1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.276 2005/12/15 11:18:21 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.
50 PH/01 Two changes to the default runtime configuration:
52 (1) Move the checks for relay_from_hosts and authenticated clients from
53 after to before the (commented out) DNS black list checks.
55 (2) Add control=submission to the relay_from_hosts and authenticated
56 clients checks, on the grounds that messages accepted by these
57 statements are most likely to be submissions.
59 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
61 (1) Generate an error if the third argument for the ${prvs expansion is
64 (2) Treat a missing third argument of ${prvscheck as if it were an empty
67 (3) Reset the variables that are obtained from the first argument of
68 ${prvscheck and used in the second argument before leaving the code,
69 because their memory is reclaimed, so using them afterwards may do
72 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
73 one (it's much easier than Tom thought :-).
75 (5) Because of (4), we can now allow for the use of $prvscheck_result
76 inside the third argument.
78 PH/03 For some reason, the default setting of PATH when running a command from
79 a pipe transport was just "/usr/bin". I have changed it to
82 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
83 anything to be listed in the output from -bV.
85 PH/05 When a filter generated an autoreply, the entire To: header line was
86 quoted in the delivery log line, like this:
88 => >A.N.Other <ano@some.domain> <original@ddress> ...
90 This has been changed so that it extracts the operative address. There
91 may be more than one such address. If so, they are comma-separated, like
94 => >ano@some.domain,ona@other.domain <original@ddress> ...
96 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
97 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
98 not being looked up in the rDNS to get a host name, Exim was showing the
99 IP address twice in Received: lines, even though the IP addresses were
100 identical. For example:
102 Received: from [1.2.3.4] (helo=[1.2.3.4])
104 However, if the real host name was known, it was omitting the HELO data
105 if it matched the actual IP address. This has been tidied up so that it
106 doesn't show the same IP address twice.
108 PH/07 When both +timestamp and +memory debugging was on, the value given by
109 $tod_xxx expansions could be wrong, because the tod_stamp() function was
110 called by the debug printing, thereby overwriting the timestamp buffer.
111 Debugging no longer uses the tod_stamp() function when +timestamp is set.
113 PH/08 When the original message was included in an autoreply transport, it
114 always said "this is a copy of the message, including all the headers",
115 even if body_only or headers_only was set. It now gives an appropriate
118 PH/09 Applied a patch from the Sieve maintainer which:
120 o fixes some comments
121 o adds the (disabled) notify extension core
122 o adds some debug output for the result of if/elsif tests
123 o points to the current vacation draft in the documentation
124 and documents the missing references header update
128 o fixes a bug in processing the envelope test (when testing
129 multiple envelope elements, the last element determinted the
132 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
133 Electronic Mail") by including:
135 Auto-submitted: auto-generated
137 in the messages that it generates (bounce messages and others, such as
138 warnings). In the case of bounce messages for non-SMTP mesages, there was
139 also a typo: it was using "Auto_submitted" (underscore instead of
140 hyphen). Since every message generated by Exim is necessarily in response
141 to another message, thes have all been changed to:
143 Auto-Submitted: auto-replied
145 in accordance with these statements in the RFC:
147 The auto-replied keyword:
149 - SHOULD be used on messages sent in direct response to another
150 message by an automatic process,
152 - MUST NOT be used on manually-generated messages,
154 - MAY be used on Delivery Status Notifications (DSNs) and Message
155 Disposition Notifications (MDNs),
157 - MUST NOT be used on messages generated by automatic or periodic
158 processes, except for messages which are automatic responses to
161 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
162 to the default Received: header definition.
164 PH/12 Added log selector acl_warn_skipped (default on).
166 PH/13 After a successful wildlsearch lookup, discard the values of numeric
167 variables because (a) they are in the wrong storage pool and (b) even if
168 they were copied, it wouldn't work properly because of the caching.
170 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
171 checking when decoding. Apparently there are clients that generate
172 overlong encoded strings. Why am I not surprised?
174 PH/15 If the first argument of "${if match_address" was not empty, but did not
175 contain an "@" character, Exim crashed. Now it writes a panic log message
176 and treats the condition as false.
178 PH/16 In autoreply, treat an empty string for "once" the same as unset.
180 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
181 extension "envelope-auth". The code is finished and in agreement with
182 other implementations, but there is no documentation so far and in fact,
183 nobody wrote the draft yet. This extension is currently #undef'ed, thus
184 not changing the active code.
186 Print executed "if" and "elsif" statements when debugging is used. This
187 helps a great deal to understand what a filter does.
189 Document more things not specified clearly in RFC3028. I had all this
190 sorted out, when out of a sudden new issues came to my mind. Oops."
192 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
195 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
196 canonical form (as documented). However, after a host name lookup from
197 the IP address, check_host() was doing a simple string comparison with
198 addresses acquired from the DNS when checking that the found name did
199 have the original IP as one of its addresses. Since any found IPv6
200 addresses are likely to be in abbreviated form, the comparison could
201 fail. Luckily, there already exists a function for doing the comparison
202 by converting both addresses to binary, so now that is used instead of
205 PH/20 There was another similar case to PH/19, when a complete host name was
206 given in a host list; looking up its IP address could give an abbreviated
207 form, whereas the current host's name might or might not be abbreviated.
208 The same fix has been applied.
214 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
215 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
218 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
219 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
221 PH/03 Typo: missing ".o" in src/pcre/Makefile.
223 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
224 header line, restrict the check to what is listed in RFCs 2369 and 2929.
225 Also, for "Auto-Submitted", treat anything other than "no" as
226 non-personal, in accordance with RFC 3834. (Previously it treated
227 anything starting "auto-" as non-personal.)
229 TF/01 The control=submission/name=... option had a problem with syntax
230 errors if the name included a slash character. The /name= option
231 now slurps the rest of the string, so it can include any characters
232 but it must come last in the list of options (after /sender_retain
235 PH/05 Some modifications to the interface to the fake nameserver for the new
238 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
239 into the default Received: header string.
246 TK/01 Added the "success_on_redirect" address verification option. See
247 NewStuff for rationale and an example.
249 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
251 PH/02 Patch to exigrep to allow it to work on syslog lines.
253 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
254 fread() to skip over the body file's header line, because in Cygwin the
255 header line is locked and is inaccessible.
257 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
258 co-exist for some time) to make it clear that it is the Exim ID that is
259 referenced, not the Message-ID: header line.
261 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
262 string_format() function, because snprintf() does not exist on all
265 PH/06 The use of forbid_filter_existstest now also locks out the use of the
266 ${stat: expansion item.
268 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
269 protocol synchronization error", to keep the pedants happy.
271 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
272 well as for IRIX systems, when gcc is being used. See the host.c source
275 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
277 PH/10 Named domain lists were not working if used in a queue_smtp_domains
280 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
281 transport and to the smtp transport in LMTP mode.
283 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
285 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
286 run a filter in a subprocess. This could lead to confusion in subsequent
287 lookups in the parent process. There should also be a search_tidyup() at
288 the end of the subprocess.
290 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
291 only if the host matched helo_try_verify_hosts, which caused the
292 verification to occur when the EHLO/HELO command was issued. The ACL just
293 tested the remembered result. Now, if a previous verification attempt has
294 not happened, "verify = helo" does it there and then.
296 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
298 TK/03 Fix log output including CR from clamd.
300 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
301 exist provoked a memory error which could cause a segfault.
303 PH/15 Installed PCRE 6.2
305 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
307 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
308 of the problem. Specifically, suggested +O2 rather than +O1 for the
311 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
313 PH/20 If a delivery was routed to a non-standard port by means of an SRV
314 record, the port was not correctly logged when the outgoing_port log
315 selector was set (it logged the transort's default port).
317 PH/21 Added support for host-specific ports to manualroute, queryprogram,
318 fallback_hosts, and "hosts" in the smtp transport.
320 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
321 host errors such as "Connection refused".
323 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
324 authentication with radiusclient 0.4.9:
326 - Error returned from rc_read_config was caught wrongly
327 - Username/password not passed on to radius server due to wrong length.
329 The presumption is that some radiusclient API changes for 4.51/PH/17
330 were not taken care of correctly. The code is still untested by me (my
331 Linux distribution still has 0.3.2 of radiusclient), but it was
332 contributed by a Radius user.
334 PH/24 When doing a callout, the value of $domain wasn't set correctly when
335 expanding the "port" option of the smtp transport.
337 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
338 while reading a MIME header. Thanks to Tom Hughes for a patch.
340 PH/24 Include config.h inside local_scan.h so that configuration settings are
343 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
344 that in an ACL for RCPT (for example), you can examine exactly what was
347 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
348 commands, but it was not correctly comparing the address with the actual
349 client host address. Thus, it would show the EHLO address in Received:
350 header lines when this was not necessary.
352 PH/27 Added the % operator to ${eval:}.
354 PH/28 Exim tries to create and chdir to its spool directory when it starts;
355 it should be ignoring failures (because with -C, for example, it has lost
356 privilege). It wasn't ignoring creation failures other than "already
359 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
362 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
363 because an input line was too long, either on its own, or by virtue of
364 too many continuations, the temporary file was not being removed, and the
365 return code was incorrect.
367 PH/31 Missing "BOOL" in function definition in filtertest.c.
369 PH/32 Applied Sieve patches from the maintainer.
371 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
373 PH/33 Added "verify = not_blind".
375 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
376 Local/Makefile (with some defaults set). These are used in built scripts
377 such as exicyclog, but they have never been used in the exim_install
378 script (though there are many overriding facilities there). I have
379 arranged that the exim_install script now takes note of these two
382 PH/35 Installed configuration files for Dragonfly.
384 PH/36 When a locally submitted message by a trusted user did not contain a
385 From: header, and the sender address was obtained from -f or from an SMTP
386 MAIL command, and the trusted user did not use -F to supply a sender
387 name, $originator_name was incorrectly used when constructing a From:
388 header. Furthermore, $originator_name was used for submission mode
389 messages from external hosts without From: headers in a similar way,
390 which is clearly wrong.
392 PH/37 Added control=suppress_local_fixups.
394 PH/38 When log_selector = +received_sender was set, and the addition of the
395 sender made the log line's construction buffer exactly full, or one byte
396 less than full, an overflow happened when the terminating "\n" was
399 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
400 when the result of a list match is failure because a DNS lookup failed.
402 PH/40 RM_COMMAND is now used in the building process.
404 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
405 the "build-* directories that it finds.
407 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
408 address in a domain literal was a prefix of an interface address.
410 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
411 when verifying a sender address, unless rewrite_headers is false.
413 PH/44 Wrote a long comment about why errors_to addresses are verified as
414 recipients, not senders.
416 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
417 the ratelimit ACL was added.
419 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
421 PH/46 Added extra information about PostgreSQL errors to the error string.
423 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
424 avoiding the need to install special zones in a real server. This is
425 backwards compatible; if it can't find the fake resolver, it drops back.
426 Thus, both old and new test suites can be run.
428 TF/02 Added util/ratelimit.pl
430 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
431 clock is set back in time.
433 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
434 Candler <B.Candler@pobox.com>.
436 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
437 OK for addresses that are the result of redirections.
439 PH/48 A number of further additions for the benefit of the new test suite,
440 including a fake gethostbyname() that interfaces to the fake DNS resolver
443 TF/06 The fix for widen_domains has also been applied to qualify_single and
444 search_parents which are the other dnslookup options that can cause
447 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
450 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
452 TF/07 Exim produced the error message "an SRV record indicated no SMTP
453 service" if it encountered an MX record with an empty target hostname.
454 The message is now "an MX or SRV record indicated no SMTP service".
456 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
457 if the DNS of the sending site is misconfigured. This is quite a
458 common situation. This change restores the behaviour of treating a
459 helo verification defer as a failure.
461 PH/51 If self=fail was set on a router, the bounce message did not include the
462 actual error message.
468 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
470 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
471 command itself ended in error, the underlying message about the transport
472 filter timeout was being overwritten with the pipe command error. Now the
473 underlying error message should be appended to the second error message.
475 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
477 PH/02 Reduce the amount of output that "make" produces by default. Full output
478 can still be requested.
480 PH/03 The warning log line about a condition test deferring for a "warn" verb
481 was being output only once per connection, rather than after each
482 occurrence (because it was using the same function as for successful
483 "warn" verbs). This seems wrong, so I have changed it.
485 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
486 it should not have, which might have caused a crash in the right
487 circumstances, but probably never did.
489 PH/04 Installed a modified version of Tony Finch's patch to make submission
490 mode fix the return path as well as the Sender: header line, and to
491 add a /name= option so that you can make the user's friendly name appear
494 TF/03 Added the control = fakedefer ACL modifier.
496 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
497 Mark Lowes for thorough testing.
499 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
501 TK/03 Merged latest SRS patch from Miles Wilton.
503 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
504 with the definition in sysexits.h (which is #included earlier).
505 Fortunately, Exim does not actually use EX_OK. The code used to try to
506 preserve the sysexits.h value, by assumimg that macro definitions were
507 scanned for macro replacements. I have been disabused of this notion,
508 so now the code just undefines EX_OK before #including unistd.h.
510 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
511 in the smtp transport. When a block could not be written in a single
512 write() function, the timeout was being re-applied to each part-write.
513 This seems wrong - if the receiver was accepting one byte at a time it
514 would take for ever. The timeout is now adjusted when this happens. It
515 doesn't have to be particularly precise.
517 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
518 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
520 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
521 without a domain if the check to <postmaster@domain> fails.
523 SC/01 Eximstats: added -xls and the ability to specify output files
524 (patch written by Frank Heydlauf).
526 SC/02 Eximstats: use FileHandles for outputing results.
528 SC/03 Eximstats: allow any combination of xls, txt, and html output.
530 SC/04 Eximstats: fixed display of large numbers with -nvr option
532 SC/05 Eximstats: fixed merging of reports with empty tables.
534 SC/06 Eximstats: added the -include_original_destination flag
536 SC/07 Eximstats: removed tabs and trailing whitespace.
538 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
540 TK/06 MBOX spool code: Add real "From " MBOX separator line
541 so the .eml file is really in mbox format (even though
542 most programs do not really care). Patch from Alex Miller.
544 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
545 The latter is generated from $received_to and is only set if the
546 message has one envelope recipient. SA can use these headers,
547 obviously out-of-the-box. Patch from Alex Miller.
549 PH/08 The ${def test on a variable was returning false if the variable's
550 value was "0", contrary to what the specification has always said!
551 The result should be true unless the variable is empty.
553 PH/09 The syntax error of a character other than { following "${if
554 def:variable_name" (after optional whitespace) was not being diagnosed.
555 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
556 accidental colon was present, for example, could give incorrect results.
558 PH/10 Tidied the code in a number of places where the st_size field of a stat()
559 result is used (not including appendfile, where other changes are about
562 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
563 This involved changing a lot of size variables from int to off_t. It
564 should work with maildirs and everything.
566 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
567 spamd dying while we are connected to it.
569 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
572 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
573 function for BDB 4.3.
575 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
577 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
578 the sources for PCRE have been re-arranged and more files are now
581 PH/15 The code I had for printing potentially long long variables in PH/11
582 above was not the best (it lost precision). The length of off_t variables
583 is now inspected at build time, and an appropriate printing format (%ld
584 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
585 to be "long long int" or "long int". This is needed for the internal
586 formatting function string_vformat().
588 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
589 the configuration file to be ":syslog", then the script "guesses" where
590 the logs files are, rather than using the compiled in default. In our
591 case the guess is not the same as the compiled default, so the script
592 suddenly stopped working when I started to use syslog. The patch checks
593 to see if log_file_path is "". If so, it attempts to read it from exim
594 with no configuration file to get the compiled in version, before it
595 falls back to the previous guessing code."
597 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
598 implementing BATV in an Exim configuration. See NewStuff for the gory
601 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
602 Makefile that are specific to HP-UX.
604 PH/18 If the "use_postmaster" option was set for a recipient callout together
605 with the "random" option, the postmaster address was used as the MAIL
606 FROM address for the random test, but not for the subsequent recipient
607 test. It is now used for both.
609 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
610 patch removes a few documentation additions to RFC 3028, because the
611 latest draft now contains them. It adds the new en;ascii-case comparator
612 and a new error check for 8bit text in MIME parts. Comparator and
613 require names are now matched exactly. I enabled the subaddress
614 extension, but it is not well tested yet (read: it works for me)."
616 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
617 rework some of the code of TK/09 above to avoid the hardwired use of
618 "%lld" and "long long". Replaced the call to snprintf() with a call to
621 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
622 records point to non-existent hosts", "retry timeout exceeded", and
623 "retry time not reached for any host after a long failure period".
625 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
626 experimental DomainKeys support:
628 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
629 (2) On an error such as an illegally used "control", the wrong name for
630 the control was given.
632 These problems did NOT occur unless DomainKeys support was compiled.
634 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
636 PH/24 Added ${if match_ip condition.
638 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
639 databases so that it will be absolutely obvious if a crash occurs in the
640 DB library. This is a regular occurrence (often caused by mis-matched
643 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
644 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
645 picked up on a user's system that detects such things. There doesn't seem
646 to be a gcc warning option for this - only an attribute that has to be
647 put on the function's prototype. It seems that in Fedora Core 4 they have
648 set this on a number of new functions. No doubt there will be more in due
651 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
652 specify a transport. However, if an address that was verified by such a
653 router was the subject of a callout, Exim crashed because it tried to
654 read the rcpt_include_affixes from the non-existent transport. Now it
655 just assumes that the setting of that option is false. This bug was
656 introduced by 4.51/PH/31.
658 PH/28 Changed -d+all to exclude +memory, because that information is very
659 rarely of interest, but it makes the output a lot bigger. People tend to
660 do -d+all out of habit.
662 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
663 code in os-type was giving problems when libc.so lives in lib64, like on
666 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
667 aren't the modern standard, and it seems that some systems' include files
668 don't always have them. Exim was already checking for some of the newer
669 ones like T_AAAA, and defining it itself. I've added checks for all the
670 record types that Exim uses.
672 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
673 not automatically generating a new one, as it is supposed to. This
674 prevented TLS from working. If the file did exist, but contained invalid
675 data, a new version was generated, as expected. It was only the case of a
676 non-existent file that was broken.
678 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
679 with a change in libdomainkeys > 0.64.
681 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
682 from DNS. If the selector record carries the flag, it now has
683 precedence over the domain-wide flag.
685 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
687 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
688 the use of a port name that isn't defined in /etc/services) occurred, the
689 message was deferred as in a normal delivery, and thus remained on the
690 spool, instead of being failed because of the mua_wrapper setting. This
691 is now fixed, and I tidied up some of the mua_wrapper messages at the
694 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
695 the messages in a hash of arrays rather than using individual hashes.
696 This is a bit cleaner and results in dramatic memory savings, albeit
697 at a slight CPU cost.
699 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
700 as requested by Marc Sherman.
702 SC/10 Eximstats: added histograms for user specified patterns as requested
705 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
707 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
708 fopen() in the content-scanning modules that did not already have it.
714 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
715 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
717 TK/02 Fix ACL "control" statement not being available in MIME ACL.
719 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
721 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
722 to test Sieve filters that use "vacation".
724 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
725 that changes the way the GnuTLS parameters are stored in the cache file.
726 The new format can be generated externally. For backward compatibility,
727 if the data in the cache doesn't make sense, Exim assumes it has read an
728 old-format file, and it generates new data and writes a new file. This
729 means that you can't go back to an older release without removing the
732 PH/03 A redirect router that has both "unseen" and "one_time" set does not
733 work if there are any delivery delays because "one_time" forces the
734 parent to be marked "delivered", so its unseen clone is never tried
735 again. For this reason, Exim now forbids the simultaneous setting of
738 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
739 redirected to themselves ("homonym" addresses). Read the long ChangeLog
740 entry if you want to know the details. The fix, however, neglected to
741 consider the case when local delivery batching is involved. The test for
742 "previously delivered" was not happening when checking to see if an
743 address could be batched with a previous (undelivered) one; under
744 certain circumstances this could lead to multiple deliveries to the same
747 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
748 in its include files, and this causes problems building Exim.
750 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
751 header_syntax) but Exim was just ignoring anything given after a slash.
752 In particular, this caused confusion with an attempt to use "verify =
753 reverse_host_lookup/defer_ok". An error is now given when options are
754 supplied for verify items that do not have them. (Maybe reverse_host_
755 lookup should have a defer_ok option, but that's a different point.)
757 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
758 defined by RFC 821) to 2048, because there were problems with some AUTH
759 commands, and RFC 1869 says the size should be increased for extended
760 SMTP commands that take arguments.
762 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
765 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
766 "unknown" error; now it says that the functionality isn't in the binary.
768 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
769 an address' error message when a string expansion fails (syntax or
770 whatever). Otherwise the password may appear in the log. Following change
771 PH/42 below, there is no longer a chance of it appearing in a bounce
774 PH/11 Installed exipick version 20050225.0 from John Jetmore.
776 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
777 of its addresses was ever tried. (Bugzilla bug #2.)
779 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
780 the result incorrectly in the debug output. (It correctly added a newline
781 to what was transported.)
783 TF/01 Added $received_time.
785 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
786 commented out examples of how to interface to a virus scanner and to
787 SpamAssassin. Also added commented examples of av_scanner and
788 spamd_address settings.
790 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
791 and controls are allowed in which ACLs. There were a couple of minor
792 errors. Some of the entries in the conditions table (which is a table of
793 where they are NOT allowed) were getting very unwieldy; rewrote them as a
794 negation of where the condition IS allowed.
796 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
798 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
799 header file does not have a version number, so I've had to invent a new
800 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
801 API. The code is untested by me (my Linux distribution still has 0.3.2 of
802 radiusclient), but it was contributed by a Radius user.
804 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
805 files or directories, for OpenSSL.
807 PH/19 When an Exim process that is running as root has to create an Exim log
808 file, it does so in a subprocess that runs as exim:exim so as to get the
809 ownership right at creation (otherwise, other Exim processes might see
810 the file with the wrong ownership). There was no test for failure of this
811 fork() call, which would lead to the process getting stuck as it waited
812 for a non-existent subprocess. Forks do occasionally fail when resources
813 run out. I reviewed all the other calls to fork(); they all seem to check
816 PH/20 When checking for unexpected SMTP input at connect time (before writing
817 the banner), Exim was not dealing correctly with a non-positive return
818 from the read() function. If the client had disconnected by this time,
819 the result was a log entry for a synchronization error with an empty
820 string after "input=" when read() returned zero. If read() returned -1
821 (an event I could not check), uninitialized data bytes were printed.
822 There were reports of junk text (parts of files, etc) appearing after
825 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
827 PH/22 Added support for macro redefinition, and (re)definition in between
828 driver and ACL definitions.
830 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
831 forgetting to use the resulting value; it was using the unexpanded value.
833 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
834 hadn't been configured. The fix is from Juergen Kreileder, who
835 understands it better than I do:
837 "Here's what I see happening with three configured cyrus_sasl
838 authenticators configured (plain, login, cram-md5):
840 On startup auth_cyrus_sasl_init() gets called for each of these.
841 This means three calls to sasl_listmech() without a specified mech_list.
842 => SASL tests which mechs of all available mechs actually work
843 => three warnings about OTP not working
844 => the returned list contains: plain, login, cram-md5, digest-md5, ...
846 With the patch, sasl_listmech() also gets called three times. But now
847 SASL's mech_list option is set to the server_mech specified in the the
848 authenticator. Or in other words, the answer from sasl_listmech()
849 gets limited to just the mech you're testing for (which is different
851 => the return list contains just 'plain' or 'login', 'cram-md5' or
852 nothing depending on the value of ob->server_mech.
854 I've just tested the patch: Authentication still works fine,
855 unavailable mechs specified in the exim configuration are still
856 caught, and the auth.log warnings about OTP are gone."
858 PH/25 When debugging is enabled, the contents of the command line are added
859 to the debugging output, even when log_selector=+arguments is not
862 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
863 answer is "GNU", and only if the return is "GNU/something" is the answer
866 PH/27 $acl_verify_message is now set immediately after the failure of a
867 verification in an ACL, and so is available in subsequent modifiers. In
868 particular, the message can be preserved by coding like this:
870 warn !verify = sender
871 set acl_m0 = $acl_verify_message
873 Previously, $acl_verify_message was set only while expanding "message"
874 and "log_message" when a very denied access.
876 PH/28 Modified OS/os.c-Linux with
878 -#ifndef OS_LOAD_AVERAGE
879 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
881 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
882 nomenclature these days.)
884 PH/29 Installed patch from the Sieve maintainer that adds the options
885 sieve_useraddress and sieve_subaddress to the redirect router.
887 PH/30 In these circumstances:
888 . Two addresses routed to the same list of hosts;
889 . First host does not offer TLS;
890 . First host accepts first address;
891 . First host gives temporary error to second address;
892 . Second host offers TLS and a TLS session is established;
893 . Second host accepts second address.
894 Exim incorrectly logged both deliveries with the TLS parameters (cipher
895 and peerdn, if requested) that were in fact used only for the second
898 PH/31 When doing a callout as part of verifying an address, Exim was not paying
899 attention to any local part prefix or suffix that was matched by the
900 router that accepted the address. It now behaves in the same way as it
901 does for delivery: the affixes are removed from the local part unless
902 rcpt_include_affixes is set on the transport.
904 PH/32 Add the sender address, as F=<...>, to the log line when logging a
905 timeout during the DATA phase of an incoming message.
907 PH/33 Sieve envelope tests were broken for match types other than :is. I have
908 applied a patch sanctioned by the Sieve maintainer.
910 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
911 the uid or gid is negative. A case of a negative gid caused this to be
912 noticed. The fix allows for either to be negative.
914 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
915 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
918 PH/36 The change PH/12 above was broken. Fixed it.
920 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
921 the grounds that routing the same address twice would always produce the
922 same answer. This might have been true once, but it is certainly no
923 longer true now. Routing a child address may depend on the previous
924 routing that produced that child. Some complicated redirection strategies
925 went wrong when messages had multiple recipients, and made Exim's
926 behaviour dependent on the order in which the addresses were given.
928 I have moved the duplicate checking until after the routing is complete.
929 Exim scans the addresses that are assigned to local and remote
930 transports, and removes any duplicates. This means that more work will be
931 done, as duplicates will always all be routed, but duplicates are
932 presumably rare, so I don't expect this is of any significance.
934 For deliveries to pipes, files, and autoreplies, the duplicate checking
935 still happens during the routing process, since they are not going to be
938 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
939 It corrects a timeout issue with spamd. This is Ian's comment: "The
940 background is that sometimes spamd either never reads data from a
941 connection it has accepted, or it never writes response data. The exiscan
942 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
943 blindly assumes that writes won't block so it may never time out."
945 PH/39 Allow G after quota size as well as K and M.
947 PH/40 The value set for $authenticated_id in an authenticator may not contain
948 binary zeroes or newlines because the value is written to log lines and
949 to spool files. There was no check on this. Now the value is run through
950 the string_printing() function so that such characters are converted to
951 printable escape sequences.
953 PH/41 $message_linecount is a new variable that contains the total number of
954 lines in the message. Compare $body_linecount, which is the count for the
957 PH/42 Exim no longer gives details of delivery errors for specific addresses in
958 bounce and delay warning messages, except in certain special cases, which
961 (a) An SMTP error message from a remote host;
962 (b) A message specified in a :fail: redirection;
963 (c) A message specified in a "fail" command in a system filter;
964 (d) A message specified in a FAIL return from the queryprogram router;
965 (e) A message specified by the cannot_route_message router option.
967 In these cases only, Exim does include the error details in bounce and
968 warning messages. There are also a few cases where bland messages such
969 as "unrouteable address" or "local delivery error" are given.
971 PH/43 $value is now also set for the "else" part of a ${run expansion.
973 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
974 being worked on, but at least Exim now implements the latest version to
977 PH/45 In a pipe transport, although a timeout while waiting for the pipe
978 process to complete was treated as a delivery failure, a timeout while
979 writing the message to the pipe was logged, but erroneously treated as a
980 successful delivery. Such timeouts include transport filter timeouts. For
981 consistency with the overall process timeout, these timeouts are now
982 treated as errors, giving rise to delivery failures by default. However,
983 there is now a new Boolean option for the pipe transport called
984 timeout_defer, which, if set TRUE, converts the failures into defers for
985 both kinds of timeout. A transport filter timeout is now identified in
988 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
989 systems where "make" and "gmake" are different, calling "gmake" at top
990 level broke things. I've arranged for the value of $(MAKE) to be passed
991 from the Makefile to this script so that it can call the same version of
995 A note about Exim versions 4.44 and 4.50
996 ----------------------------------------
998 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
999 changes of various kinds. As a consequence, a big documentation update was
1000 needed. This delayed the release for rather longer than seemed good, especially
1001 in the light of a couple of (minor) security issues. Therefore, the changes
1002 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1003 release. So 4.44 and 4.50 are in effect two different branches that both start
1006 I have left the 4.50 change log unchanged; it contains all the changes since
1007 4.43. The change log for 4.44 is below; many of its items are identical to
1008 those for 4.50. This seems to be the most sensible way to preserve the
1009 historical information.
1015 1. Minor wording change to the doc/README.SIEVE file.
1017 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1018 computation of the current number of files was incorrect.
1020 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1021 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1024 4. Give more explanation in the error message when the command for a transport
1025 filter fails to execute.
1027 5. There are several places where Exim runs a non-Exim command in a
1028 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1029 was being done only for the command run by the queryprogram router. It is
1030 now done for all such subprocesses. The other cases are: ${run, transport
1031 filters, and the commands run by the lmtp and pipe transports.
1033 6. Added CONFIGURE_GROUP build-time option.
1035 7. Some older OS have a limit of 256 on the maximum number of file
1036 descriptors. Exim was using setrlimit() to set 1000 as a large value
1037 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1038 systems. I've change it so that if it can't get 1000, it tries for 256.
1040 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1041 was an oversight, and furthermore, ever since the addition of extra
1042 controls (e.g. 4.43/32), the checks on when to allow different forms of
1043 "control" were broken. There should now be diagnostics for all cases when a
1044 control that does not make sense is encountered.
1046 9. Added the /retain_sender option to "control=submission".
1048 10. $recipients is now available in the predata ACL (oversight).
1050 11. Tidy the search cache before the fork to do a delivery from a message
1051 received from the command line. Otherwise the child will trigger a lookup
1052 failure and thereby defer the delivery if it tries to use (for example) a
1053 cached ldap connection that the parent has called unbind on.
1055 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1056 of $address_data from the recipient verification was clobbered by the
1057 sender verification.
1059 13. The value of address_data from a sender verification is now available in
1060 $sender_address_data in subsequent conditions in the ACL statement.
1062 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1064 15. Added a new option "connect=<time>" to callout options, to set a different
1067 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1068 was its contents. (It was OK if the option was not defined at all.)
1070 17. A "Completed" log line is now written for messages that are removed from
1071 the spool by the -Mrm option.
1073 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1074 information about exactly what failed.
1076 19. Added -dd to debug only the daemon process.
1078 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1079 handles timeouts, both on the server side and network timeouts. Renamed the
1080 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1082 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1083 It is now set to "smtps".
1085 22. $host_address is now set to the target address during the checking of
1086 ignore_target_hosts.
1088 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1089 being passed; this would have caused $sender_host_name to have been used if
1090 matching the list had actually called for a host name (not very likely,
1091 since this list is usually IP addresses). A host name is now passed as
1094 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1095 code that specifies a non-restarting handler (typically sigaction() in
1096 modern systems) in an attempt to fix a rare and obscure crash bug.
1098 25. Narrowed the window for a race in the daemon that could cause it to ignore
1099 SIGCHLD signals. This is not a major problem, because they are used only to
1100 wake it up if nothing else does.
1102 26. A malformed maildirsize file could cause Exim to calculate negative values
1103 for the mailbox size or file count. Odd effects could occur as a result.
1104 The maildirsize information is now recalculated if the size or filecount
1107 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1108 support for a long time. Removed HAVE_SYS_VFS_H.
1110 28. Installed the latest version of exipick from John Jetmore.
1112 29. In an address list, if the pattern was not a regular expression, an empty
1113 subject address (from a bounce message) matched only if the pattern was an
1114 empty string. Non-empty patterns were not even tested. This was the wrong
1115 because it is perfectly reasonable to use an empty address as part of a
1116 database query. An empty address is now tested by patterns that are
1117 lookups. However, all the other forms of pattern expect the subject to
1118 contain a local part and a domain, and therefore, for them, an empty
1119 address still always fails if the pattern is not itself empty.
1121 30. Exim went into a mad DNS loop when attempting to do a callout where the
1122 host was specified on an smtp transport, and looking it up yielded more
1123 than one IP address.
1125 31. Re-factored the code for checking spool and log partition space into a
1126 function that finds that data and another that does the check. The former
1127 is then used to implement four new variables: $spool_space, $log_space,
1128 $spool_inodes, and $log_inodes.
1130 32. The RFC2047 encoding function was originally intended for short strings
1131 such as real names; it was not keeping to the 75-character limit for
1132 encoded words that the RFC imposes. It now respects the limit, and
1133 generates multiple encoded words if necessary. To be on the safe side, I
1134 have increased the buffer size for the ${rfc2047: expansion operator from
1137 33. It is now permitted to omit both strings after an "if" condition; if the
1138 condition is true, the result is "true". As before, when the second string
1139 is omitted, a false condition yields an empty string. This makes it less
1140 cumbersome to write custom ACL and router conditions.
1142 34. Failure to deliver a bounce message always caused it to be frozen, even if
1143 there was an errors_to setting on the router. The errors_to setting is now
1146 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1147 canonical form (fully expanded) before being placed in
1148 $sender_host_address.
1150 36. The table in the code that translates DNS record types into text (T_A to
1151 "A" for instance) was missing entries for NS and CNAME. It is just possible
1152 that this could have caused confusion if both these types were looked up
1153 for the same domain, because the text type is used as part of Exim's
1154 per-process caching. But the chance of anyone hitting this buglet seems
1157 37. The dnsdb lookup has been extended in a number of ways.
1159 (1) There is a new type, "zns", which walks up the domain tree until it
1160 finds some nameserver records. It should be used with care.
1162 (2) There is a new type, "mxh", which is like "mx" except that it returns
1163 just the host names, not the priorities.
1165 (3) It is now possible to give a list of domains (or IP addresses) to be
1166 looked up. The behaviour when one of the lookups defers can be
1167 controlled by a keyword.
1169 (4) It is now possible to specify the separator character for use when
1170 multiple records are returned.
1172 38. The dnslists ACL condition has been extended: it is now possible to supply
1173 a list of IP addresses and/or domains to be looked up in a particular DNS
1176 39. Added log_selector=+queue_time_overall.
1178 40. When running the queue in the test harness, wait just a tad after forking a
1179 delivery process, to get repeatability of debugging output.
1181 41. Include certificate and key file names in error message when GnuTLS fails
1182 to set them up, because the GnuTLS error message doesn't include the name
1183 of the failing file when there is a problem reading it.
1185 42. Allow both -bf and -bF in the same test run.
1187 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1189 44. The "Exiscan patch" is now merged into the mainline Exim source.
1191 45. Sometimes the final signoff response after QUIT could fail to get
1192 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1193 < 0 before doing a fflush(). This bug looks as though it goes back to the
1194 introduction of TLS in release 3.20, but "sometimes" must have been rare
1195 because the tests only now provoked it.
1197 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1198 (this can affect the format of dates).
1200 47. exim_tidydb, when checking for the continued existence of a message for
1201 which it has found a message-specific retry record, was not finding
1202 messages that were in split spool directories. Consequently, it was
1203 deleting retry records that should have stayed in existence.
1205 48. Steve fixed some bugs in eximstats.
1207 49. The SPA authentication driver was not abandoning authentication and moving
1208 on to the next authenticator when an expansion was forced to fail,
1209 contradicting the general specification for all authenticators. Instead it
1210 was generating a temporary error. It now behaves as specified.
1212 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1213 (the order specifies the preference for clients). The order is now AES256,
1214 AES128, 3DES, ARCFOUR128.
1216 51. Small patch to Sieve code - explicitly set From: when generating an
1219 52. Exim crashed if a remote delivery caused a very long error message to be
1220 recorded - for instance if somebody sent an entire SpamAssassin report back
1221 as a large number of 550 error lines. This bug was coincidentally fixed by
1222 increasing the size of one of Exim's internal buffers (big_buffer) that
1223 happened as part of the Exiscan merge. However, to be on the safe side, I
1224 have made the code more robust (and fixed the comments that describe what
1227 53. Now that there can be additional text after "Completed" in log lines (if
1228 the queue_time_overall log selector is set), a one-byte patch to exigrep
1229 was needed to allow it to recognize "Completed" as not the last thing in
1232 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1233 patch that reportedly fixes this has been added. I am not expert enough to
1234 create a test for it. This is what the patch creator wrote:
1236 "I found a little strange behaviour of ldap code when working with
1237 Windows 2003 AD Domain, where users was placed in more than one
1238 Organization Units. When I tried to give exim partial DN, the exit code
1239 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1240 But simultaneously result of request was absolutely normal ldap result,
1241 so I produce this patch..."
1243 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1244 REFERENCE, so I have modified the code to exclude the patch when that macro
1247 55. Some experimental protocols are using DNS PTR records for new purposes. The
1248 keys for these records are domain names, not reversed IP addresses. The
1249 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1250 leaves it alone. Component reversal etc. now happens only for IP addresses.
1253 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1255 57. Double the size of the debug message buffer (to 2048) so that more of very
1256 long debug lines gets shown.
1258 58. The exicyclog utility now does better if the number of log files to keep
1259 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1261 59. Two changes related to the smtp_active_hostname option:
1263 (1) $smtp_active_hostname is now available as a variable.
1264 (2) The default for smtp_banner uses $smtp_active_hostname instead
1265 of $primary_hostname.
1267 60. The host_aton() function is supposed to be passed a string that is known
1268 to be a valid IP address. However, in the case of IPv6 addresses, it was
1269 not checking this. This is a hostage to fortune. Exim now panics and dies
1270 if the condition is not met. A case was found where this could be provoked
1271 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1272 components; fortuitously, this particular loophole had already been fixed
1273 by change 4.50/55 above.
1275 If there are any other similar loopholes, the new check in host_aton()
1276 itself should stop them being exploited. The report I received stated that
1277 data on the command line could provoke the exploit when Exim was running as
1278 exim, but did not say which command line option was involved. All I could
1279 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1280 running as the user.
1283 61. There was a buffer overflow vulnerability in the SPA authentication code
1284 (which came originally from the Samba project). I have added a test to the
1285 spa_base64_to_bits() function which I hope fixes it.
1288 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1289 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1291 63. The daemon start-up calls getloadavg() while still root for those OS that
1292 need the first call to be done as root, but it missed one case: when
1293 deliver_queue_load_max is set with deliver_drop_privilege. This is
1294 necessary for the benefit of the queue runner, because there is no re-exec
1295 when deliver_drop_privilege is set.
1297 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1298 This has been fixed.
1300 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1301 was in use, was not putting the data itself into the right store pool;
1302 consequently, it could be overwritten for a subsequent message in the same
1303 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1306 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1308 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1309 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1310 the function were treating the return as a boolean value, which happened to
1311 work because 0=false and not-0=true, but is not correct code.
1313 68. The host_aton() function was not handling scoped IPv6 addresses (those
1314 with, for example, "%eth0" on the end) correctly.
1316 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1317 negated items (that is, ~something) in unsigned ints. Some compilers
1318 apparently mutter when there is no cast.
1320 70. If an address verification called from an ACL failed, and did not produce a
1321 user-specific message (i.e. there was only a "system" message), nothing was
1322 put in $acl_verify_message. In this situation, it now puts the system
1325 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1326 session; change 4.31/43 added the unwanted input to the log line - except
1327 that it did not do this in the start of session case. It now does.
1329 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1330 This is wrong and can cause the other end to generate a synchronization
1331 error if it is another Exim or anything else that does the synchronization
1332 check. A QUIT command is no longer sent after a timeout.
1334 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1335 during host lookups.
1337 74. The defer_ok option of callout verification was not working if it was used
1338 when verifying addresses in header lines, that is, for this case:
1340 verify = header_sender/callout=defer_ok
1342 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1343 those file descriptors could be used for SMTP connections. If anything
1344 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1345 could be sent to the SMTP client, causing chaos. The daemon now opens
1346 stdin, stdout, and stderr to /dev/null when it puts itself into the
1349 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1350 log by default. The user can override this with suitable Perl magic.
1352 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1353 the log message when discard triggers, was not working for the DATA ACL or
1354 for the non-SMTP ACL.
1356 78. Error message wording change in sieve.c.
1358 79. If smtp_accept_max_per_host was set, the number of connections could be
1359 restricted to fewer than expected, because the daemon was trying to set up
1360 a new connection before checking whether the processes handling previous
1361 connections had finished. The check for completed processes is now done
1362 earlier. On busy systems, this bug wouldn't be noticed because something
1363 else would have woken the daemon, and it would have reaped the completed
1366 80. If a message was submitted locally by a user whose login name contained one
1367 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1368 It caused a spool format error. I have fixed the spool reading code. A
1369 related problem was that the "from" clause in the Received: line became
1370 illegal because of the space(s). It is now covered by ${quote_local_part.
1372 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1375 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1377 83. Patch from Sieve maintainer to fix unterminated string problem in
1378 "vacation" handling.
1380 84. Some minor changes to the Linux configuration files to help with other
1381 OS variants using glibc.
1383 85. One more patch for Sieve to update vacation handling to latest spec.
1386 ----------------------------------------------------
1387 See the note above about the 4.44 and 4.50 releases.
1388 ----------------------------------------------------
1394 1. Change 4.43/35 introduced a bug that caused file counts to be
1395 incorrectly computed when quota_filecount was set in an appendfile
1398 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1399 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1402 3. Give more explanation in the error message when the command for a transport
1403 filter fails to execute.
1405 4. There are several places where Exim runs a non-Exim command in a
1406 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1407 was being done only for the command run by the queryprogram router. It is
1408 now done for all such subprocesses. The other cases are: ${run, transport
1409 filters, and the commands run by the lmtp and pipe transports.
1411 5. Some older OS have a limit of 256 on the maximum number of file
1412 descriptors. Exim was using setrlimit() to set 1000 as a large value
1413 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1414 systems. I've change it so that if it can't get 1000, it tries for 256.
1416 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1417 was an oversight, and furthermore, ever since the addition of extra
1418 controls (e.g. 4.43/32), the checks on when to allow different forms of
1419 "control" were broken. There should now be diagnostics for all cases when a
1420 control that does not make sense is encountered.
1422 7. $recipients is now available in the predata ACL (oversight).
1424 8. Tidy the search cache before the fork to do a delivery from a message
1425 received from the command line. Otherwise the child will trigger a lookup
1426 failure and thereby defer the delivery if it tries to use (for example) a
1427 cached ldap connection that the parent has called unbind on.
1429 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1430 of $address_data from the recipient verification was clobbered by the
1431 sender verification.
1433 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1434 was its contents. (It was OK if the option was not defined at all.)
1436 11. A "Completed" log line is now written for messages that are removed from
1437 the spool by the -Mrm option.
1439 12. $host_address is now set to the target address during the checking of
1440 ignore_target_hosts.
1442 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1443 being passed; this would have caused $sender_host_name to have been used if
1444 matching the list had actually called for a host name (not very likely,
1445 since this list is usually IP addresses). A host name is now passed as
1448 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1449 code that specifies a non-restarting handler (typically sigaction() in
1450 modern systems) in an attempt to fix a rare and obscure crash bug.
1452 15. Narrowed the window for a race in the daemon that could cause it to ignore
1453 SIGCHLD signals. This is not a major problem, because they are used only to
1454 wake it up if nothing else does.
1456 16. A malformed maildirsize file could cause Exim to calculate negative values
1457 for the mailbox size or file count. Odd effects could occur as a result.
1458 The maildirsize information is now recalculated if the size or filecount
1461 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1462 support for a long time. Removed HAVE_SYS_VFS_H.
1464 18. Updated exipick to current release from John Jetmore.
1466 19. Allow an empty sender to be matched against a lookup in an address list.
1467 Previously the only cases considered were a regular expression, or an
1470 20. Exim went into a mad DNS lookup loop when doing a callout where the
1471 host was specified on the transport, if the DNS lookup yielded more than
1474 21. The RFC2047 encoding function was originally intended for short strings
1475 such as real names; it was not keeping to the 75-character limit for
1476 encoded words that the RFC imposes. It now respects the limit, and
1477 generates multiple encoded words if necessary. To be on the safe side, I
1478 have increased the buffer size for the ${rfc2047: expansion operator from
1481 22. Failure to deliver a bounce message always caused it to be frozen, even if
1482 there was an errors_to setting on the router. The errors_to setting is now
1485 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1486 canonical form (fully expanded) before being placed in
1487 $sender_host_address.
1489 24. Updated eximstats to version 1.33
1491 25. Include certificate and key file names in error message when GnuTLS fails
1492 to set them up, because the GnuTLS error message doesn't include the name
1493 of the failing file when there is a problem reading it.
1495 26. Expand error message when OpenSSL has problems setting up cert/key files.
1498 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1499 (this can affect the format of dates).
1501 28. exim_tidydb, when checking for the continued existence of a message for
1502 which it has found a message-specific retry record, was not finding
1503 messages that were in split spool directories. Consequently, it was
1504 deleting retry records that should have stayed in existence.
1506 29. eximstats updated to version 1.35
1507 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1508 1.35 - bugfix such that pie charts by volume are generated correctly
1510 30. The SPA authentication driver was not abandoning authentication and moving
1511 on to the next authenticator when an expansion was forced to fail,
1512 contradicting the general specification for all authenticators. Instead it
1513 was generating a temporary error. It now behaves as specified.
1515 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1516 (the order specifies the preference for clients). The order is now AES256,
1517 AES128, 3DES, ARCFOUR128.
1519 31. Small patch to Sieve code - explicitly set From: when generating an
1522 32. Exim crashed if a remote delivery caused a very long error message to be
1523 recorded - for instance if somebody sent an entire SpamAssassin report back
1524 as a large number of 550 error lines. This bug was coincidentally fixed by
1525 increasing the size of one of Exim's internal buffers (big_buffer) that
1526 happened as part of the Exiscan merge. However, to be on the safe side, I
1527 have made the code more robust (and fixed the comments that describe what
1530 33. Some experimental protocols are using DNS PTR records for new purposes. The
1531 keys for these records are domain names, not reversed IP addresses. The
1532 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1533 leaves it alone. Component reversal etc. now happens only for IP addresses.
1536 34. The host_aton() function is supposed to be passed a string that is known
1537 to be a valid IP address. However, in the case of IPv6 addresses, it was
1538 not checking this. This is a hostage to fortune. Exim now panics and dies
1539 if the condition is not met. A case was found where this could be provoked
1540 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1541 components; fortuitously, this particular loophole had already been fixed
1542 by change 4.50/55 or 4.44/33 above.
1544 If there are any other similar loopholes, the new check in host_aton()
1545 itself should stop them being exploited. The report I received stated that
1546 data on the command line could provoke the exploit when Exim was running as
1547 exim, but did not say which command line option was involved. All I could
1548 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1549 running as the user.
1552 35. There was a buffer overflow vulnerability in the SPA authentication code
1553 (which came originally from the Samba project). I have added a test to the
1554 spa_base64_to_bits() function which I hope fixes it.
1557 36. The daemon start-up calls getloadavg() while still root for those OS that
1558 need the first call to be done as root, but it missed one case: when
1559 deliver_queue_load_max is set with deliver_drop_privilege. This is
1560 necessary for the benefit of the queue runner, because there is no re-exec
1561 when deliver_drop_privilege is set.
1563 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1564 was in use, was not putting the data itself into the right store pool;
1565 consequently, it could be overwritten for a subsequent message in the same
1566 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1569 38. Sometimes the final signoff response after QUIT could fail to get
1570 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1571 < 0 before doing a fflush(). This bug looks as though it goes back to the
1572 introduction of TLS in release 3.20, but "sometimes" must have been rare
1573 because the tests only now provoked it.
1579 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1580 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1581 Now it returns an integer. A number of calls were still expecting a T/F
1582 return. Fortuitously, in all cases, the tests worked in OK situations,
1583 which is the norm. However, things would have gone wrong on any write
1584 failures on the smtp file descriptor. This function is used when sending
1585 messages over SMTP and also when doing verify callouts.
1587 2. When Exim is called to do synchronous delivery of a locally submitted
1588 message (the -odf or -odi options), it no longer closes stderr before doing
1591 3. Implemented the mua_wrapper option.
1593 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1595 5. Implemented the functions header_remove(), header_testname(),
1596 header_add_at_position(), and receive_remove_recipient(), and exported them
1599 6. If an ACL "warn" statement specified the addition of headers, Exim already
1600 inserted X-ACL-Warn: at the start if there was no header name. However, it
1601 was not making this test for the second and subsequent header lines if
1602 there were newlines in the string. This meant that an invalid header could
1603 be inserted if Exim was badly configured.
1605 7. Allow an ACL "warn" statement to add header lines at the start or after all
1606 the Received: headers, as well as at the end.
1608 8. Added the rcpt_4xx retry error code.
1610 9. Added postmaster_mailfrom=xxx to callout verification option.
1612 10. Added mailfrom=xxxx to the callout verification option, for verify=
1615 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1616 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1618 12. Inserted some casts to stop certain compilers warning when using pointer
1619 differences as field lengths or precisions in printf-type calls (mostly
1620 affecting debugging statements).
1622 13. Added optional readline() support for -be (dynamically loaded).
1624 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1625 same clock tick as a message's arrival, so that its received time was the
1626 same as the "first fail" time on the retry record, and that message
1627 remained on the queue past the ultimate address timeout, every queue runner
1628 would try a delivery (because it was past the ultimate address timeout) but
1629 after another failure, the ultimate address timeout, which should have then
1630 bounced the address, did not kick in. This was a "< instead of <=" error;
1631 in most cases the first failure would have been in the next clock tick
1632 after the received time, and all would be well.
1634 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1635 being recognized when the domain list was tested by the match_domain
1636 condition in an expansion string.
1638 16. Added the ${str2b64: operator.
1640 17. Exim was always calling setrlimit() to set a large limit for the number of
1641 processes, without checking whether the existing limit was already
1642 adequate. (It did check for the limit on file descriptors.) Furthermore,
1643 errors from getrlimit() and setrlimit() were being ignored. Now they are
1644 logged to the main and panic logs, but Exim does carry on, to try to do its
1645 job under whatever limits there are.
1647 18. Imported PCRE 5.0.
1649 19. Trivial typo in log message " temporarily refused connection" (the leading
1652 20. If the log selector return_path_on_delivery was set and an address was
1653 redirected to /dev/null, the delivery process crashed because it assumed
1654 that a return path would always be set for a "successful" delivery. In this
1655 case, the whole delivery is bypassed as an optimization, and therefore no
1658 21. Internal re-arrangement: the function for sending a challenge and reading
1659 a response while authentication was assuming a zero-terminated challenge
1660 string. It's now changed to take a pointer and a length, to allow for
1661 binary data in such strings.
1663 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1665 23. Exim was not respecting finduser_retries when seeking the login of the
1666 uid under which it was called; it was always trying 10 times. (The default
1667 setting of finduser_retries is zero.) Also, it was sleeping after the final
1668 failure, which is pointless.
1670 24. Implemented tls_on_connect_ports.
1672 25. Implemented acl_smtp_predata.
1674 26. If the domain in control=submission is set empty, Exim assumes that the
1675 authenticated id is a complete email address when it generates From: or
1676 Sender: header lines.
1678 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1679 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1680 chown and chgrp in /bin and hostname in /usr/bin.
1682 28. Exim was keeping the "process log" file open after each use, just as it
1683 does for the main log. This opens the possibility of it remaining open for
1684 long periods when the USR1 signal hits a daemon. Occasional processlog
1685 errors were reported, that could have been caused by this. Anyway, it seems
1686 much more sensible not to leave this file open at all, so that is what now
1689 29. The long-running daemon process does not normally write to the log once it
1690 has entered its main loop, and it closes the log before doing so. This is
1691 so that log files can straightforwardly be renamed and moved. However,
1692 there are a couple of unusual error situations where the daemon does write
1693 log entries, and I had neglected to close the log afterwards.
1695 30. The text of an SMTP error response that was received during a remote
1696 delivery was being truncated at 512 bytes. This is too short for some of
1697 the long messages that one sometimes sees. I've increased the limit to
1700 31. It is now possible to make retry rules that apply only when a message has a
1701 specific sender, in particular, an empty sender.
1703 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1704 it possible to be selective about when SMTP synchronization is enforced.
1706 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1708 32. Implemented hosts_connection_nolog.
1710 33. Added an ACL for QUIT.
1712 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1715 35. Added mailbox_size and mailbox_filecount to appendfile.
1717 36. Added control = no_multiline_responses to ACLs.
1719 37. There was a bug in the logic of the code that waits for the clock to tick
1720 in the case where the clock went backwards by a substantial amount such
1721 that the microsecond fraction of "now" was more than the microsecond
1722 fraction of "then" (but the whole seconds number was less).
1724 38. Added support for the libradius Radius client library this is found on
1725 FreeBSD (previously only the radiusclient library was supported).
1731 1. When certain lookups returned multiple values in the form name=value, the
1732 quoting of the values was not always being done properly. Specifically:
1733 (a) If the value started with a double quote, but contained no whitespace,
1735 (b) If the value contained whitespace other than a space character (i.e.
1736 tabs or newlines or carriage returns) it was not quoted.
1737 This fix has been applied to the mysql and pgsql lookups by writing a
1738 separate quoting function and calling it from the lookup code. The fix
1739 should probably also be applied to nisplus, ibase and oracle lookups, but
1740 since I cannot test any of those, I have not disturbed their existing code.
1742 2. A hit in the callout cache for a specific address caused a log line with no
1743 reason for rejecting RCPT. Now it says "Previous (cached) callout
1744 verification failure".
1746 3. There was an off-by-one bug in the queryprogram router. An over-long
1747 return line was truncated at 256 instead of 255 characters, thereby
1748 overflowing its buffer with the terminating zero. As well as fixing this, I
1749 have increased the buffer size to 1024 (and made a note to document this).
1751 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1752 when Exim is waiting for an SMTP response from a remote server, Exim
1753 restarts its select() call on the socket, thereby resetting its timeout.
1754 This is not a problem when such interrupts are rare. Somebody set up a cron
1755 job to run exiwhat every 2 minutes, which is less than the normal select()
1756 timeout (5 or 10 minutes). This meant that the select() timeout never
1757 kicked in because it was always reset. I have fixed this by comparing the
1758 time when an interrupt arrives with the time at the start of the first call
1759 to select(). If more time than the timeout has elapsed, the interrupt is
1760 treated as a timeout.
1762 5. Some internal re-factoring in preparation for the addition of Sieve
1763 extensions (by MH). In particular, the "personal" test is moved to a
1764 separate function, and given an option for scanning Cc: and Bcc: (which is
1765 not set for Exim filters).
1767 6. When Exim created an email address using the login of the caller as the
1768 local part (e.g. when creating a From: or Sender: header line), it was not
1769 quoting the local part when it contained special characters such as @.
1771 7. Installed new OpenBSD configuration files.
1773 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1774 try to make them clearer.
1776 9. Callout options, other than the timeout value, were being ignored when
1777 verifying sender addresses in header lines. For example, when using
1779 verify = header_sender/callout=no_cache
1781 the cache was (incorrectly) being used.
1783 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1784 only the Cygwin environment.
1786 11. When return_path_on_delivery was set as a log selector, if different remote
1787 addresses in the same message used different return paths and parallel
1788 remote delivery occurred, the wrong values would sometimes be logged.
1789 (Whenever a remote delivery process finished, the return path value from
1790 the most recently started remote delivery process was logged.)
1792 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1793 lines when AUTH and/or TLS are in use. This is the "received protocol"
1794 field. Exim used to use "asmtp" for authenticated SMTP, without any
1795 indication (in the protocol name) for TLS use. Now it follows the RFC and
1796 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1797 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1798 names appear in log lines as well as in Received: header lines.
1800 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1801 extensions, and comparison tests, and to fix some bugs.
1803 14. Changes to the "personal" filter test:
1805 (1) The test was buggy in that it was just doing the equivalent of
1806 "contains" tests on header lines. For example, if a user's address was
1807 anne@some.where, the "personal" test would incorrectly be true for
1809 To: susanne@some.where
1811 This test is now done by extracting each address from the header in turn,
1812 and checking the entire address. Other tests that are part of "personal"
1813 are now done using regular expressions (for example, to check local parts
1814 of addresses in From: header lines).
1816 (2) The list of non-personal local parts in From: addresses has been
1817 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1818 taken from the Sieve specification recommendations.
1820 (3) If the message contains any header line starting with "List-" it is
1821 treated as non-personal.
1823 (4) The test for "circular" in the Subject: header line has been removed
1824 because it now seems ill-conceived.
1826 15. Minor typos in src/EDITME comments corrected.
1828 16. Installed latest exipick from John Jetmore.
1830 17. If headers_add on a router specified a text string that was too long for
1831 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1832 of string_sprintf() is now avoided.
1834 18. $message_body_size was not set (it was always zero) when running the DATA
1835 ACL and the local_scan() function.
1837 19. For the "mail" command in an Exim filter, no default was being set for
1838 the once_repeat time, causing a random time value to be used if "once" was
1839 specified. (If the value happened to be <= 0, no repeat happened.) The
1840 default is now 0s, meaning "never repeat". The "vacation" command was OK
1841 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1842 (I found it when inspecting the code).
1844 20. There is now an overall timeout for performing a callout verification. It
1845 defaults to 4 times the callout timeout, which applies to individual SMTP
1846 commands during the callout. The overall timeout applies when there is more
1847 than one host that can be tried. The timeout is checked before trying the
1848 next host. This prevents very long delays if there are a large number of
1849 hosts and all are timing out (e.g. when the network connections are timing
1850 out). The value of the overall timeout can be changed by specifying an
1851 additional sub-option for "callout", called "maxwait". For example:
1853 verify = sender/callout=5s,maxwait=20s
1855 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1856 to the end before writing, but this should make it even safer).
1858 22. Exim was forgetting that it had advertised PIPELINING for the second and
1859 subsequent messages on an SMTP connection. It was also not resetting its
1860 memory on STARTTLS and an internal HELO.
1862 23. When Exim logs an SMTP synchronization error within a session, it now
1863 records whether PIPELINING has been advertised or not.
1865 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1866 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1867 rather than long int.
1869 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1871 26. Added the never_mail option to autoreply.
1877 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1878 crash if the getsockname() call failed; this can happen if a connection is
1879 closed very soon after it is established. The problem was simply in the
1880 order in which certain operations were done, causing Exim to try to write
1881 to the SMTP stream before it had set up the file descriptor. The bug has
1882 been fixed by making things happen in the correct order.
1888 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1889 before the connection was closed, thus losing the rejection response.
1891 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1892 some early Solaris releases, but causes trouble in current releases where
1893 socklen_t is defined.
1895 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1898 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1901 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1902 that releasing the top of it at the end releases what was used for sub-
1903 expansions (unless the block got too big). However, discard this block if
1904 the first thing is a variable or header, so that we can use its block when
1905 it is dynamic (useful for very large $message_headers, for example).
1907 6. Lookups now cache *every* query, not just the most recent. A new, separate
1908 store pool is used for this. It can be recovered when all lookup caches are
1909 flushed. Lookups now release memory at the end of their result strings.
1910 This has involved some general refactoring of the lookup sources.
1912 7. Some code has been added to the store_xxx() functions to reduce the amount
1913 of flapping under certain conditions.
1915 8. log_incoming_interface used to affect only the <= reception log lines. Now
1916 it causes the local interface and port to be added to several more SMTP log
1917 lines, for example "SMTP connection from", and rejection lines.
1919 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1921 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1923 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1924 could be overwritten at the end of the current message (or the start of a
1925 new message if it was set in a HELO ACL). The value is now preserved for
1926 the duration of the SMTP connection.
1928 12. If a transport had a headers_rewrite setting, and a matching header line
1929 contained an unqualified address, that address was qualified, even if it
1930 did not match any rewriting rules. The underlying bug was that the values
1931 of the flags that permit the existence of unqualified sender and recipient
1932 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1933 non-local messages, and by -bnq for local messages) were not being
1934 preserved with the message after it was received.
1936 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1937 "next input=" as part of the text comprising the host identity instead of
1938 the correct text. The code was using the same buffer for two different
1939 strings. However, depending on which order the printing function evaluated
1940 its arguments, the bug did not always show up. Under Linux, for example, my
1941 test suite worked just fine.
1943 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1944 doesn't work with some older versions of Perl. It has been changed to "my",
1945 which in any case is probably the better facility to use.
1947 15. A really picky compiler found some instances of statements for creating
1948 error messages that either had too many or two few arguments for the format
1951 16. The size of the buffer for calls to the DNS resolver has been increased
1952 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1953 for addresses that have a lot of PTR records. This alleviates a problem; it
1954 does not fully solve it.
1956 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1957 buffer now truncates the list and logs the incident, which is the same
1958 action as happens when Exim is looking up a host name and its aliases.
1959 Previously in this situation something unpredictable would happen;
1960 sometimes it was "internal error: store_reset failed".
1962 18. If a server dropped the connection unexpectedly when an Exim client was
1963 using GnuTLS and trying to read a response, the client delivery process
1964 crashed while trying to generate an error log message.
1966 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1967 string, for example:
1969 warn message = H1: something\nH2: something
1971 the text was added as a single header line from Exim's point of view
1972 though it ended up OK in the delivered message. However, searching for the
1973 second and subsequent header lines using $h_h2: did not work. This has been
1974 fixed. Similarly, if a system filter added multiple headers in this way,
1975 the routers could not see them.
1977 20. Expanded the error message when iplsearch is called with an invalid key to
1978 suggest using net-iplsearch in a host list.
1980 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1981 ACLs are no longer actually imposed (and a message to that effect is
1984 22. If a "gecos" field in a passwd entry contained escaped characters, in
1985 particular, if it contained a \" sequence, Exim got it wrong when building
1986 a From: or a Sender: header from that name. A second bug also caused
1987 incorrect handling when an unquoted " was present following a character
1988 that needed quoting.
1990 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1991 was not being matched caselessly.
1993 24. Arranged for all hyphens in the exim.8 source to be escaped with
1996 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1997 properly. Recipient callout cache records were still being keyed to include
1998 the sender, even when use_sender was set false. This led to far more
1999 callouts that were necessary. The sender is no longer included in the key
2000 when use_sender is false.
2002 26. Added "control = submission" modifier to ACLs.
2004 27. Added the ${base62d: operator to decode base 62 numbers.
2006 28. dnsdb lookups can now access SRV records.
2008 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2009 the configuration file.
2011 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2012 (-v) mode. This makes the output for a verbose queue run more intelligible.
2014 31. Added a use_postmaster feature to recipient callouts.
2016 32. Added the $body_zerocount variable, containing the number of binary zero
2017 bytes in the message body.
2019 33. The time of last modification of the "new" subdirectory is now used as the
2020 "mailbox time last read" when there is a quota error for a maildir
2023 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2025 35. Added +ignore_unknown as a special item in host lists.
2027 36. Code for decoding IPv6 addresses in host lists is now included, even if
2028 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2029 address was recognized as an IP address, but was then not correctly decoded
2030 into binary, causing unexpected and incorrect effects when compared with
2037 1. Very minor rewording of debugging text in manualroute to say "list of
2038 hosts" instead of "hostlist".
2040 2. If verify=header_syntax was set, and a header line with an unqualified
2041 address (no domain) and a large number of spaces between the end of the
2042 name and the colon was received, the reception process suffered a buffer
2043 overflow, and (when I tested it) crashed. This was caused by some obsolete
2044 code that should have been removed. The fix is to remove it!
2046 3. When running in the test harness, delay a bit after writing a bounce
2047 message to get a bit more predictability in the log output.
2049 4. Added a call to search_tidyup() just before forking a reception process. In
2050 theory, someone could use a lookup in the expansion of smtp_accept_max_
2051 per_host which, without the tidyup, could leave open a database connection.
2053 5. Added the variables $recipient_data and $sender_data which get set from a
2054 lookup success in an ACL "recipients" or "senders" condition, or a router
2055 "senders" option, similar to $domain_data and $local_part_data.
2057 6. Moved the writing of debug_print from before to after the "senders" test
2060 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2061 problems for message scanning, either using a data ACL, or using
2062 local_scan() because the Received: header was not generated till after they
2063 were called (in order to set the time as the time of reception completion).
2064 I have revised the way this works. The header is now generated after the
2065 body is received, but before the ACL or local_scan() are called. After they
2066 are run, the timestamp in the header is updated.
2072 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2073 before starting a queue runner without re-exec. This happened only when
2074 deliver_drop_privilege was set or when the Exim user was set to root. The
2075 effect of the bug was that timeouts during subsequent deliveries caused
2076 crashes instead of being properly handled. The handler is now left at its
2077 default (and expected) setting.
2079 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2080 message, again when deliver_drop_privilege is set or Exim is run as root.
2081 The bug described in (1) was not present in this case, but the tidying up
2082 of the other signals was missing. I have made the two cases consistent.
2084 3. The ignore_target_hosts setting on a manualroute router was being ignored
2085 for hosts that were looked up using the /MX notation.
2087 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2090 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2091 operated on the sender address. After changing the $sender_address to <>
2092 for the sender address verify, Exim was re-instated it as the original
2093 (before rewriting) address, but remembering that it had rewritten it, so it
2094 wasn't rewriting it again. This bug also had the effect of breaking the
2095 sender address verification caching when the sender address was rewritten.
2097 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2098 This has been changed so that if the ip literal address matches
2099 ignore_target_hosts, the router declines.
2101 7. Added expansion conditions match_domain, match_address, and match_local_
2102 part (NOT match_host).
2104 8. The placeholder for the Received: header didn't have a length field set.
2106 9. Added code to Exim itself and to exim_lock to test for a specific race
2107 condition that could lead to file corruption when using MBX delivery. The
2108 issue is with the lockfile that is created in /tmp. If this file is removed
2109 after a process has opened it but before that process has acquired a lock,
2110 there is the potential for a second process to recreate the file and also
2111 acquire a lock. This could lead to two Exim processes writing to the file
2112 at the same time. The added code performs the same test as UW imapd; it
2113 checks after acquiring the lock that its file descriptor still refers to
2114 the same named file.
2116 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2117 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2120 11. Added the smtp_active_hostname option. If used, this will typically be made
2121 to depend on the incoming interface address. Because $interface_address is
2122 not set up until the daemon has forked a reception process, error responses
2123 that can happen earlier (such as "too many connections") no longer contain
2126 12. If an expansion in a condition on a "warn" statement fails because a lookup
2127 defers, the "warn" statement is abandoned, and the next ACL statement is
2128 processed. Previously this caused the whole ACL to be aborted.
2130 13. Added the iplsearch lookup type.
2132 14. Added ident_timeout as a log selector.
2134 15. Added tls_certificate_verified as a log selector.
2136 16. Added a global option tls_require_ciphers (compare the smtp transport
2137 option of the same name). This controls incoming TLS connections.
2139 17. I finally figured out how to make tls_require_ciphers do a similar thing
2140 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2141 before starting the TLS session.
2143 18. Tabs are now shown as \t in -bP output.
2145 19. If the log selector return_path_on_delivery was set, Exim crashed when
2146 bouncing a message because it had too many Received: header lines.
2148 20. If two routers both had headers_remove settings, and the first one included
2149 a superfluous trailing colon, the final name in the first list and the
2150 first name in the second list were incorrectly joined into one item (with a
2151 colon in the middle).
2157 1. Added -C and -D options to the exinext utility, mainly to make it easier
2158 to include in the automated testing, but these could be helpful when
2159 multiple configurations are in use.
2161 2. The exinext utility was not formatting the output nicely when there was
2162 an alternate port involved in the retry record key, nor when there was a
2163 message id as well (for retries that were specific to a specific message
2164 and a specific host). It was also confused by IPv6 addresses, because of
2165 the additional colons they contain. I have fixed the IPv4 problem, and
2166 patched it up to do a reasonable job for IPv6.
2168 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2169 delivery, the log line now contains "pipelined" if PIPELINING was used.
2171 4. An SMTP transport process used to panic and die if the bind() call to set
2172 an explicit outgoing interface failed. This has been changed; it is now
2173 treated in the same way as a connect() failure.
2175 5. A reference to $sender_host_name in the part of a conditional expansion
2176 that was being skipped was still causing a DNS lookup. This no longer
2179 6. The def: expansion condition was not recognizing references to header lines
2180 that used bh_ and bheader_.
2182 7. Added the _cache feature to named lists.
2184 8. The code for checking quota_filecount in the appendfile transport was
2185 allowing one more file than it should have been.
2187 9. For compatibility with Sendmail, the command line option
2195 and sets the incoming protocol and host name (for trusted callers). The
2196 host name and its colon can be omitted when only the protocol is to be set.
2197 Note the Exim already has two private options, -pd and -ps, that refer to
2198 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2199 "s", but I don't think that's a major issue.
2201 10. A number of refactoring changes to the code, none of which should affect
2204 (a) The number of logging options was getting close to filling up the
2205 32-bit word that was used as a bit map. I have split them into two classes:
2206 those that are passed in the argument to log_write(), and those that are
2207 only ever tested independently outside of that function. These are now in
2208 separate 32-bit words, so there is plenty of room for expansion again.
2209 There is no change in the user interface or the logging behaviour.
2211 (b) When building, for example, log lines, the code previously used a
2212 macro that called string_cat() twice, in order to add two strings. This is
2213 not really sufficiently general. Furthermore, there was one instance where
2214 it was actually wrong because one of the argument was used twice, and in
2215 one call a function was used. (As it happened, calling the function twice
2216 did not affect the overall behaviour.) The macro has been replaced by a
2217 function that can join an arbitrary number of extra strings onto a growing
2220 (c) The code for expansion conditions now uses a table and a binary chop
2221 instead of a serial search (which was left over from when there were very
2222 few conditions). Also, it now recognizes conditions like "pam" even when
2223 the relevant support is not compiled in: a suitably worded error message is
2224 given if an attempt is made to use such a condition.
2226 11. Added ${time_interval:xxxxx}.
2228 12. A bug was causing one of the ddress fields not to be passed back correctly
2229 from remote delivery subprocesses. The field in question was not being
2230 subsequently used, so this caused to problems in practice.
2232 13. Added new log selectors queue_time and deliver_time.
2234 14. Might have fixed a bug in maildirsizefile handling that threw up
2235 "unexpected character" debug warnings, and recalculated the data
2236 unnecessarily. In any case, I expanded the warning message to give more
2239 15. Added the message "Restricted characters in address" to the statements in
2240 the default ACL that block characters like @ and % in local parts.
2242 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2243 Three changes have been made:
2245 (a) There was a serious bug; a negative response to MAIL caused the whole
2246 recipient domain to be cached as invalid, thereby blocking all messages
2247 to all local parts at the same domain, from all senders. This bug has
2248 been fixed. The domain is no longer cached after a negative response to
2249 MAIL if the sender used is not empty.
2251 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2254 (c) A new callout option, "use_sender" has been added for people who want
2255 the modified behaviour.
2261 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2264 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2265 indeed breaks things for older releases.
2267 3. Added additional logging to the case where there is a problem reading data
2268 from a filter that is running in a subprocess using a pipe, in order to
2269 try to track down a specific problem.
2271 4. Testing facility fudge: when running in the test harness and attempting
2272 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2273 getting "No route to host". Convert this to a timeout.
2275 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2278 6. Some OS don't have socklen_t but use size_t instead. This affects the
2279 fifth argument of getsockopt() amongst other things. This is now
2280 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2281 can be set for individual OS. I have set it for SunOS5, OSF1, and
2282 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2283 some earlier ones do not.
2285 7. Change 4.30/15 was not doing the test caselessly.
2287 8. The standard form for an IPv6 address literal was being rejected by address
2288 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2289 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2290 this, as well as the form without the "IPv6" on the front (but only when
2291 address literals are enabled, of course).
2293 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2295 10. Exim crashed if a message with an empty sender address specified by -f
2296 encountered a router with an errors_to setting. This could be provoked only
2297 by a command such as
2301 where an empty string was supplied; "<>" did not hit this bug.
2303 11. Installed PCRE release 4.5.
2305 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2306 remained set. It is now erased.
2308 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2309 times from message ids (which are base 36 rather than the normal 62).
2311 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2312 were being counted as actual protocol errors, and logged if the log
2313 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2314 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2315 and DATA following a set of rejected RCPTs do not count as protocol errors.
2316 In other words, Exim assumes they were pipelined, though this may not
2317 actually be the case. Of course, in all cases the client gets an
2318 appropriate error code.
2320 15. If a lookup fails in an ACL condition, a message about the failure may
2321 be available; it is used if testing the ACL cannot continue, because most
2322 such messages specify what the cause of the deferral is. However, some
2323 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2324 that caused an old message to be retained and used if a later statement
2325 caused a defer, replacing the real cause of the deferral.
2327 16. If an IP address had so many PTR records that the DNS lookup buffer
2328 was not large enough to hold them, Exim could crash while trying to process
2329 the truncated data. It now detects and logs this case.
2331 17. Further to 4.21/58, another change has been made: if (and only if) the
2332 first line of a message (the first header line) ends with CRLF, a bare LF
2333 in a subsequent header line has a space inserted after it, so as not to
2334 terminate the header.
2336 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2337 unnecessarily, used atoi() instead of strtol(), and didn't check the
2338 termination when getting file sizes from file names by regex.
2340 19. Completely re-implemented the support for maildirsize files, in the light
2341 of a number of problems with the previous contributed implementation
2342 (4.30/29). In particular:
2344 . If the quota is zero, the maildirsize file is maintained, but no quota is
2347 . If the maildir directory does not exist, it is created before any attempt
2348 to write a maildirsize file.
2350 . The quota value in the file is just a cache; if the quota is changed in
2351 the transport, the new value overrides.
2353 . A regular expression is available for excluding directories from the
2356 20. The autoreply transport checks the characters in options that define the
2357 message's headers; it allows continued headers, but it was checking with
2358 isspace() after an embedded newline instead of explicitly looking for a
2361 21. If all the "regular" hosts to which an address was routed had passed their
2362 expiry times, and had not reached their retry times, the address was
2363 bounced, even if fallback hosts were defined. Now Exim should go on to try
2366 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2367 equivalent code in the SMTP transport. Some hosts send humungous responses
2368 to HELO/EHLO, more than 1024 it seems.
2370 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2371 gives compiler warnings in some environments. I've now done it "properly",
2374 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2375 (because of problems with the built-in one) was declared to return uschar *
2376 instead of char *, causing compiler failure.
2378 25. Fixed a file descriptor leak when processing alias/forward files.
2380 26. Fixed a minor format string issue in dbfn.c.
2382 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2384 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2385 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2388 29. When a sender address is verified, it is cached, to save repeating the test
2389 when there is more than one recipient in a message. However, when the
2390 verification involves a callout, it is possible for different callout
2391 options to be set for different recipients. It is too complicated to keep
2392 track of this in the cache, so now Exim always runs a verification when a
2393 callout is required, relying on the callout cache for the optimization.
2394 The overhead is duplication of the address routing, but this should not be
2397 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2398 to be verified with callout=postmaster, and the main callout worked but the
2399 postmaster check failed, the verification correctly failed. However, if a
2400 subsequent RCPT command asked for sender verification *without* the
2401 postmaster check, incorrect caching caused this verification also to fail,
2404 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2405 it was not caching the DNS options (qualify_single, search_parents) that
2406 were used when the lookup failed. A subsequent lookup with different
2407 options therefore always gave the same answer, though there were cases
2408 where it should not have. (Example: a "domains = !$mx_any" option on a
2409 dnslookup router: the "domains" option is always processed without any
2410 widening, but the router might have qualify_single set.) Now Exim uses the
2411 cached value only when the same options are set.
2413 32. Added John Jetmore's "exipick" utility to the distribution.
2415 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2416 than a timeout (e.g. a certificate is required, and is not provided), an
2417 Exim server now closes the connection immediately. Previously it waited for
2418 the client to close - but if the client is SSL, it seems that they each
2419 wait for each other, leading to a delay before one of them times out.
2421 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2422 maintained 0.8.x compatibility because I don't think many are using it, and
2423 it is clearly obsolete.
2425 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2428 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2429 client certificate was expired. A simple patch fixes this, though I don't
2430 understand the full logic of why the verify callback is called multiple
2433 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2434 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2435 which causes problems with some clients (such as the Certicom SSL Plus
2436 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2437 disables the coutermeasure allowing Eudora to connect."
2439 38. Exim was not checking that a write() to a log file succeeded. This could
2440 lead to Bad Things if a log got too big, in particular if it hit a file
2441 size limit. Exim now panics and dies if it cannot write to a log file, just
2442 as it does if it cannot open a log file.
2444 39. Modified OS/Makefile-Linux so that it now contains
2446 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2448 The two -D definitions ensure that Exim is compiled with large file
2449 support, which makes it possible to handle log files that are bigger than
2452 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2453 instance) a domain was checked against a named list that involved a lookup,
2454 causing $domain_data to be set, then another domain was checked against the
2455 same list, then the first domain was re-checked, the value of $domain_data
2456 after the final check could be wrong. In particular, if the second check
2457 failed, it could be set empty. This bug probably also applied to
2460 41. The strip_trailing_dot option was not being applied to the address given
2461 with the -f command-line option.
2463 42. The code for reading a message's header from the spool was incrementing
2464 $received_count, but never initializing it. This meant that the value was
2465 incorrect (doubled) while delivering a message in the same process in which
2466 it was received. In the most common configuration of Exim, this never
2467 happens - a fresh exec is done - but it can happen when
2468 deliver_drop_privilege is set.
2470 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2471 it now includes up to 150 characters of the unexpected data in the log
2474 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2475 and building data strings. The size of both of these buffers was 10 000
2476 bytes - far larger than anybody would *ever* want, thought I. Needless to
2477 say, somebody hit the limit. I have increased the maximum line length to
2478 20 000 and the maximum data length of concatenated lines to 100 000. I have
2479 also fixed two bugs, because there was no checking on these buffers. Tsk,
2480 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2481 buffer is too small.
2483 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2484 lsearch lookups. Now it does.
2486 46. When parsing a route_list item in a manualroute router, a fixed-length
2487 buffer was used for the list of hosts. I made this 1024 bytes long,
2488 thinking that nobody would ever have a list of hosts that long. Wrong.
2489 Somebody had a whole pile of complicated expansion conditions, and the
2490 string was silently truncated, leading to an expansion error. It turns out
2491 that it is easier to change to an unlimited length (owing to other changes
2492 that have happened since this code was originally written) than to build
2493 structure for giving a limitation error. The length of the item that
2494 expands into the list of hosts is now unlimited.
2496 47. The lsearch lookup could not handle data where the length of text line was
2497 more than 4095 characters. Such lines were truncated, leading to shortened
2498 data being returned. It should now handle lines of any length.
2500 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2501 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2504 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2505 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2506 no longer happens in comment lines. A list of replacements is now placed
2507 at the head of all of the source files, except those whose only change is
2508 to replace PERL_COMMAND in the very first #! line.
2510 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2511 messages on the queue, with a bottom-up merge sort, using code contributed
2512 by Michael Haardt. This should make operations like -bp somewhat faster on
2513 large queues. It won't affect queue runners, except when queue_run_in_order
2516 51. Installed eximstats 1.31 in the distribution.
2518 52. Added support for SRV lookups to the dnslookup router.
2520 53. If an ACL referred to $message_body or $message_body_end, the value was not
2521 reset for any messages that followed in the same SMTP session.
2523 54. The store-handling optimization for building very long strings was not
2524 differentiating between the different store pools. I don't think this
2525 actually made any difference in practice, but I've tidied it.
2527 55. While running the routers to verify a sender address, $sender_address
2528 was still set to the sender address. This is wrong, because when routing to
2529 send a bounce to the sender, it would be empty. Therefore, I have changed
2530 it so that, while verifying a sender address, $sender_address is set to <>.
2531 (There is no change to what happens when verifying a recipient address.)
2533 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2534 target A or AAAA records (if not already returned) without resetting the
2535 qualify_single or search_parents options of the DNS resolver. These are
2536 inappropriate in this case because the targets of MX and SRV records must
2537 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2538 target that, when qualified, matched something in the local domain. These
2539 two options are now turned off when doing these lookups.
2541 57. It seems that at least some releases of Reiserfs (which does not have the
2542 concept of a fixed number of inodes) returns zero and not -1 for the
2543 number of available inodes. This interacted badly with check_spool_inodes,
2544 which assumed that -1 was the "no such thing" setting. What I have done is
2545 to check that the total number of inodes is greater than zero before doing
2546 the test of how many are available.
2548 58. When a "warn" ACL statement has a log_message modifier, the message is
2549 remembered, and not repeated. This is to avoid a lot of repetition when a
2550 message has many recipients that cause the same warning to be written.
2551 Howewer, Exim was preserving the list of already written lines for an
2552 entire SMTP session, which doesn't seem right. The memory is now reset if a
2553 new message is started.
2555 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2556 debugging output unless log_rewrite was also set.
2558 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2559 of (int)(handle) when we know that handle contains (void *)(-1).
2561 61. The Exim daemon panic-logs an error return when it closes the incoming
2562 connection. However "connection reset by peer" seems to be common, and
2563 isn't really an error worthy of noting specially, so that particular error
2566 62. When Exim is trying to find all the local interfaces, it used to panic and
2567 die if the ioctl to get the interface flags failed. However, it seems that
2568 on at least one OS (Solaris 9) it is possible to have an interface that is
2569 included in the list of interfaces, but for which you get a failure error
2570 for this call. This happens when the interface is not "plumbed" into a
2571 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2572 failure of the "get flags" call assumes that the interface is down.
2574 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2575 makes life easier for people who are doing arithmetic on fields extracted
2576 from dates, where you often get leading zeros that should not be
2577 interpreted as octal.
2579 64. Added qualify_domain to the redirect router, to override the global
2582 65. If a pathologically long header line contained very many addresses (the
2583 report of this problem mentioned 10 000) and each of them was rewritten,
2584 Exim could use up a very large amount of memory. (It kept on making new
2585 copies of the header line as it rewrote, and never released the old ones.)
2586 At the expense of a bit more processing, the header rewriting function has
2587 been changed so that it no longer eats memory in this way.
2589 66. The generation of the Received: header has been moved from the time that a
2590 message starts to be received, to the time that it finishes. The timestamp
2591 in the Received: header should now be very close to that of the <= log
2592 line. There are two side-effects of this change:
2594 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2595 logged header lines no longer include the local Received: line, because
2596 it has not yet been created. The same applies to a copy of the message
2597 that is returned to a non-SMTP sender when a message is rejected.
2599 (b) When a filter file is tested using -bf, no additional Received: header
2600 is added to the test message. After some thought, I decided that this
2603 This change does not affect the value of $received_for. It is still set
2604 after address rewriting, but before local_scan() is called.
2606 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2608 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2609 gave an unhelpful panic error message, and a defer error. I have managed to
2610 change this behaviour so that it now rejects any supplied certificate,
2611 which seems right, as the list of acceptable certificates is empty.
2613 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2614 gave an unhelpful defer error. I have not managed to make this reject any
2615 supplied certificates, but the error message it gives is "no certificate
2616 supplied", which is not helpful.
2618 70. exigrep's output now also includes lines that are not associated with any
2619 message, but which match the given pattern. Implemented by a patch from
2620 Martin Sluka, which also tidied up the Perl a bit.
2622 71. Recipient callout verification, like sender verification, was using <> in
2623 the MAIL FROM command. This isn't really the right thing, since the actual
2624 sender may affect whether the remote host accepts the recipient or not. I
2625 have changed it to use the actual sender in the callout; this means that
2626 the cache record is now keyed on a recipient/sender pair, not just the
2627 recipient address. There doesn't seem to be a real danger of callout loops,
2628 since a callout by the remote host to check the sender would use <>.
2629 [SEE ABOVE: changed after hitting problems.]
2631 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2632 temporary errors. However, in the case of such a code being given after
2633 the end of a data transmission (i.e. after ".") Exim was failing to write
2634 a retry record for the message. (Yes, there was some broken host that was
2635 actually sending 8xx at this point.)
2637 73. An unknown lookup type in a host list could cause Exim to panic-die when
2638 the list was checked. (An example that provoked this was putting <; in the
2639 middle of a list instead of at the start.) If this happened during a DATA
2640 ACL check, a -D file could be left lying around. This kind of configuration
2641 error no longer causes Exim to die; instead it causes a defer errror. The
2642 incident is still logged to the main and panic logs.
2644 74. Buglet left over from Exim 3 conversion. The message "too many messages
2645 in one connection" was written to the rejectlog but not the mainlog, except
2646 when address rewriting (yes!) was being logged.
2648 75. Added write_rejectlog option.
2650 76. When a system filter was run not as root (that is, when system_filter_user
2651 was set), the values of the $n variables were not being returned to the
2652 main process; thus, they were not subsequently available in the $sn
2655 77. Added +return_path_on_delivery log selector.
2657 78. A connection timeout was being treated differently from recipients deferred
2658 when testing hosts_max_try with a message that was older than the host's
2659 retry timeout. (The host should not be counted, thus allowing all hosts to
2660 be tried at least once before bouncing.) This may have been the cause of an
2661 occasionally reported bug whereby a message would remain on the queue
2662 longer than the retry timeout, but would be bounced if a delivery was
2663 forced. I say "may" because I never totally pinned down the problem;
2664 setting up timeout/retry tests is difficult. See also the next item.
2666 79. The ultimate address timeout was not being applied to errors that involved
2667 a combination of host plus message (for example, a timeout on a MAIL
2668 command). When an address resolved to a number of possible hosts, and they
2669 were not all tried for each delivery (e.g. because of hosts_max_try), a
2670 message could remain on the queue longer than the retry timeout.
2672 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2675 81. Fixed an obscure SMTP outgoing bug which required at least the following
2676 conditions: (a) there was another message waiting for the same server;
2677 (b) the server returned 5xx to all RCPT commands in the first message so
2678 that the message was not completed; (c) the server dropped the connection
2679 or gave a negative response to the RSET that Exim sends to abort the
2680 transaction. The observed case was a dropped connection after DATA that had
2681 been sent in pipelining mode. That is, the server had advertised PIPELINING
2682 but was not implementing it correctly. The effect of the bug was incorrect
2683 behaviour, such as trying another host, and this could lead to a crash.
2689 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2690 and daemon.c were passed as pointers to ints; they should have been
2691 pointers to socklen_t variables (which are typically unsigned ints).
2693 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2696 3. Fixed a really odd bug that affected only the testing scheme; patching a
2697 certain fixed string in the binary changed the value of another string that
2698 happened to be identical to the end of the original first string.
2700 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2701 name", it returns that address as the IP address. On some operating
2702 systems (e.g. Solaris), it also passes back the IP address string as the
2703 "host name". However, on others (e.g. Linux), it passes back an empty
2704 string. Exim wasn't checking for this, and was changing the host name to an
2705 empty string, assuming it had been canonicized.
2707 5. Although rare, it is permitted to have more than one PTR record for a given
2708 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2709 all the names associated with an address, because they do in Solaris.
2710 However, it seems that they do not in Linux for data that comes from the
2711 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2712 I found this out when I moved to a new Linux workstation and tried to run
2713 the Exim test suite.
2715 To get round this problem I have changed the code so that it now does its
2716 own call to the DNS to look up PTR records when searching for a host name.
2717 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2718 addresses that are only in /etc/hosts are still found.
2720 This behaviour is, however, controlled by an option called host_lookup_
2721 order, which defaults to "bydns:byaddr". If people want to use the other
2722 order, or indeed, just use one or the other means of lookup, they can
2723 specify it in this variable.
2725 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2726 some operating systems, this comes back from gethostbyaddr() as an empty
2727 string, and this is what Exim used to test for. However, it seems that in
2728 other systems, "." is yielded. Exim now tests for this case too.
2730 7. The values of check_spool_space and check_log_space are now held internally
2731 as a number of kilobytes instead of an absolute number of bytes. If a
2732 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2733 kilobyte. This means that much larger values can be stored.
2735 8. Exim monitor: an attempt to get the action menu when not actually pointing
2736 at a message produces an empty menu entitled "No message selected". This
2737 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2738 no entries in it ("Shell widget menu has zero width and/or height"). So I
2739 have added a single, blank menu entry in this case.
2741 9. Added ${quote_local_part.
2743 10. MIME decoding is now applied to the contents of Subject: header lines when
2746 11. Now that a reference to $sender_host_address automatically causes a reverse
2747 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2748 host lookup before query-style lookups in lists that might use this
2749 variable. This has therefore been abolished, and the "net-" prefix is no
2750 longer necessary for query-style lookups.
2752 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2753 have been a typo for LFLAGS, so it has been changed.
2755 13. The install script calls Exim with "-C /dev/null" in order to find the
2756 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2757 to be output. Howeve, since Exim outputs its version number before the
2758 error, it didn't break the script. It just looked ugly. I fixed this by
2759 always allowing "-C /dev/null" if the caller is root.
2761 14. Ignore overlarge ACL variable number when reading spool file - insurance
2762 against a later release with more variables having written the file.
2764 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2765 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2766 this, as well as the form without the "IPv6" on the front.
2768 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2769 OS/Makefile-Darwin file.
2771 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2772 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2774 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2775 variable. However, some OS use ints for inodes. I've added cast to long int
2776 to get rid of the compiler warning.
2778 19. I had forgotten to lock out "/../" in configuration file names when
2779 ALT_CONFIG_PREFIX was set.
2781 20. Routers used for verification do not need to specify transports. However,
2782 if such a router generated a host list, and callout was configured, Exim
2783 crashed, because it could not find a port number from the (non-existent)
2784 transport. It now assumes port 25 in this circumstance.
2786 21. Added the -t option to exigrep.
2788 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2789 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2790 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2791 them is set, LOOKUP_LSEARCH is forced.
2793 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2795 24. Added sender and host information to the "rejected by local_scan()" log
2796 line; previously there was no indication of these.
2798 25. Added .include_if_exists.
2800 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2801 Linux. It turns out that not all file systems support this. Apparently some
2802 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2803 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2804 not supported on the file descriptor, is now ignored when Exim is trying to
2805 sync a directory. This applies only to Linux.
2807 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2809 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2810 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2813 29. Added support for maildirsize files from supplied patch (modified a bit).
2815 30. The use of :fail: followed by an empty string could lead Exim to respond to
2816 sender verification failures with (e.g.):
2818 550 Verification failed for <xxx>
2819 550 Sender verify failed
2821 where the first response line was missing the '-' that indicates it is not
2822 the final line of the response.
2824 31. The loop for finding the name of the user that called Exim had a hardwired
2825 limit of 10; it now uses the value of finduser_retries, which is used for
2826 all other user lookups.
2828 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2831 33. Exim was neglecting to zero errno before one call of strtol() when
2832 expanding a string and expecting an integer value. On some systems this
2833 resulted in spurious "integer overflow" errors. Also, it was casting the
2834 result into an int without checking.
2836 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2837 did not work. The code looks as if it has *never* worked, though it appears
2838 to have been documented since at least releast 1.62. I have made it work.
2840 35. The "timeout_DNS" error in retry rules, also documented since at least
2841 1.62, also never worked. As it isn't clear exactly what this means, and
2842 clearly it isn't a major issue, I have abolished the feature by treating it
2843 as "timeout", and writing a warning to the main and panic logs.
2845 36. The display of retry rules for -brt wasn't always showing the error code
2848 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2849 timeout_connect_A, timeout_connect_MX.
2851 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2852 to the empty sender.
2854 39. The daemon was not analysing the content of -oX till after it had closed
2855 stderr and disconnected from the controlling terminal. This meant that any
2856 syntax errors were only noted on the panic log, and the return code from
2857 the command was 0. By re-arranging the code a little, I've made the
2858 decoding happen first, so such errors now appear on stderr, and the return
2859 code is 1. However, the actual setting up of the sockets still happens in
2860 the disconnected process, so errors there are still only recorded on the
2863 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2864 connections (as happens on some IP stacks) was logged at start up time as
2865 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2866 it from "IPv6 and IPv4", which means that two separate sockets are being
2869 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2870 says whether AF_INET or AF_INET6 was passed as an argument.
2872 42. Exiwhat output was messed up when time zones were included in log
2875 43. Exiwhat now gives more information about the daemon's listening ports,
2876 and whether -tls-on-connect was used.
2878 44. The "port" option of the smtp transport is now expanded.
2880 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2881 silently ignored. Now an error message is written to the main and panic
2884 46. There's a new ACL modifier called "logwrite" which writes to a log file
2885 as soon as it is encountered.
2887 47. Added $local_user_uid and $local_user_gid at routing time.
2889 48. Exim crashed when trying to verify a sender address that was being
2892 49. Exim was recognizing only a space character after ".include". It now also
2893 recognizes a tab character.
2895 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2896 extracting the relevant information from the specification. The man page no
2897 longer contains scrambled data for the -d option, and I've added a section
2898 at the front about calling Exim under different names.
2900 51. Added "extra_headers" argument to the "mail" command in filter files.
2902 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2905 53. Installed eximstats 1.29.
2907 54. Added transport_filter_timeout as a generic transport option.
2909 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2910 Cc: header lines. This was required by RFC 822, but it not required by RFC
2913 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2914 incoming messages that did not have them. Now it does so only if the
2915 message originates locally, that is, if there is no associated remote host
2916 address. When Resent- header lines are present, this applies to the Resent-
2917 lines rather than the non-Resent- lines.
2919 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2920 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2922 58. Messages for configuration errors now include the name of the main
2923 configuration file - useful now that there may be more than one file in a
2924 list (.included file names were always shown).
2926 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2927 for those rare installations that do not start the daemon as root or run it
2928 setuid root. I've cut out the call to initgroups() if the daemon is not
2931 60. The Exim user and group can now be bound into the binary as text strings
2932 that are looked up at the start of Exim's processing.
2934 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2936 62. Added $mailstore_basename variable.
2938 63. Installed patch to sieve.c from Michael Haardt.
2940 64. When Exim failed to open the panic log after failing to open the main log,
2941 the original message it was trying to log was written to stderr and debug
2942 output, but if they were not available (the usual case in production), it
2943 was lost. Now it is written to syslog before the two lines that record the
2944 failures to open the logs.
2946 65. Users' Exim filters run in subprocesses under the user's uid. It is
2947 possible for a "deliver" command or an alias in a "personal" command to
2948 provoke an address rewrite. If logging of address rewriting is configured,
2949 this fails because the process is not running as root or exim. There may be
2950 a better way of dealing with this, but for the moment (because 4.30 needs
2951 to be released), I have disabled address rewrite logging when running a
2952 filter in a non-root, non-exim process.
2958 1. The buildconfig auxiliary program wasn't quoting the value set for
2959 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2960 not defined. This bug was masked in 4.22 by the effect that was fixed in
2963 2. Some messages that were rejected after a message id was allocated were
2964 shown as "incomplete" by exigrep. It no longer does this for messages that
2965 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2967 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2968 have allow_domain_literals set, the ID did not get logged in the <= line.
2969 Domain literals are now always recognized in Message-ID: header lines.
2971 4. The first argument for a ${extract expansion item is the key name or field
2972 number. Leading and trailing spaces in this item were not being ignored,
2973 causing some misleading effects.
2975 5. When deliver_drop_privilege was set, single queue runner processes started
2976 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2977 same command in the process it spins off) were not dropping privilege.
2979 6. When the daemon running as "exim" started a queue runner, it always
2980 re-executed Exim in the spun-off process. This is a waste of effort when
2981 deliver_drop_privilege is set. The new process now just calls the
2982 queue-runner function directly.
2988 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2991 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2992 ignored. Though the use of -oP was forcing the writing of a pid file, it
2993 was always written to the default place.
2995 3. If the message "no IP address found for host xxxx" is generated during
2996 incoming verification, it is now followed by identification of the incoming
2997 connection (so you can more easily find what provoked it).
2999 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3001 5. Added some features to "harden" Exim a bit more against certain attacks:
3003 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3004 be put in Local/Makefile. This is like the never_users runtime option,
3005 but it cannot be overridden. The default setting is "root".
3007 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3008 prefix string with which any file named in a -C command line option
3011 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3012 is retained for -C and -D only if the caller of Exim is root. Without
3013 it, the exim user may also use -C and -D and retain privilege.
3015 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3016 command line option is disabled.
3018 6. Macro names set by the -D option must start with an upper case letter, just
3019 like macro names defined in the configuration file.
3021 7. Added "dereference=" facility to LDAP.
3023 8. Two instances of the typo "uknown" in the source files are fixed.
3025 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3026 the Configure-Makefile script screwed up while processing it.
3028 10. Incorporated PCRE 4.4.
3030 11. The SMTP synchronization check was not operating right at the start of an
3031 SMTP session. For example, it could not catch a HELO sent before the client
3032 waited for the greeting. There is now a check for outstanding input at the
3033 point when the greeting is written. Because of the duplex, asynchronous
3034 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3035 way, but not yet received, when the check is performed.
3037 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3038 on TCP/IP sockets, because this apparently causes some broken clients to
3041 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3042 unchanged) from the Cygwin maintainer.
3044 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3045 was supported instead of testing for mbx. Effectively a typo.
3047 15. The spa authenticator server code was not checking that the input it
3048 received was valid base64.
3050 16. The debug output line for the "set" modifier in ACLs was not showing the
3051 name of the variable that was being set.
3053 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3055 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3056 lookup on its own. Something else was needed to trigger the lookup. For
3057 example, a match in host_lookup or the need for a host name in a host list.
3058 Now, if $sender_host_name is referenced and the host name has not yet been
3059 looked up, a lookup is performed. If the lookup fails, the variable remains
3060 empty, and $host_lookup_failed is set to "1".
3062 19. Added "eqi" as a case-independent comparison operator.
3064 20. The saslauthd authentication condition could segfault if neither service
3065 nor realm was specified.
3067 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3068 error message that was logged was misleading, and incoming SMTP
3069 connections were dropped. The message is now more accurate, and temporary
3070 errors are given to SMTP connections.
3072 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3073 (except RSET) with a 421 error, until QUIT is received. However, it was
3074 failing to send a response to QUIT.
3076 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3077 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3078 ACL is now run after the helo_try_verify_hosts code.
3080 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3081 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3082 case-independent, so other case variants are also recognized). Apparently
3083 some systems use these upper case variants.
3085 25. If more than two messages were waiting for the same host, and a transport
3086 filter was specified for the transport, Exim sent two messages over the
3087 same TCP/IP connection, and then failed with "socket operation on non-
3088 socket" when it tried to send the third.
3090 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3092 27. The extern definition of crypt16() in expand.c was not being excluded when
3093 the OS had its own crypt16() function.
3095 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3096 as a preferred synonym for return_size_limit, both as an option and as an
3099 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3101 30. Changed the default configuration ACL to relax the local part checking rule
3102 for addresses that are not in any local domains. For these addresses,
3103 slashes and pipe symbols are allowed within local parts, but the sequence
3104 /../ is explicitly forbidden.
3106 31. SPA server authentication was not clearing the challenge buffer before
3109 32. log_message in a "warn" ACL statement was writing to the reject log as
3110 well as to the main log, which contradicts the documentation and doesn't
3111 seem right (because no rejection is happening). So I have stopped it.
3113 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3114 However, I am unable to do any testing of this.
3116 34. Fixed an infelicity in the appendfile transport. When checking directories
3117 for a mailbox, to see if any needed to be created, it was accidentally
3118 using path names with one or more superfluous leading slashes; tracing
3119 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3121 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3122 contents are added to the log line that is written for every discarded
3123 recipient. (Previously a log_message setting was ignored.)
3125 36. The ${quote: operator now quotes the string if it is empty.
3127 37. The install script runs exim in order to find its version number. If for
3128 some reason other than non-existence or emptiness, which it checks, it
3129 could not run './exim', it was installing it with an empty version number,
3130 i.e. as "exim-". This error state is now caught, and the installation is
3133 38. An argument was missing from the function that creates an error message
3134 when Exim fails to connect to the socket for saslauthd authentication.
3135 This could cause Exim to crash, or give a corrupted message.
3137 39. Added isip, isip4, and isip6 to ${if conditions.
3139 40. The ACL variables $acl_xx are now saved with the message, and can be
3140 accessed later in routers, transports, and filters.
3142 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3143 strings in the file are not string-expanded.
3145 42. If a MAIL command specified a SIZE value that was too large to fit into an
3146 int variable, the check against message_size_limit failed. Such values are
3147 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3148 day this will have to be increased, but I don't think I want to be around
3149 when emails are that large.
3156 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3157 iconv() is not standard in FreeBSD.
3159 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3160 IPv6 enabled. The observed symptom was a segmentation fault on return from
3161 the function os_common_find_running_interfaces() in src/os.c.
3163 3. In the check_special_case() function in daemon.c I had used "errno" as an
3164 argument name, which causes warnings on some systems. This was basically a
3165 typo, since it was named "eno" in the comments!
3167 4. The code that waits for the clock to tick (at a resolution of some fraction
3168 of a second) so as to ensure message-id uniqueness was always waiting for
3169 at least one whole tick, when it could have waited for less. [This is
3170 almost certainly not relevant at current processor speeds, where it is
3171 unlikely to ever wait at all. But we try to future-proof.]
3173 5. The function that sleeps for a time interval that includes fractions of a
3174 second contained a race. It did not block SIGALRM between setting the
3175 timer, and suspending (a couple of lines later). If the interval was short
3176 and the sigsuspend() was delayed until after it had expired, the suspension
3177 never ended. On busy systems this could lead to processes getting stuck for
3180 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3181 process, before it forks any delivery processes. The open lookup caching
3182 mechanism meant that the open file or database connection was passed into
3183 the delivery process. The problem was that delivery processes always tidy
3184 up cached lookup data. This could cause a problem for the next delivery
3185 process started by the queue runner, because the external queue runner
3186 process does not know about the closure. So the next delivery process
3187 still has data in the lookup cache. In the case of a file lookup, there was
3188 no problem because closing a file descriptor in a subprocess doesn't affect
3189 the parent. However, if the lookup was caching a connection to a database,
3190 the connection was closed, and the second delivery process was likely to
3191 see errors such as "PGSQL: query failed: server closed the connection
3192 unexpectedly". The problem has been fixed by closing all cached lookups
3193 in a queue runner before running a delivery process.
3195 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3196 seem to have the "const" qualifier which it has on other OS. I've
3199 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3200 *of the same type* (client or server) with the same public name that an
3201 error should be diagnosed.
3203 9. When Exim looked up a host name for an IP address, but failed to find the
3204 original IP address when looking up the host name (a safety check), it
3205 output the message "<ip address> does not match any IP for NULL", which was
3206 confusing, to say the least. The bug was that the host name should have
3207 appeared instead of "NULL".
3209 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3210 user that is built into the binary, and the -C or -D options is used, root
3211 privilege is dropped before the configuration file is read. In addition,
3212 logging is switched to stderr instead of the normal log files. If the
3213 configuration then re-defines the Exim user, the unprivileged environment
3214 is probably not what is expected, so Exim logs a panic warning message (but
3217 However, if deliver_drop_privilege is set, the unprivileged state may well
3218 be exactly what is intended, so the warning has been cut out in that case,
3219 and Exim is allowed to try to write to its normal log files.
3225 1. smtp_return_error_details was not giving details for temporary sender
3226 or receiver verification errors.
3228 2. Diagnose a configuration error if two authenticators have the same public
3231 3. Exim used not to create the message log file for a message until the first
3232 delivery attempt. This could be confusing when incoming messages were held
3233 for policy or load reasons. The message log file is now created at the time
3234 the message is received, and an initial "Received" line is written to it.
3236 4. The automatically generated man page for command line options had a minor
3237 bug that caused no ill effects; however, a more serious problem was that
3238 the procedure for building the man page automatically didn't always
3239 operate. Consequently, release 4.20 contains an out-of-date version. This
3240 shouldn't happen again.
3242 5. When building Exim with embedded Perl support, the script that builds the
3243 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3244 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3246 6. The freeze_tell option was not being used for messages that were frozen on
3247 arrival, either by an ACL or by local_scan().
3249 7. Added the smtp_incomplete_transaction log selector.
3251 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3252 was accepting AUTH without a new EHLO.
3254 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3255 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3258 10. Logging of TCP/IP connections (when configured) now happens in the main
3259 daemon process instead of the child process, so that the TCP/IP connection
3260 count is more accurate (but it can never be perfect).
3262 11. The use of "drop" in a nested ACL was not being handled correctly in the
3263 outer ACL. Now, if condition failure induced by the nested "drop" causes
3264 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3265 or "require"), the connection is dropped.
3267 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3268 that yield "discard" can now be used with an "accept" or a "discard" verb,
3269 but an error is generated for any others (because I can't see a useful way
3270 to define what should happen).
3272 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3273 are now processed in the same way as lines in the Exim configuration file.
3274 In particular, continuation lines are supported.
3276 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3278 15. Added -ti meaning -t -i.
3280 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3281 domains, and warn by logging if others are found.
3283 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3284 returned from GIFCONF when Exim is trying to find the list of interfaces on
3285 a host. The code in os.c has been modified to copy each ifreq to an aligned
3286 structure in all cases.
3288 Also, in some cases, the returned ifreq's were being copied to a 'struct
3289 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3290 means the last couple of bytes of an IPv6 address could be chopped if the
3291 ifreq contained only a normal sockaddr (14 bytes storage).
3293 18. Named domain lists were not supported in the hosts_treat_as_local option.
3294 An entry such as +xxxx was not recognized, and was treated as a literal
3297 19. Ensure that header lines added by a DATA ACL are included in the reject log
3298 if the ACL subsequently rejects the message.
3300 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3301 MD5 (which is deprecated).
3303 21. When testing a filter file using -bf, Exim was writing a message when it
3304 took the sender from a "From " line in the message, but it was not doing so
3305 when it took $return_path from a Return-Path: header line. It now does.
3307 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3308 with a valid header line field name (a series of printing characters
3309 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3311 23. Changed "disc" in the source to "disk" to conform to the documentation and
3312 the book and for uniformity.
3314 24. Ignore Sendmail's -Ooption=value command line item.
3316 25. When execve() failed while trying to run a command in a pipe transport,
3317 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3318 could be confused with a return value of 69 from the command itself. This
3319 has been changed to 127, the value the shell returns if it is asked to run
3320 a non-existent command. The wording for the related log line suggests a
3321 non-existent command as the problem.
3323 26. If received_header_text expands to an empty string, do not add a Received:
3324 header line to the message. (Well, it adds a token one on the spool, but
3325 marks it "old" so that it doesn't get used or transmitted.)
3327 27. Installed eximstats 1.28 (addition of -nt option).
3329 28. There was no check for failure on the call to getsockname() in the daemon
3330 code. This can fail if there is a shortage of resources on the system, with
3331 ENOMEM, for example. A temporary error is now given on failure.
3333 29. Contrary to the C standard, it seems that in some environments, the
3334 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3335 program. Exim now does this explicitly; it affects the formatting of
3336 timestamps using strftime().
3338 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3339 complaints. I've now initialized all the variables, to avoid this.
3341 32. Header lines added by a system filter were not being "seen" during
3342 transport-time rewrites.
3344 33. The info_callback() function passed to OpenSSL is set up with type void
3345 (*)(SSL *, int, int), as described somewhere. However, when calling the
3346 function (actually a macro) that sets it up, the type void(*)() is
3347 expected. I've put in a cast to prevent warnings from picky compilers.
3349 34. If a DNS black list lookup found a CNAME record, but there were no A
3350 records associated with the domain it pointed at, Exim crashed.
3352 35. If a DNS black list lookup returned more than one A record, Exim ignored
3353 all but the first. It now scans all returned addresses if a particular IP
3354 value is being sought. In this situation, the contents of the
3355 $dnslist_value variable are a list of all the addresses, separated by a
3358 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3359 to accept a host name and all its aliases if the forward lookup for any of
3360 them yielded the IP address of the incoming connection. Now it accepts only
3361 those names whose forward lookup yields the correct IP address. Any other
3362 names are discarded. This closes a loophole whereby a rogue DNS
3363 administrator could create reverse DNS records to break through a
3364 wildcarded host restriction in an ACL.
3366 37. If a user filter or a system filter that ran in a subprocess used any of
3367 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3368 the wrong values were passed to the pipe command ($thisaddress had the
3369 value of $0, $0 had the value of $1, etc). This bug was introduced by
3370 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3372 38. Improved the line breaking for long SMTP error messages from ACLs.
3373 Previously, if there was no break point between 40 and 75 characters, Exim
3374 left the rest of the message alone. Two changes have been made: (a) I've
3375 reduced the minimum length to 35 characters; (b) if it can't find a break
3376 point between 35 and 75 characters, it looks ahead and uses the first one
3377 that it finds. This may give the occasional overlong line, but at least the
3378 remaining text gets split now.
3380 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3381 file descriptors might be low, and that setting 1000 would always raise it.
3382 It turns out that in some environments, the limit is already over 1000 and
3383 that lowering it causes trouble. So now Exim takes care not to decrease it.
3385 40. When delivering a message, the value of $return_path is set to $sender_
3386 address at the start of routing (routers may change the value). By an
3387 oversight, this default was not being set up when an address was tested by
3388 -bt or -bv, which affected the outcome if any router or filter referred to
3391 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3392 log only when "message" or "log_message" are set. However, if one of the
3393 conditions was an address verification, or a call to a nested ACL, the
3394 messages generated by the underlying test were being passed through. This
3395 no longer happens. The underlying message is available in $acl_verify_
3396 message for both "message" and "log_message" expansions, so it can be
3397 passed through if needed.
3399 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3400 new expansion $bh_ to give the encoded byte string without charset
3401 translation. Translation happens only if iconv() is available; HAVE_ICONV
3402 indicates this at build time. HEADERS_CHARSET gives the charset to
3403 translate to; headers_charset can change it in the configuration, and
3404 "headers charset" can change it in an individual filter file.
3406 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3407 that creates RFC 2047 encoded "words" labels them as that charset instead
3408 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3409 expansion operator; (ii) when Exim creates a From: line for a local
3410 message; (iii) when a header line is rewritten to include a "phrase" part.
3412 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3413 buggy, causing it to skip the first lines of messages whose message ID
3414 ended in 'D'. This would not have bitten before Exim release 4.14, because
3415 message IDs were unlikely to end in 'D' before then. The effect was to have
3416 incorrect size information for certain domains.
3418 45. #include "config.h" was missing at the start of the crypt16.c module. This
3419 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3422 46. If there was a timeout during a "random" callout check, Exim treated it as
3423 a failure of the random address, and carried on sending RSET and the real
3424 address. If the delay was just some slowness somewhere, the response to the
3425 original RCPT would be taken as a response to RSET and so on, causing
3426 mayhem of various kinds.
3428 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3429 when I implemented it. It didn't allow for the fact that some option values
3430 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3431 the right test for positive values.
3433 48. Domain names in DNS records are case-independent. Exim always looks them up
3434 in lower case. Some resolvers return domain names in exactly the case they
3435 appear in the zone file, that is, they may contain uppercase letters. Not
3436 all resolvers do this - some return always lower case. Exim was treating a
3437 change of case by a resolver as a change of domain, similar to a widening
3438 of a domain abbreviation. This triggered its re-routing code and so it was
3439 trying to route what was effectively the same domain again. This normally
3440 caused routing to fail (because the router wouldn't handle the domain
3441 twice). Now Exim checks for this case specially, and just changes the
3442 casing of the domain that it ultimately uses when it transmits the message
3445 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3448 50. If a filter generated a file delivery with a non-absolute name (possible if
3449 no home directory exists for the router), the forbid_file option was not
3452 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3453 the existing equality matching.
3455 52. Exim was using ints instead of ino_t variables in some places where it was
3456 dealing with inode numbers.
3458 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3459 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3460 called TMPDIR, and if it finds it is different, it changes its value.
3462 54. The smtp_printf() function is now made available to local_scan() so
3463 additional output lines can be written before returning. There is also an
3464 smtp_fflush() function to enable the detection of a dropped connection.
3465 The variables smtp_input and smtp_batched_input are exported to
3468 55. Changed the default runtime configuration: the message "Unknown user"
3469 has been removed from the ACL, and instead placed on the localuser router,
3470 using the cannot_route_message feature. This means that any verification
3471 failures that generate their own messages won't get overridden. Similarly,
3472 the "Unrouteable address" message that was in the ACL for unverifiable
3473 relay addresses has also been removed.
3475 56. Added hosts_avoid_esmtp to the smtp transport.
3477 57. The exicyclog script was not checking for the esoteric option
3478 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3479 will work only if exicyclog is run under the appropriate euid.
3481 58. Following a discussion on the list, the rules by which Exim recognises line
3482 endings on incoming messages have been changed. The -dropcr and drop_cr
3483 options are now no-ops, retained only for backwards compatibility. The
3484 following line terminators are recognized: LF CRLF CR. However, special
3485 processing applies to CR:
3487 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3488 nor a local message in the state where . is a terminator.
3490 (ii) If a bare CR is encountered in a header line, an extra space is added
3491 after the line terminator so as not to end the header. The reasoning
3492 behind this is that bare CRs in header lines are most likely either
3493 to be mistakes, or people trying to play silly games.
3495 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3496 was being incorrectly given as 18 bytes larger than it should have been.
3497 This is a VOB (very old bug).
3499 60. This may never have affected anything current, but just in case it has:
3500 When the local host is found other than at the start of a list of hosts,
3501 the local host, those with the same MX, and any that follow, are discarded.
3502 When the list in question was part of a longer list of hosts, the following
3503 hosts (not currently being processed) were also being discarded. This no
3504 longer happens. I'm not sure if this situation could ever has previously
3507 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3510 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3511 header. This is something that is recommended in a new Internet Draft, and
3512 is something that is documented as being done by Sendmail. There are two
3513 possible values. For messages generated by the autoreply transport, Exim
3516 Auto-Submitted: auto-replied
3518 whereas for all other generated messages (e.g. bounces) it adds
3520 Auto-Submitted: auto-generated
3522 63. The "personal" condition in filters now includes a test for the
3523 Auto-Submitted: header. If it contains the string "auto-" the message it
3524 not considered personal.
3526 64. Added rcpt_include_affixes as a generic transport option.
3528 65. Added queue_only_override (default true).
3530 66. Added the syslog_duplication option.
3532 67. If what should have been the first header line of a message consisted of
3533 a space followed by a colon, Exim was mis-interpreting it as a header line.
3534 It isn't of course - it is syntactically invalid and should therefore be
3535 treated as the start of the message body. The misbehaviour could have
3536 caused a number of strange effects, including loss of data in subsequent
3537 header lines, and spool format errors.
3539 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3540 client host had authenticated. This control can now be exercised by an ACL
3541 for more flexibility.
3543 69. By default, callouts do not happen when testing with -bh. There is now a
3544 variant, -bhc, which does actually run the callout code, including
3545 consulting and updating the callout cache.
3547 70. Added support for saslauthd authentication, courtesy of Alexander
3550 71. If statvfs() failed on the spool or log directories while checking their
3551 size for availability, Exim confusingly gave the error "space shortage".
3552 Furthermore, in debugging mode it crashed with a floating point exception.
3553 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3554 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3555 serious problem, Exim now writes to the main and panic logs when this
3556 happens, with details of the failure. It then refuses to accept the
3557 incoming message, giving the message "spool directory problem" or "log
3558 directory problem" with a 421 code for SMTP messages.
3560 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3561 0, 1, and 2 exist, because some OS complain for execs without them (see
3562 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3563 descriptors. However, the code omitted to check that the open succeeded,
3564 causing mysterious errors if for some reason the permissions on /dev/null
3565 got screwed. Now Exim writes a message to the main and panic logs, and
3566 bombs out if it can't open /dev/null.
3568 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3569 interact so that it is all more flexible. It is supposed to remain
3570 backwards compatible. Also added extra_local_interfaces.
3572 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3573 to bomb out with an assertion failure - to the client this appears as a
3574 connection drop. This problem occurs in the part of the code that was taken
3575 from the Samba project. Fortunately, the assertion is in a very simple
3576 function, so I have fixed this by reproducing the function inline in the
3577 one place where it is called, and arranging for authentication to fail
3578 instead of killing the process with assert().
3580 75. The SPA client code was not working when the server requested OEM rather
3581 than Unicode encoding.
3583 76. Added code to make require_files with a specific uid setting more usable in
3584 the case where statting the file as root fails - usually a non-root-mounted
3585 NFS file system. When this happens and the failure is EACCES, Exim now
3586 forks a subprocess and does the per-uid checking as the relevant uid.
3588 77. Added process_log_path.
3590 78. If log_file_path was not explicitly set, a setting of check_log_space or
3591 check_log_inodes was ignored.
3593 79. If a space check for the spool or log partitions fails, the incident is now
3594 logged. Of course, in the latter case the data may get lost...
3596 80. Added the %p formatting code to string_format() so that it can be used to
3597 print addresses in debug_print(). Adjusted all the address printing in the
3598 debugging in store.c to use %p rather than %d.
3600 81. There was a concern that a line of code in smtp_in.c could overflow a
3601 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3602 initially expressed, the concern was not well-founded, because trailing
3603 spaces are removed early. However, if the trailing spaces were followed by
3604 a NULL, they did not get removed, so the overflow was possible. Two fixes
3607 (a) I re-wrote the offending code in a cleaner fashion.
3608 (b) If an incoming SMTP command contains a NULL character, it is rejected
3611 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3612 runs initgroups(), so that if the Exim user is in any additional groups,
3613 they will be used during message reception.
3619 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob