $sender_address
2542 was still set to the sender address. This is wrong, because when routing to
2543 send a bounce to the sender, it would be empty. Therefore, I have changed
2544 it so that, while verifying a sender address, $sender_address is set to <>.
2545 (There is no change to what happens when verifying a recipient address.)
2547 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2548 target A or AAAA records (if not already returned) without resetting the
2549 qualify_single or search_parents options of the DNS resolver. These are
2550 inappropriate in this case because the targets of MX and SRV records must
2551 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2552 target that, when qualified, matched something in the local domain. These
2553 two options are now turned off when doing these lookups.
2555 57. It seems that at least some releases of Reiserfs (which does not have the
2556 concept of a fixed number of inodes) returns zero and not -1 for the
2557 number of available inodes. This interacted badly with check_spool_inodes,
2558 which assumed that -1 was the "no such thing" setting. What I have done is
2559 to check that the total number of inodes is greater than zero before doing
2560 the test of how many are available.
2562 58. When a "warn" ACL statement has a log_message modifier, the message is
2563 remembered, and not repeated. This is to avoid a lot of repetition when a
2564 message has many recipients that cause the same warning to be written.
2565 Howewer, Exim was preserving the list of already written lines for an
2566 entire SMTP session, which doesn't seem right. The memory is now reset if a
2567 new message is started.
2569 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2570 debugging output unless log_rewrite was also set.
2572 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2573 of (int)(handle) when we know that handle contains (void *)(-1).
2575 61. The Exim daemon panic-logs an error return when it closes the incoming
2576 connection. However "connection reset by peer" seems to be common, and
2577 isn't really an error worthy of noting specially, so that particular error
2580 62. When Exim is trying to find all the local interfaces, it used to panic and
2581 die if the ioctl to get the interface flags failed. However, it seems that
2582 on at least one OS (Solaris 9) it is possible to have an interface that is
2583 included in the list of interfaces, but for which you get a failure error
2584 for this call. This happens when the interface is not "plumbed" into a
2585 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2586 failure of the "get flags" call assumes that the interface is down.
2588 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2589 makes life easier for people who are doing arithmetic on fields extracted
2590 from dates, where you often get leading zeros that should not be
2591 interpreted as octal.
2593 64. Added qualify_domain to the redirect router, to override the global
2596 65. If a pathologically long header line contained very many addresses (the
2597 report of this problem mentioned 10 000) and each of them was rewritten,
2598 Exim could use up a very large amount of memory. (It kept on making new
2599 copies of the header line as it rewrote, and never released the old ones.)
2600 At the expense of a bit more processing, the header rewriting function has
2601 been changed so that it no longer eats memory in this way.
2603 66. The generation of the Received: header has been moved from the time that a
2604 message starts to be received, to the time that it finishes. The timestamp
2605 in the Received: header should now be very close to that of the <= log
2606 line. There are two side-effects of this change:
2608 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2609 logged header lines no longer include the local Received: line, because
2610 it has not yet been created. The same applies to a copy of the message
2611 that is returned to a non-SMTP sender when a message is rejected.
2613 (b) When a filter file is tested using -bf, no additional Received: header
2614 is added to the test message. After some thought, I decided that this
2617 This change does not affect the value of $received_for. It is still set
2618 after address rewriting, but before local_scan() is called.
2620 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2622 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2623 gave an unhelpful panic error message, and a defer error. I have managed to
2624 change this behaviour so that it now rejects any supplied certificate,
2625 which seems right, as the list of acceptable certificates is empty.
2627 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2628 gave an unhelpful defer error. I have not managed to make this reject any
2629 supplied certificates, but the error message it gives is "no certificate
2630 supplied", which is not helpful.
2632 70. exigrep's output now also includes lines that are not associated with any
2633 message, but which match the given pattern. Implemented by a patch from
2634 Martin Sluka, which also tidied up the Perl a bit.
2636 71. Recipient callout verification, like sender verification, was using <> in
2637 the MAIL FROM command. This isn't really the right thing, since the actual
2638 sender may affect whether the remote host accepts the recipient or not. I
2639 have changed it to use the actual sender in the callout; this means that
2640 the cache record is now keyed on a recipient/sender pair, not just the
2641 recipient address. There doesn't seem to be a real danger of callout loops,
2642 since a callout by the remote host to check the sender would use <>.
2643 [SEE ABOVE: changed after hitting problems.]
2645 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2646 temporary errors. However, in the case of such a code being given after
2647 the end of a data transmission (i.e. after ".") Exim was failing to write
2648 a retry record for the message. (Yes, there was some broken host that was
2649 actually sending 8xx at this point.)
2651 73. An unknown lookup type in a host list could cause Exim to panic-die when
2652 the list was checked. (An example that provoked this was putting <; in the
2653 middle of a list instead of at the start.) If this happened during a DATA
2654 ACL check, a -D file could be left lying around. This kind of configuration
2655 error no longer causes Exim to die; instead it causes a defer errror. The
2656 incident is still logged to the main and panic logs.
2658 74. Buglet left over from Exim 3 conversion. The message "too many messages
2659 in one connection" was written to the rejectlog but not the mainlog, except
2660 when address rewriting (yes!) was being logged.
2662 75. Added write_rejectlog option.
2664 76. When a system filter was run not as root (that is, when system_filter_user
2665 was set), the values of the $n variables were not being returned to the
2666 main process; thus, they were not subsequently available in the $sn
2669 77. Added +return_path_on_delivery log selector.
2671 78. A connection timeout was being treated differently from recipients deferred
2672 when testing hosts_max_try with a message that was older than the host's
2673 retry timeout. (The host should not be counted, thus allowing all hosts to
2674 be tried at least once before bouncing.) This may have been the cause of an
2675 occasionally reported bug whereby a message would remain on the queue
2676 longer than the retry timeout, but would be bounced if a delivery was
2677 forced. I say "may" because I never totally pinned down the problem;
2678 setting up timeout/retry tests is difficult. See also the next item.
2680 79. The ultimate address timeout was not being applied to errors that involved
2681 a combination of host plus message (for example, a timeout on a MAIL
2682 command). When an address resolved to a number of possible hosts, and they
2683 were not all tried for each delivery (e.g. because of hosts_max_try), a
2684 message could remain on the queue longer than the retry timeout.
2686 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2689 81. Fixed an obscure SMTP outgoing bug which required at least the following
2690 conditions: (a) there was another message waiting for the same server;
2691 (b) the server returned 5xx to all RCPT commands in the first message so
2692 that the message was not completed; (c) the server dropped the connection
2693 or gave a negative response to the RSET that Exim sends to abort the
2694 transaction. The observed case was a dropped connection after DATA that had
2695 been sent in pipelining mode. That is, the server had advertised PIPELINING
2696 but was not implementing it correctly. The effect of the bug was incorrect
2697 behaviour, such as trying another host, and this could lead to a crash.
2703 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2704 and daemon.c were passed as pointers to ints; they should have been
2705 pointers to socklen_t variables (which are typically unsigned ints).
2707 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2710 3. Fixed a really odd bug that affected only the testing scheme; patching a
2711 certain fixed string in the binary changed the value of another string that
2712 happened to be identical to the end of the original first string.
2714 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2715 name", it returns that address as the IP address. On some operating
2716 systems (e.g. Solaris), it also passes back the IP address string as the
2717 "host name". However, on others (e.g. Linux), it passes back an empty
2718 string. Exim wasn't checking for this, and was changing the host name to an
2719 empty string, assuming it had been canonicized.
2721 5. Although rare, it is permitted to have more than one PTR record for a given
2722 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2723 all the names associated with an address, because they do in Solaris.
2724 However, it seems that they do not in Linux for data that comes from the
2725 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2726 I found this out when I moved to a new Linux workstation and tried to run
2727 the Exim test suite.
2729 To get round this problem I have changed the code so that it now does its
2730 own call to the DNS to look up PTR records when searching for a host name.
2731 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2732 addresses that are only in /etc/hosts are still found.
2734 This behaviour is, however, controlled by an option called host_lookup_
2735 order, which defaults to "bydns:byaddr". If people want to use the other
2736 order, or indeed, just use one or the other means of lookup, they can
2737 specify it in this variable.
2739 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2740 some operating systems, this comes back from gethostbyaddr() as an empty
2741 string, and this is what Exim used to test for. However, it seems that in
2742 other systems, "." is yielded. Exim now tests for this case too.
2744 7. The values of check_spool_space and check_log_space are now held internally
2745 as a number of kilobytes instead of an absolute number of bytes. If a
2746 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2747 kilobyte. This means that much larger values can be stored.
2749 8. Exim monitor: an attempt to get the action menu when not actually pointing
2750 at a message produces an empty menu entitled "No message selected". This
2751 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2752 no entries in it ("Shell widget menu has zero width and/or height"). So I
2753 have added a single, blank menu entry in this case.
2755 9. Added ${quote_local_part.
2757 10. MIME decoding is now applied to the contents of Subject: header lines when
2760 11. Now that a reference to $sender_host_address automatically causes a reverse
2761 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2762 host lookup before query-style lookups in lists that might use this
2763 variable. This has therefore been abolished, and the "net-" prefix is no
2764 longer necessary for query-style lookups.
2766 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2767 have been a typo for LFLAGS, so it has been changed.
2769 13. The install script calls Exim with "-C /dev/null" in order to find the
2770 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2771 to be output. Howeve, since Exim outputs its version number before the
2772 error, it didn't break the script. It just looked ugly. I fixed this by
2773 always allowing "-C /dev/null" if the caller is root.
2775 14. Ignore overlarge ACL variable number when reading spool file - insurance
2776 against a later release with more variables having written the file.
2778 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2779 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2780 this, as well as the form without the "IPv6" on the front.
2782 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2783 OS/Makefile-Darwin file.
2785 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2786 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2788 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2789 variable. However, some OS use ints for inodes. I've added cast to long int
2790 to get rid of the compiler warning.
2792 19. I had forgotten to lock out "/../" in configuration file names when
2793 ALT_CONFIG_PREFIX was set.
2795 20. Routers used for verification do not need to specify transports. However,
2796 if such a router generated a host list, and callout was configured, Exim
2797 crashed, because it could not find a port number from the (non-existent)
2798 transport. It now assumes port 25 in this circumstance.
2800 21. Added the -t option to exigrep.
2802 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2803 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2804 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2805 them is set, LOOKUP_LSEARCH is forced.
2807 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2809 24. Added sender and host information to the "rejected by local_scan()" log
2810 line; previously there was no indication of these.
2812 25. Added .include_if_exists.
2814 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2815 Linux. It turns out that not all file systems support this. Apparently some
2816 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2817 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2818 not supported on the file descriptor, is now ignored when Exim is trying to
2819 sync a directory. This applies only to Linux.
2821 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2823 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2824 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2827 29. Added support for maildirsize files from supplied patch (modified a bit).
2829 30. The use of :fail: followed by an empty string could lead Exim to respond to
2830 sender verification failures with (e.g.):
2832 550 Verification failed for <xxx>
2833 550 Sender verify failed
2835 where the first response line was missing the '-' that indicates it is not
2836 the final line of the response.
2838 31. The loop for finding the name of the user that called Exim had a hardwired
2839 limit of 10; it now uses the value of finduser_retries, which is used for
2840 all other user lookups.
2842 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2845 33. Exim was neglecting to zero errno before one call of strtol() when
2846 expanding a string and expecting an integer value. On some systems this
2847 resulted in spurious "integer overflow" errors. Also, it was casting the
2848 result into an int without checking.
2850 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2851 did not work. The code looks as if it has *never* worked, though it appears
2852 to have been documented since at least releast 1.62. I have made it work.
2854 35. The "timeout_DNS" error in retry rules, also documented since at least
2855 1.62, also never worked. As it isn't clear exactly what this means, and
2856 clearly it isn't a major issue, I have abolished the feature by treating it
2857 as "timeout", and writing a warning to the main and panic logs.
2859 36. The display of retry rules for -brt wasn't always showing the error code
2862 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2863 timeout_connect_A, timeout_connect_MX.
2865 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2866 to the empty sender.
2868 39. The daemon was not analysing the content of -oX till after it had closed
2869 stderr and disconnected from the controlling terminal. This meant that any
2870 syntax errors were only noted on the panic log, and the return code from
2871 the command was 0. By re-arranging the code a little, I've made the
2872 decoding happen first, so such errors now appear on stderr, and the return
2873 code is 1. However, the actual setting up of the sockets still happens in
2874 the disconnected process, so errors there are still only recorded on the
2877 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2878 connections (as happens on some IP stacks) was logged at start up time as
2879 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2880 it from "IPv6 and IPv4", which means that two separate sockets are being
2883 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2884 says whether AF_INET or AF_INET6 was passed as an argument.
2886 42. Exiwhat output was messed up when time zones were included in log
2889 43. Exiwhat now gives more information about the daemon's listening ports,
2890 and whether -tls-on-connect was used.
2892 44. The "port" option of the smtp transport is now expanded.
2894 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2895 silently ignored. Now an error message is written to the main and panic
2898 46. There's a new ACL modifier called "logwrite" which writes to a log file
2899 as soon as it is encountered.
2901 47. Added $local_user_uid and $local_user_gid at routing time.
2903 48. Exim crashed when trying to verify a sender address that was being
2906 49. Exim was recognizing only a space character after ".include". It now also
2907 recognizes a tab character.
2909 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2910 extracting the relevant information from the specification. The man page no
2911 longer contains scrambled data for the -d option, and I've added a section
2912 at the front about calling Exim under different names.
2914 51. Added "extra_headers" argument to the "mail" command in filter files.
2916 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2919 53. Installed eximstats 1.29.
2921 54. Added transport_filter_timeout as a generic transport option.
2923 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2924 Cc: header lines. This was required by RFC 822, but it not required by RFC
2927 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2928 incoming messages that did not have them. Now it does so only if the
2929 message originates locally, that is, if there is no associated remote host
2930 address. When Resent- header lines are present, this applies to the Resent-
2931 lines rather than the non-Resent- lines.
2933 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2934 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2936 58. Messages for configuration errors now include the name of the main
2937 configuration file - useful now that there may be more than one file in a
2938 list (.included file names were always shown).
2940 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2941 for those rare installations that do not start the daemon as root or run it
2942 setuid root. I've cut out the call to initgroups() if the daemon is not
2945 60. The Exim user and group can now be bound into the binary as text strings
2946 that are looked up at the start of Exim's processing.
2948 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2950 62. Added $mailstore_basename variable.
2952 63. Installed patch to sieve.c from Michael Haardt.
2954 64. When Exim failed to open the panic log after failing to open the main log,
2955 the original message it was trying to log was written to stderr and debug
2956 output, but if they were not available (the usual case in production), it
2957 was lost. Now it is written to syslog before the two lines that record the
2958 failures to open the logs.
2960 65. Users' Exim filters run in subprocesses under the user's uid. It is
2961 possible for a "deliver" command or an alias in a "personal" command to
2962 provoke an address rewrite. If logging of address rewriting is configured,
2963 this fails because the process is not running as root or exim. There may be
2964 a better way of dealing with this, but for the moment (because 4.30 needs
2965 to be released), I have disabled address rewrite logging when running a
2966 filter in a non-root, non-exim process.
2972 1. The buildconfig auxiliary program wasn't quoting the value set for
2973 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2974 not defined. This bug was masked in 4.22 by the effect that was fixed in
2977 2. Some messages that were rejected after a message id was allocated were
2978 shown as "incomplete" by exigrep. It no longer does this for messages that
2979 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2981 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2982 have allow_domain_literals set, the ID did not get logged in the <= line.
2983 Domain literals are now always recognized in Message-ID: header lines.
2985 4. The first argument for a ${extract expansion item is the key name or field
2986 number. Leading and trailing spaces in this item were not being ignored,
2987 causing some misleading effects.
2989 5. When deliver_drop_privilege was set, single queue runner processes started
2990 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2991 same command in the process it spins off) were not dropping privilege.
2993 6. When the daemon running as "exim" started a queue runner, it always
2994 re-executed Exim in the spun-off process. This is a waste of effort when
2995 deliver_drop_privilege is set. The new process now just calls the
2996 queue-runner function directly.
3002 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3005 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3006 ignored. Though the use of -oP was forcing the writing of a pid file, it
3007 was always written to the default place.
3009 3. If the message "no IP address found for host xxxx" is generated during
3010 incoming verification, it is now followed by identification of the incoming
3011 connection (so you can more easily find what provoked it).
3013 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3015 5. Added some features to "harden" Exim a bit more against certain attacks:
3017 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3018 be put in Local/Makefile. This is like the never_users runtime option,
3019 but it cannot be overridden. The default setting is "root".
3021 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3022 prefix string with which any file named in a -C command line option
3025 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3026 is retained for -C and -D only if the caller of Exim is root. Without
3027 it, the exim user may also use -C and -D and retain privilege.
3029 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3030 command line option is disabled.
3032 6. Macro names set by the -D option must start with an upper case letter, just
3033 like macro names defined in the configuration file.
3035 7. Added "dereference=" facility to LDAP.
3037 8. Two instances of the typo "uknown" in the source files are fixed.
3039 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3040 the Configure-Makefile script screwed up while processing it.
3042 10. Incorporated PCRE 4.4.
3044 11. The SMTP synchronization check was not operating right at the start of an
3045 SMTP session. For example, it could not catch a HELO sent before the client
3046 waited for the greeting. There is now a check for outstanding input at the
3047 point when the greeting is written. Because of the duplex, asynchronous
3048 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3049 way, but not yet received, when the check is performed.
3051 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3052 on TCP/IP sockets, because this apparently causes some broken clients to
3055 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3056 unchanged) from the Cygwin maintainer.
3058 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3059 was supported instead of testing for mbx. Effectively a typo.
3061 15. The spa authenticator server code was not checking that the input it
3062 received was valid base64.
3064 16. The debug output line for the "set" modifier in ACLs was not showing the
3065 name of the variable that was being set.
3067 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3069 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3070 lookup on its own. Something else was needed to trigger the lookup. For
3071 example, a match in host_lookup or the need for a host name in a host list.
3072 Now, if $sender_host_name is referenced and the host name has not yet been
3073 looked up, a lookup is performed. If the lookup fails, the variable remains
3074 empty, and $host_lookup_failed is set to "1".
3076 19. Added "eqi" as a case-independent comparison operator.
3078 20. The saslauthd authentication condition could segfault if neither service
3079 nor realm was specified.
3081 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3082 error message that was logged was misleading, and incoming SMTP
3083 connections were dropped. The message is now more accurate, and temporary
3084 errors are given to SMTP connections.
3086 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3087 (except RSET) with a 421 error, until QUIT is received. However, it was
3088 failing to send a response to QUIT.
3090 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3091 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3092 ACL is now run after the helo_try_verify_hosts code.
3094 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3095 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3096 case-independent, so other case variants are also recognized). Apparently
3097 some systems use these upper case variants.
3099 25. If more than two messages were waiting for the same host, and a transport
3100 filter was specified for the transport, Exim sent two messages over the
3101 same TCP/IP connection, and then failed with "socket operation on non-
3102 socket" when it tried to send the third.
3104 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3106 27. The extern definition of crypt16() in expand.c was not being excluded when
3107 the OS had its own crypt16() function.
3109 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3110 as a preferred synonym for return_size_limit, both as an option and as an
3113 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3115 30. Changed the default configuration ACL to relax the local part checking rule
3116 for addresses that are not in any local domains. For these addresses,
3117 slashes and pipe symbols are allowed within local parts, but the sequence
3118 /../ is explicitly forbidden.
3120 31. SPA server authentication was not clearing the challenge buffer before
3123 32. log_message in a "warn" ACL statement was writing to the reject log as
3124 well as to the main log, which contradicts the documentation and doesn't
3125 seem right (because no rejection is happening). So I have stopped it.
3127 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3128 However, I am unable to do any testing of this.
3130 34. Fixed an infelicity in the appendfile transport. When checking directories
3131 for a mailbox, to see if any needed to be created, it was accidentally
3132 using path names with one or more superfluous leading slashes; tracing
3133 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3135 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3136 contents are added to the log line that is written for every discarded
3137 recipient. (Previously a log_message setting was ignored.)
3139 36. The ${quote: operator now quotes the string if it is empty.
3141 37. The install script runs exim in order to find its version number. If for
3142 some reason other than non-existence or emptiness, which it checks, it
3143 could not run './exim', it was installing it with an empty version number,
3144 i.e. as "exim-". This error state is now caught, and the installation is
3147 38. An argument was missing from the function that creates an error message
3148 when Exim fails to connect to the socket for saslauthd authentication.
3149 This could cause Exim to crash, or give a corrupted message.
3151 39. Added isip, isip4, and isip6 to ${if conditions.
3153 40. The ACL variables $acl_xx are now saved with the message, and can be
3154 accessed later in routers, transports, and filters.
3156 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3157 strings in the file are not string-expanded.
3159 42. If a MAIL command specified a SIZE value that was too large to fit into an
3160 int variable, the check against message_size_limit failed. Such values are
3161 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3162 day this will have to be increased, but I don't think I want to be around
3163 when emails are that large.
3170 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3171 iconv() is not standard in FreeBSD.
3173 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3174 IPv6 enabled. The observed symptom was a segmentation fault on return from
3175 the function os_common_find_running_interfaces() in src/os.c.
3177 3. In the check_special_case() function in daemon.c I had used "errno" as an
3178 argument name, which causes warnings on some systems. This was basically a
3179 typo, since it was named "eno" in the comments!
3181 4. The code that waits for the clock to tick (at a resolution of some fraction
3182 of a second) so as to ensure message-id uniqueness was always waiting for
3183 at least one whole tick, when it could have waited for less. [This is
3184 almost certainly not relevant at current processor speeds, where it is
3185 unlikely to ever wait at all. But we try to future-proof.]
3187 5. The function that sleeps for a time interval that includes fractions of a
3188 second contained a race. It did not block SIGALRM between setting the
3189 timer, and suspending (a couple of lines later). If the interval was short
3190 and the sigsuspend() was delayed until after it had expired, the suspension
3191 never ended. On busy systems this could lead to processes getting stuck for
3194 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3195 process, before it forks any delivery processes. The open lookup caching
3196 mechanism meant that the open file or database connection was passed into
3197 the delivery process. The problem was that delivery processes always tidy
3198 up cached lookup data. This could cause a problem for the next delivery
3199 process started by the queue runner, because the external queue runner
3200 process does not know about the closure. So the next delivery process
3201 still has data in the lookup cache. In the case of a file lookup, there was
3202 no problem because closing a file descriptor in a subprocess doesn't affect
3203 the parent. However, if the lookup was caching a connection to a database,
3204 the connection was closed, and the second delivery process was likely to
3205 see errors such as "PGSQL: query failed: server closed the connection
3206 unexpectedly". The problem has been fixed by closing all cached lookups
3207 in a queue runner before running a delivery process.
3209 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3210 seem to have the "const" qualifier which it has on other OS. I've
3213 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3214 *of the same type* (client or server) with the same public name that an
3215 error should be diagnosed.
3217 9. When Exim looked up a host name for an IP address, but failed to find the
3218 original IP address when looking up the host name (a safety check), it
3219 output the message "<ip address> does not match any IP for NULL", which was
3220 confusing, to say the least. The bug was that the host name should have
3221 appeared instead of "NULL".
3223 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3224 user that is built into the binary, and the -C or -D options is used, root
3225 privilege is dropped before the configuration file is read. In addition,
3226 logging is switched to stderr instead of the normal log files. If the
3227 configuration then re-defines the Exim user, the unprivileged environment
3228 is probably not what is expected, so Exim logs a panic warning message (but
3231 However, if deliver_drop_privilege is set, the unprivileged state may well
3232 be exactly what is intended, so the warning has been cut out in that case,
3233 and Exim is allowed to try to write to its normal log files.
3239 1. smtp_return_error_details was not giving details for temporary sender
3240 or receiver verification errors.
3242 2. Diagnose a configuration error if two authenticators have the same public
3245 3. Exim used not to create the message log file for a message until the first
3246 delivery attempt. This could be confusing when incoming messages were held
3247 for policy or load reasons. The message log file is now created at the time
3248 the message is received, and an initial "Received" line is written to it.
3250 4. The automatically generated man page for command line options had a minor
3251 bug that caused no ill effects; however, a more serious problem was that
3252 the procedure for building the man page automatically didn't always
3253 operate. Consequently, release 4.20 contains an out-of-date version. This
3254 shouldn't happen again.
3256 5. When building Exim with embedded Perl support, the script that builds the
3257 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3258 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3260 6. The freeze_tell option was not being used for messages that were frozen on
3261 arrival, either by an ACL or by local_scan().
3263 7. Added the smtp_incomplete_transaction log selector.
3265 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3266 was accepting AUTH without a new EHLO.
3268 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3269 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3272 10. Logging of TCP/IP connections (when configured) now happens in the main
3273 daemon process instead of the child process, so that the TCP/IP connection
3274 count is more accurate (but it can never be perfect).
3276 11. The use of "drop" in a nested ACL was not being handled correctly in the
3277 outer ACL. Now, if condition failure induced by the nested "drop" causes
3278 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3279 or "require"), the connection is dropped.
3281 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3282 that yield "discard" can now be used with an "accept" or a "discard" verb,
3283 but an error is generated for any others (because I can't see a useful way
3284 to define what should happen).
3286 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3287 are now processed in the same way as lines in the Exim configuration file.
3288 In particular, continuation lines are supported.
3290 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3292 15. Added -ti meaning -t -i.
3294 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3295 domains, and warn by logging if others are found.
3297 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3298 returned from GIFCONF when Exim is trying to find the list of interfaces on
3299 a host. The code in os.c has been modified to copy each ifreq to an aligned
3300 structure in all cases.
3302 Also, in some cases, the returned ifreq's were being copied to a 'struct
3303 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3304 means the last couple of bytes of an IPv6 address could be chopped if the
3305 ifreq contained only a normal sockaddr (14 bytes storage).
3307 18. Named domain lists were not supported in the hosts_treat_as_local option.
3308 An entry such as +xxxx was not recognized, and was treated as a literal
3311 19. Ensure that header lines added by a DATA ACL are included in the reject log
3312 if the ACL subsequently rejects the message.
3314 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3315 MD5 (which is deprecated).
3317 21. When testing a filter file using -bf, Exim was writing a message when it
3318 took the sender from a "From " line in the message, but it was not doing so
3319 when it took $return_path from a Return-Path: header line. It now does.
3321 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3322 with a valid header line field name (a series of printing characters
3323 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3325 23. Changed "disc" in the source to "disk" to conform to the documentation and
3326 the book and for uniformity.
3328 24. Ignore Sendmail's -Ooption=value command line item.
3330 25. When execve() failed while trying to run a command in a pipe transport,
3331 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3332 could be confused with a return value of 69 from the command itself. This
3333 has been changed to 127, the value the shell returns if it is asked to run
3334 a non-existent command. The wording for the related log line suggests a
3335 non-existent command as the problem.
3337 26. If received_header_text expands to an empty string, do not add a Received:
3338 header line to the message. (Well, it adds a token one on the spool, but
3339 marks it "old" so that it doesn't get used or transmitted.)
3341 27. Installed eximstats 1.28 (addition of -nt option).
3343 28. There was no check for failure on the call to getsockname() in the daemon
3344 code. This can fail if there is a shortage of resources on the system, with
3345 ENOMEM, for example. A temporary error is now given on failure.
3347 29. Contrary to the C standard, it seems that in some environments, the
3348 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3349 program. Exim now does this explicitly; it affects the formatting of
3350 timestamps using strftime().
3352 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3353 complaints. I've now initialized all the variables, to avoid this.
3355 32. Header lines added by a system filter were not being "seen" during
3356 transport-time rewrites.
3358 33. The info_callback() function passed to OpenSSL is set up with type void
3359 (*)(SSL *, int, int), as described somewhere. However, when calling the
3360 function (actually a macro) that sets it up, the type void(*)() is
3361 expected. I've put in a cast to prevent warnings from picky compilers.
3363 34. If a DNS black list lookup found a CNAME record, but there were no A
3364 records associated with the domain it pointed at, Exim crashed.
3366 35. If a DNS black list lookup returned more than one A record, Exim ignored
3367 all but the first. It now scans all returned addresses if a particular IP
3368 value is being sought. In this situation, the contents of the
3369 $dnslist_value variable are a list of all the addresses, separated by a
3372 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3373 to accept a host name and all its aliases if the forward lookup for any of
3374 them yielded the IP address of the incoming connection. Now it accepts only
3375 those names whose forward lookup yields the correct IP address. Any other
3376 names are discarded. This closes a loophole whereby a rogue DNS
3377 administrator could create reverse DNS records to break through a
3378 wildcarded host restriction in an ACL.
3380 37. If a user filter or a system filter that ran in a subprocess used any of
3381 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3382 the wrong values were passed to the pipe command ($thisaddress had the
3383 value of $0, $0 had the value of $1, etc). This bug was introduced by
3384 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3386 38. Improved the line breaking for long SMTP error messages from ACLs.
3387 Previously, if there was no break point between 40 and 75 characters, Exim
3388 left the rest of the message alone. Two changes have been made: (a) I've
3389 reduced the minimum length to 35 characters; (b) if it can't find a break
3390 point between 35 and 75 characters, it looks ahead and uses the first one
3391 that it finds. This may give the occasional overlong line, but at least the
3392 remaining text gets split now.
3394 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3395 file descriptors might be low, and that setting 1000 would always raise it.
3396 It turns out that in some environments, the limit is already over 1000 and
3397 that lowering it causes trouble. So now Exim takes care not to decrease it.
3399 40. When delivering a message, the value of $return_path is set to $sender_
3400 address at the start of routing (routers may change the value). By an
3401 oversight, this default was not being set up when an address was tested by
3402 -bt or -bv, which affected the outcome if any router or filter referred to
3405 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3406 log only when "message" or "log_message" are set. However, if one of the
3407 conditions was an address verification, or a call to a nested ACL, the
3408 messages generated by the underlying test were being passed through. This
3409 no longer happens. The underlying message is available in $acl_verify_
3410 message for both "message" and "log_message" expansions, so it can be
3411 passed through if needed.
3413 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3414 new expansion $bh_ to give the encoded byte string without charset
3415 translation. Translation happens only if iconv() is available; HAVE_ICONV
3416 indicates this at build time. HEADERS_CHARSET gives the charset to
3417 translate to; headers_charset can change it in the configuration, and
3418 "headers charset" can change it in an individual filter file.
3420 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3421 that creates RFC 2047 encoded "words" labels them as that charset instead
3422 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3423 expansion operator; (ii) when Exim creates a From: line for a local
3424 message; (iii) when a header line is rewritten to include a "phrase" part.
3426 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3427 buggy, causing it to skip the first lines of messages whose message ID
3428 ended in 'D'. This would not have bitten before Exim release 4.14, because
3429 message IDs were unlikely to end in 'D' before then. The effect was to have
3430 incorrect size information for certain domains.
3432 45. #include "config.h" was missing at the start of the crypt16.c module. This
3433 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3436 46. If there was a timeout during a "random" callout check, Exim treated it as
3437 a failure of the random address, and carried on sending RSET and the real
3438 address. If the delay was just some slowness somewhere, the response to the
3439 original RCPT would be taken as a response to RSET and so on, causing
3440 mayhem of various kinds.
3442 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3443 when I implemented it. It didn't allow for the fact that some option values
3444 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3445 the right test for positive values.
3447 48. Domain names in DNS records are case-independent. Exim always looks them up
3448 in lower case. Some resolvers return domain names in exactly the case they
3449 appear in the zone file, that is, they may contain uppercase letters. Not
3450 all resolvers do this - some return always lower case. Exim was treating a
3451 change of case by a resolver as a change of domain, similar to a widening
3452 of a domain abbreviation. This triggered its re-routing code and so it was
3453 trying to route what was effectively the same domain again. This normally
3454 caused routing to fail (because the router wouldn't handle the domain
3455 twice). Now Exim checks for this case specially, and just changes the
3456 casing of the domain that it ultimately uses when it transmits the message
3459 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3462 50. If a filter generated a file delivery with a non-absolute name (possible if
3463 no home directory exists for the router), the forbid_file option was not
3466 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3467 the existing equality matching.
3469 52. Exim was using ints instead of ino_t variables in some places where it was
3470 dealing with inode numbers.
3472 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3473 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3474 called TMPDIR, and if it finds it is different, it changes its value.
3476 54. The smtp_printf() function is now made available to local_scan() so
3477 additional output lines can be written before returning. There is also an
3478 smtp_fflush() function to enable the detection of a dropped connection.
3479 The variables smtp_input and smtp_batched_input are exported to
3482 55. Changed the default runtime configuration: the message "Unknown user"
3483 has been removed from the ACL, and instead placed on the localuser router,
3484 using the cannot_route_message feature. This means that any verification
3485 failures that generate their own messages won't get overridden. Similarly,
3486 the "Unrouteable address" message that was in the ACL for unverifiable
3487 relay addresses has also been removed.
3489 56. Added hosts_avoid_esmtp to the smtp transport.
3491 57. The exicyclog script was not checking for the esoteric option
3492 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3493 will work only if exicyclog is run under the appropriate euid.
3495 58. Following a discussion on the list, the rules by which Exim recognises line
3496 endings on incoming messages have been changed. The -dropcr and drop_cr
3497 options are now no-ops, retained only for backwards compatibility. The
3498 following line terminators are recognized: LF CRLF CR. However, special
3499 processing applies to CR:
3501 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3502 nor a local message in the state where . is a terminator.
3504 (ii) If a bare CR is encountered in a header line, an extra space is added
3505 after the line terminator so as not to end the header. The reasoning
3506 behind this is that bare CRs in header lines are most likely either
3507 to be mistakes, or people trying to play silly games.
3509 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3510 was being incorrectly given as 18 bytes larger than it should have been.
3511 This is a VOB (very old bug).
3513 60. This may never have affected anything current, but just in case it has:
3514 When the local host is found other than at the start of a list of hosts,
3515 the local host, those with the same MX, and any that follow, are discarded.
3516 When the list in question was part of a longer list of hosts, the following
3517 hosts (not currently being processed) were also being discarded. This no
3518 longer happens. I'm not sure if this situation could ever has previously
3521 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3524 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3525 header. This is something that is recommended in a new Internet Draft, and
3526 is something that is documented as being done by Sendmail. There are two
3527 possible values. For messages generated by the autoreply transport, Exim
3530 Auto-Submitted: auto-replied
3532 whereas for all other generated messages (e.g. bounces) it adds
3534 Auto-Submitted: auto-generated
3536 63. The "personal" condition in filters now includes a test for the
3537 Auto-Submitted: header. If it contains the string "auto-" the message it
3538 not considered personal.
3540 64. Added rcpt_include_affixes as a generic transport option.
3542 65. Added queue_only_override (default true).
3544 66. Added the syslog_duplication option.
3546 67. If what should have been the first header line of a message consisted of
3547 a space followed by a colon, Exim was mis-interpreting it as a header line.
3548 It isn't of course - it is syntactically invalid and should therefore be
3549 treated as the start of the message body. The misbehaviour could have
3550 caused a number of strange effects, including loss of data in subsequent
3551 header lines, and spool format errors.
3553 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3554 client host had authenticated. This control can now be exercised by an ACL
3555 for more flexibility.
3557 69. By default, callouts do not happen when testing with -bh. There is now a
3558 variant, -bhc, which does actually run the callout code, including
3559 consulting and updating the callout cache.
3561 70. Added support for saslauthd authentication, courtesy of Alexander
3564 71. If statvfs() failed on the spool or log directories while checking their
3565 size for availability, Exim confusingly gave the error "space shortage".
3566 Furthermore, in debugging mode it crashed with a floating point exception.
3567 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3568 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3569 serious problem, Exim now writes to the main and panic logs when this
3570 happens, with details of the failure. It then refuses to accept the
3571 incoming message, giving the message "spool directory problem" or "log
3572 directory problem" with a 421 code for SMTP messages.
3574 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3575 0, 1, and 2 exist, because some OS complain for execs without them (see
3576 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3577 descriptors. However, the code omitted to check that the open succeeded,
3578 causing mysterious errors if for some reason the permissions on /dev/null
3579 got screwed. Now Exim writes a message to the main and panic logs, and
3580 bombs out if it can't open /dev/null.
3582 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3583 interact so that it is all more flexible. It is supposed to remain
3584 backwards compatible. Also added extra_local_interfaces.
3586 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3587 to bomb out with an assertion failure - to the client this appears as a
3588 connection drop. This problem occurs in the part of the code that was taken
3589 from the Samba project. Fortunately, the assertion is in a very simple
3590 function, so I have fixed this by reproducing the function inline in the
3591 one place where it is called, and arranging for authentication to fail
3592 instead of killing the process with assert().
3594 75. The SPA client code was not working when the server requested OEM rather
3595 than Unicode encoding.
3597 76. Added code to make require_files with a specific uid setting more usable in
3598 the case where statting the file as root fails - usually a non-root-mounted
3599 NFS file system. When this happens and the failure is EACCES, Exim now
3600 forks a subprocess and does the per-uid checking as the relevant uid.
3602 77. Added process_log_path.
3604 78. If log_file_path was not explicitly set, a setting of check_log_space or
3605 check_log_inodes was ignored.
3607 79. If a space check for the spool or log partitions fails, the incident is now
3608 logged. Of course, in the latter case the data may get lost...
3610 80. Added the %p formatting code to string_format() so that it can be used to
3611 print addresses in debug_print(). Adjusted all the address printing in the
3612 debugging in store.c to use %p rather than %d.
3614 81. There was a concern that a line of code in smtp_in.c could overflow a
3615 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3616 initially expressed, the concern was not well-founded, because trailing
3617 spaces are removed early. However, if the trailing spaces were followed by
3618 a NULL, they did not get removed, so the overflow was possible. Two fixes
3621 (a) I re-wrote the offending code in a cleaner fashion.
3622 (b) If an incoming SMTP command contains a NULL character, it is rejected
3625 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3626 runs initgroups(), so that if the Exim user is in any additional groups,
3627 they will be used during message reception.
3633 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob