1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.274 2005/12/12 15:58:53 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.
42 PH/01 Two changes to the default runtime configuration:
44 (1) Move the checks for relay_from_hosts and authenticated clients from
45 after to before the (commented out) DNS black list checks.
47 (2) Add control=submission to the relay_from_hosts and authenticated
48 clients checks, on the grounds that messages accepted by these
49 statements are most likely to be submissions.
51 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
53 (1) Generate an error if the third argument for the ${prvs expansion is
56 (2) Treat a missing third argument of ${prvscheck as if it were an empty
59 (3) Reset the variables that are obtained from the first argument of
60 ${prvscheck and used in the second argument before leaving the code,
61 because their memory is reclaimed, so using them afterwards may do
64 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
65 one (it's much easier than Tom thought :-).
67 (5) Because of (4), we can now allow for the use of $prvscheck_result
68 inside the third argument.
70 PH/03 For some reason, the default setting of PATH when running a command from
71 a pipe transport was just "/usr/bin". I have changed it to
74 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
75 anything to be listed in the output from -bV.
77 PH/05 When a filter generated an autoreply, the entire To: header line was
78 quoted in the delivery log line, like this:
80 => >A.N.Other <ano@some.domain> <original@ddress> ...
82 This has been changed so that it extracts the operative address. There
83 may be more than one such address. If so, they are comma-separated, like
86 => >ano@some.domain,ona@other.domain <original@ddress> ...
88 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
89 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
90 not being looked up in the rDNS to get a host name, Exim was showing the
91 IP address twice in Received: lines, even though the IP addresses were
92 identical. For example:
94 Received: from [1.2.3.4] (helo=[1.2.3.4])
96 However, if the real host name was known, it was omitting the HELO data
97 if it matched the actual IP address. This has been tidied up so that it
98 doesn't show the same IP address twice.
100 PH/07 When both +timestamp and +memory debugging was on, the value given by
101 $tod_xxx expansions could be wrong, because the tod_stamp() function was
102 called by the debug printing, thereby overwriting the timestamp buffer.
103 Debugging no longer uses the tod_stamp() function when +timestamp is set.
105 PH/08 When the original message was included in an autoreply transport, it
106 always said "this is a copy of the message, including all the headers",
107 even if body_only or headers_only was set. It now gives an appropriate
110 PH/09 Applied a patch from the Sieve maintainer which:
112 o fixes some comments
113 o adds the (disabled) notify extension core
114 o adds some debug output for the result of if/elsif tests
115 o points to the current vacation draft in the documentation
116 and documents the missing references header update
120 o fixes a bug in processing the envelope test (when testing
121 multiple envelope elements, the last element determinted the
124 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
125 Electronic Mail") by including:
127 Auto-submitted: auto-generated
129 in the messages that it generates (bounce messages and others, such as
130 warnings). In the case of bounce messages for non-SMTP mesages, there was
131 also a typo: it was using "Auto_submitted" (underscore instead of
132 hyphen). Since every message generated by Exim is necessarily in response
133 to another message, thes have all been changed to:
135 Auto-Submitted: auto-replied
137 in accordance with these statements in the RFC:
139 The auto-replied keyword:
141 - SHOULD be used on messages sent in direct response to another
142 message by an automatic process,
144 - MUST NOT be used on manually-generated messages,
146 - MAY be used on Delivery Status Notifications (DSNs) and Message
147 Disposition Notifications (MDNs),
149 - MUST NOT be used on messages generated by automatic or periodic
150 processes, except for messages which are automatic responses to
153 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
154 to the default Received: header definition.
156 PH/12 Added log selector acl_warn_skipped (default on).
158 PH/13 After a successful wildlsearch lookup, discard the values of numeric
159 variables because (a) they are in the wrong storage pool and (b) even if
160 they were copied, it wouldn't work properly because of the caching.
162 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
163 checking when decoding. Apparently there are clients that generate
164 overlong encoded strings. Why am I not surprised?
166 PH/15 If the first argument of "${if match_address" was not empty, but did not
167 contain an "@" character, Exim crashed. Now it writes a panic log message
168 and treats the condition as false.
170 PH/16 In autoreply, treat an empty string for "once" the same as unset.
172 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
173 extension "envelope-auth". The code is finished and in agreement with
174 other implementations, but there is no documentation so far and in fact,
175 nobody wrote the draft yet. This extension is currently #undef'ed, thus
176 not changing the active code.
178 Print executed "if" and "elsif" statements when debugging is used. This
179 helps a great deal to understand what a filter does.
181 Document more things not specified clearly in RFC3028. I had all this
182 sorted out, when out of a sudden new issues came to my mind. Oops."
184 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
187 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
188 canonical form (as documented). However, after a host name lookup from
189 the IP address, check_host() was doing a simple string comparison with
190 addresses acquired from the DNS when checking that the found name did
191 have the original IP as one of its addresses. Since any found IPv6
192 addresses are likely to be in abbreviated form, the comparison could
193 fail. Luckily, there already exists a function for doing the comparison
194 by converting both addresses to binary, so now that is used instead of
197 PH/20 There was another similar case to PH/19, when a complete host name was
198 given in a host list; looking up its IP address could give an abbreviated
199 form, whereas the current host's name might or might not be abbreviated.
200 The same fix has been applied.
206 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
207 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
210 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
211 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
213 PH/03 Typo: missing ".o" in src/pcre/Makefile.
215 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
216 header line, restrict the check to what is listed in RFCs 2369 and 2929.
217 Also, for "Auto-Submitted", treat anything other than "no" as
218 non-personal, in accordance with RFC 3834. (Previously it treated
219 anything starting "auto-" as non-personal.)
221 TF/01 The control=submission/name=... option had a problem with syntax
222 errors if the name included a slash character. The /name= option
223 now slurps the rest of the string, so it can include any characters
224 but it must come last in the list of options (after /sender_retain
227 PH/05 Some modifications to the interface to the fake nameserver for the new
230 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
231 into the default Received: header string.
238 TK/01 Added the "success_on_redirect" address verification option. See
239 NewStuff for rationale and an example.
241 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
243 PH/02 Patch to exigrep to allow it to work on syslog lines.
245 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
246 fread() to skip over the body file's header line, because in Cygwin the
247 header line is locked and is inaccessible.
249 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
250 co-exist for some time) to make it clear that it is the Exim ID that is
251 referenced, not the Message-ID: header line.
253 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
254 string_format() function, because snprintf() does not exist on all
257 PH/06 The use of forbid_filter_existstest now also locks out the use of the
258 ${stat: expansion item.
260 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
261 protocol synchronization error", to keep the pedants happy.
263 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
264 well as for IRIX systems, when gcc is being used. See the host.c source
267 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
269 PH/10 Named domain lists were not working if used in a queue_smtp_domains
272 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
273 transport and to the smtp transport in LMTP mode.
275 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
277 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
278 run a filter in a subprocess. This could lead to confusion in subsequent
279 lookups in the parent process. There should also be a search_tidyup() at
280 the end of the subprocess.
282 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
283 only if the host matched helo_try_verify_hosts, which caused the
284 verification to occur when the EHLO/HELO command was issued. The ACL just
285 tested the remembered result. Now, if a previous verification attempt has
286 not happened, "verify = helo" does it there and then.
288 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
290 TK/03 Fix log output including CR from clamd.
292 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
293 exist provoked a memory error which could cause a segfault.
295 PH/15 Installed PCRE 6.2
297 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
299 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
300 of the problem. Specifically, suggested +O2 rather than +O1 for the
303 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
305 PH/20 If a delivery was routed to a non-standard port by means of an SRV
306 record, the port was not correctly logged when the outgoing_port log
307 selector was set (it logged the transort's default port).
309 PH/21 Added support for host-specific ports to manualroute, queryprogram,
310 fallback_hosts, and "hosts" in the smtp transport.
312 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
313 host errors such as "Connection refused".
315 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
316 authentication with radiusclient 0.4.9:
318 - Error returned from rc_read_config was caught wrongly
319 - Username/password not passed on to radius server due to wrong length.
321 The presumption is that some radiusclient API changes for 4.51/PH/17
322 were not taken care of correctly. The code is still untested by me (my
323 Linux distribution still has 0.3.2 of radiusclient), but it was
324 contributed by a Radius user.
326 PH/24 When doing a callout, the value of $domain wasn't set correctly when
327 expanding the "port" option of the smtp transport.
329 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
330 while reading a MIME header. Thanks to Tom Hughes for a patch.
332 PH/24 Include config.h inside local_scan.h so that configuration settings are
335 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
336 that in an ACL for RCPT (for example), you can examine exactly what was
339 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
340 commands, but it was not correctly comparing the address with the actual
341 client host address. Thus, it would show the EHLO address in Received:
342 header lines when this was not necessary.
344 PH/27 Added the % operator to ${eval:}.
346 PH/28 Exim tries to create and chdir to its spool directory when it starts;
347 it should be ignoring failures (because with -C, for example, it has lost
348 privilege). It wasn't ignoring creation failures other than "already
351 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
354 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
355 because an input line was too long, either on its own, or by virtue of
356 too many continuations, the temporary file was not being removed, and the
357 return code was incorrect.
359 PH/31 Missing "BOOL" in function definition in filtertest.c.
361 PH/32 Applied Sieve patches from the maintainer.
363 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
365 PH/33 Added "verify = not_blind".
367 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
368 Local/Makefile (with some defaults set). These are used in built scripts
369 such as exicyclog, but they have never been used in the exim_install
370 script (though there are many overriding facilities there). I have
371 arranged that the exim_install script now takes note of these two
374 PH/35 Installed configuration files for Dragonfly.
376 PH/36 When a locally submitted message by a trusted user did not contain a
377 From: header, and the sender address was obtained from -f or from an SMTP
378 MAIL command, and the trusted user did not use -F to supply a sender
379 name, $originator_name was incorrectly used when constructing a From:
380 header. Furthermore, $originator_name was used for submission mode
381 messages from external hosts without From: headers in a similar way,
382 which is clearly wrong.
384 PH/37 Added control=suppress_local_fixups.
386 PH/38 When log_selector = +received_sender was set, and the addition of the
387 sender made the log line's construction buffer exactly full, or one byte
388 less than full, an overflow happened when the terminating "\n" was
391 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
392 when the result of a list match is failure because a DNS lookup failed.
394 PH/40 RM_COMMAND is now used in the building process.
396 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
397 the "build-* directories that it finds.
399 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
400 address in a domain literal was a prefix of an interface address.
402 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
403 when verifying a sender address, unless rewrite_headers is false.
405 PH/44 Wrote a long comment about why errors_to addresses are verified as
406 recipients, not senders.
408 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
409 the ratelimit ACL was added.
411 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
413 PH/46 Added extra information about PostgreSQL errors to the error string.
415 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
416 avoiding the need to install special zones in a real server. This is
417 backwards compatible; if it can't find the fake resolver, it drops back.
418 Thus, both old and new test suites can be run.
420 TF/02 Added util/ratelimit.pl
422 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
423 clock is set back in time.
425 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
426 Candler <B.Candler@pobox.com>.
428 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
429 OK for addresses that are the result of redirections.
431 PH/48 A number of further additions for the benefit of the new test suite,
432 including a fake gethostbyname() that interfaces to the fake DNS resolver
435 TF/06 The fix for widen_domains has also been applied to qualify_single and
436 search_parents which are the other dnslookup options that can cause
439 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
442 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
444 TF/07 Exim produced the error message "an SRV record indicated no SMTP
445 service" if it encountered an MX record with an empty target hostname.
446 The message is now "an MX or SRV record indicated no SMTP service".
448 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
449 if the DNS of the sending site is misconfigured. This is quite a
450 common situation. This change restores the behaviour of treating a
451 helo verification defer as a failure.
453 PH/51 If self=fail was set on a router, the bounce message did not include the
454 actual error message.
460 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
462 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
463 command itself ended in error, the underlying message about the transport
464 filter timeout was being overwritten with the pipe command error. Now the
465 underlying error message should be appended to the second error message.
467 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
469 PH/02 Reduce the amount of output that "make" produces by default. Full output
470 can still be requested.
472 PH/03 The warning log line about a condition test deferring for a "warn" verb
473 was being output only once per connection, rather than after each
474 occurrence (because it was using the same function as for successful
475 "warn" verbs). This seems wrong, so I have changed it.
477 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
478 it should not have, which might have caused a crash in the right
479 circumstances, but probably never did.
481 PH/04 Installed a modified version of Tony Finch's patch to make submission
482 mode fix the return path as well as the Sender: header line, and to
483 add a /name= option so that you can make the user's friendly name appear
486 TF/03 Added the control = fakedefer ACL modifier.
488 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
489 Mark Lowes for thorough testing.
491 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
493 TK/03 Merged latest SRS patch from Miles Wilton.
495 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
496 with the definition in sysexits.h (which is #included earlier).
497 Fortunately, Exim does not actually use EX_OK. The code used to try to
498 preserve the sysexits.h value, by assumimg that macro definitions were
499 scanned for macro replacements. I have been disabused of this notion,
500 so now the code just undefines EX_OK before #including unistd.h.
502 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
503 in the smtp transport. When a block could not be written in a single
504 write() function, the timeout was being re-applied to each part-write.
505 This seems wrong - if the receiver was accepting one byte at a time it
506 would take for ever. The timeout is now adjusted when this happens. It
507 doesn't have to be particularly precise.
509 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
510 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
512 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
513 without a domain if the check to <postmaster@domain> fails.
515 SC/01 Eximstats: added -xls and the ability to specify output files
516 (patch written by Frank Heydlauf).
518 SC/02 Eximstats: use FileHandles for outputing results.
520 SC/03 Eximstats: allow any combination of xls, txt, and html output.
522 SC/04 Eximstats: fixed display of large numbers with -nvr option
524 SC/05 Eximstats: fixed merging of reports with empty tables.
526 SC/06 Eximstats: added the -include_original_destination flag
528 SC/07 Eximstats: removed tabs and trailing whitespace.
530 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
532 TK/06 MBOX spool code: Add real "From " MBOX separator line
533 so the .eml file is really in mbox format (even though
534 most programs do not really care). Patch from Alex Miller.
536 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
537 The latter is generated from $received_to and is only set if the
538 message has one envelope recipient. SA can use these headers,
539 obviously out-of-the-box. Patch from Alex Miller.
541 PH/08 The ${def test on a variable was returning false if the variable's
542 value was "0", contrary to what the specification has always said!
543 The result should be true unless the variable is empty.
545 PH/09 The syntax error of a character other than { following "${if
546 def:variable_name" (after optional whitespace) was not being diagnosed.
547 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
548 accidental colon was present, for example, could give incorrect results.
550 PH/10 Tidied the code in a number of places where the st_size field of a stat()
551 result is used (not including appendfile, where other changes are about
554 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
555 This involved changing a lot of size variables from int to off_t. It
556 should work with maildirs and everything.
558 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
559 spamd dying while we are connected to it.
561 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
564 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
565 function for BDB 4.3.
567 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
569 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
570 the sources for PCRE have been re-arranged and more files are now
573 PH/15 The code I had for printing potentially long long variables in PH/11
574 above was not the best (it lost precision). The length of off_t variables
575 is now inspected at build time, and an appropriate printing format (%ld
576 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
577 to be "long long int" or "long int". This is needed for the internal
578 formatting function string_vformat().
580 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
581 the configuration file to be ":syslog", then the script "guesses" where
582 the logs files are, rather than using the compiled in default. In our
583 case the guess is not the same as the compiled default, so the script
584 suddenly stopped working when I started to use syslog. The patch checks
585 to see if log_file_path is "". If so, it attempts to read it from exim
586 with no configuration file to get the compiled in version, before it
587 falls back to the previous guessing code."
589 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
590 implementing BATV in an Exim configuration. See NewStuff for the gory
593 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
594 Makefile that are specific to HP-UX.
596 PH/18 If the "use_postmaster" option was set for a recipient callout together
597 with the "random" option, the postmaster address was used as the MAIL
598 FROM address for the random test, but not for the subsequent recipient
599 test. It is now used for both.
601 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
602 patch removes a few documentation additions to RFC 3028, because the
603 latest draft now contains them. It adds the new en;ascii-case comparator
604 and a new error check for 8bit text in MIME parts. Comparator and
605 require names are now matched exactly. I enabled the subaddress
606 extension, but it is not well tested yet (read: it works for me)."
608 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
609 rework some of the code of TK/09 above to avoid the hardwired use of
610 "%lld" and "long long". Replaced the call to snprintf() with a call to
613 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
614 records point to non-existent hosts", "retry timeout exceeded", and
615 "retry time not reached for any host after a long failure period".
617 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
618 experimental DomainKeys support:
620 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
621 (2) On an error such as an illegally used "control", the wrong name for
622 the control was given.
624 These problems did NOT occur unless DomainKeys support was compiled.
626 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
628 PH/24 Added ${if match_ip condition.
630 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
631 databases so that it will be absolutely obvious if a crash occurs in the
632 DB library. This is a regular occurrence (often caused by mis-matched
635 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
636 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
637 picked up on a user's system that detects such things. There doesn't seem
638 to be a gcc warning option for this - only an attribute that has to be
639 put on the function's prototype. It seems that in Fedora Core 4 they have
640 set this on a number of new functions. No doubt there will be more in due
643 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
644 specify a transport. However, if an address that was verified by such a
645 router was the subject of a callout, Exim crashed because it tried to
646 read the rcpt_include_affixes from the non-existent transport. Now it
647 just assumes that the setting of that option is false. This bug was
648 introduced by 4.51/PH/31.
650 PH/28 Changed -d+all to exclude +memory, because that information is very
651 rarely of interest, but it makes the output a lot bigger. People tend to
652 do -d+all out of habit.
654 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
655 code in os-type was giving problems when libc.so lives in lib64, like on
658 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
659 aren't the modern standard, and it seems that some systems' include files
660 don't always have them. Exim was already checking for some of the newer
661 ones like T_AAAA, and defining it itself. I've added checks for all the
662 record types that Exim uses.
664 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
665 not automatically generating a new one, as it is supposed to. This
666 prevented TLS from working. If the file did exist, but contained invalid
667 data, a new version was generated, as expected. It was only the case of a
668 non-existent file that was broken.
670 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
671 with a change in libdomainkeys > 0.64.
673 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
674 from DNS. If the selector record carries the flag, it now has
675 precedence over the domain-wide flag.
677 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
679 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
680 the use of a port name that isn't defined in /etc/services) occurred, the
681 message was deferred as in a normal delivery, and thus remained on the
682 spool, instead of being failed because of the mua_wrapper setting. This
683 is now fixed, and I tidied up some of the mua_wrapper messages at the
686 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
687 the messages in a hash of arrays rather than using individual hashes.
688 This is a bit cleaner and results in dramatic memory savings, albeit
689 at a slight CPU cost.
691 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
692 as requested by Marc Sherman.
694 SC/10 Eximstats: added histograms for user specified patterns as requested
697 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
699 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
700 fopen() in the content-scanning modules that did not already have it.
706 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
707 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
709 TK/02 Fix ACL "control" statement not being available in MIME ACL.
711 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
713 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
714 to test Sieve filters that use "vacation".
716 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
717 that changes the way the GnuTLS parameters are stored in the cache file.
718 The new format can be generated externally. For backward compatibility,
719 if the data in the cache doesn't make sense, Exim assumes it has read an
720 old-format file, and it generates new data and writes a new file. This
721 means that you can't go back to an older release without removing the
724 PH/03 A redirect router that has both "unseen" and "one_time" set does not
725 work if there are any delivery delays because "one_time" forces the
726 parent to be marked "delivered", so its unseen clone is never tried
727 again. For this reason, Exim now forbids the simultaneous setting of
730 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
731 redirected to themselves ("homonym" addresses). Read the long ChangeLog
732 entry if you want to know the details. The fix, however, neglected to
733 consider the case when local delivery batching is involved. The test for
734 "previously delivered" was not happening when checking to see if an
735 address could be batched with a previous (undelivered) one; under
736 certain circumstances this could lead to multiple deliveries to the same
739 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
740 in its include files, and this causes problems building Exim.
742 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
743 header_syntax) but Exim was just ignoring anything given after a slash.
744 In particular, this caused confusion with an attempt to use "verify =
745 reverse_host_lookup/defer_ok". An error is now given when options are
746 supplied for verify items that do not have them. (Maybe reverse_host_
747 lookup should have a defer_ok option, but that's a different point.)
749 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
750 defined by RFC 821) to 2048, because there were problems with some AUTH
751 commands, and RFC 1869 says the size should be increased for extended
752 SMTP commands that take arguments.
754 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
757 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
758 "unknown" error; now it says that the functionality isn't in the binary.
760 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
761 an address' error message when a string expansion fails (syntax or
762 whatever). Otherwise the password may appear in the log. Following change
763 PH/42 below, there is no longer a chance of it appearing in a bounce
766 PH/11 Installed exipick version 20050225.0 from John Jetmore.
768 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
769 of its addresses was ever tried. (Bugzilla bug #2.)
771 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
772 the result incorrectly in the debug output. (It correctly added a newline
773 to what was transported.)
775 TF/01 Added $received_time.
777 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
778 commented out examples of how to interface to a virus scanner and to
779 SpamAssassin. Also added commented examples of av_scanner and
780 spamd_address settings.
782 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
783 and controls are allowed in which ACLs. There were a couple of minor
784 errors. Some of the entries in the conditions table (which is a table of
785 where they are NOT allowed) were getting very unwieldy; rewrote them as a
786 negation of where the condition IS allowed.
788 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
790 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
791 header file does not have a version number, so I've had to invent a new
792 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
793 API. The code is untested by me (my Linux distribution still has 0.3.2 of
794 radiusclient), but it was contributed by a Radius user.
796 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
797 files or directories, for OpenSSL.
799 PH/19 When an Exim process that is running as root has to create an Exim log
800 file, it does so in a subprocess that runs as exim:exim so as to get the
801 ownership right at creation (otherwise, other Exim processes might see
802 the file with the wrong ownership). There was no test for failure of this
803 fork() call, which would lead to the process getting stuck as it waited
804 for a non-existent subprocess. Forks do occasionally fail when resources
805 run out. I reviewed all the other calls to fork(); they all seem to check
808 PH/20 When checking for unexpected SMTP input at connect time (before writing
809 the banner), Exim was not dealing correctly with a non-positive return
810 from the read() function. If the client had disconnected by this time,
811 the result was a log entry for a synchronization error with an empty
812 string after "input=" when read() returned zero. If read() returned -1
813 (an event I could not check), uninitialized data bytes were printed.
814 There were reports of junk text (parts of files, etc) appearing after
817 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
819 PH/22 Added support for macro redefinition, and (re)definition in between
820 driver and ACL definitions.
822 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
823 forgetting to use the resulting value; it was using the unexpanded value.
825 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
826 hadn't been configured. The fix is from Juergen Kreileder, who
827 understands it better than I do:
829 "Here's what I see happening with three configured cyrus_sasl
830 authenticators configured (plain, login, cram-md5):
832 On startup auth_cyrus_sasl_init() gets called for each of these.
833 This means three calls to sasl_listmech() without a specified mech_list.
834 => SASL tests which mechs of all available mechs actually work
835 => three warnings about OTP not working
836 => the returned list contains: plain, login, cram-md5, digest-md5, ...
838 With the patch, sasl_listmech() also gets called three times. But now
839 SASL's mech_list option is set to the server_mech specified in the the
840 authenticator. Or in other words, the answer from sasl_listmech()
841 gets limited to just the mech you're testing for (which is different
843 => the return list contains just 'plain' or 'login', 'cram-md5' or
844 nothing depending on the value of ob->server_mech.
846 I've just tested the patch: Authentication still works fine,
847 unavailable mechs specified in the exim configuration are still
848 caught, and the auth.log warnings about OTP are gone."
850 PH/25 When debugging is enabled, the contents of the command line are added
851 to the debugging output, even when log_selector=+arguments is not
854 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
855 answer is "GNU", and only if the return is "GNU/something" is the answer
858 PH/27 $acl_verify_message is now set immediately after the failure of a
859 verification in an ACL, and so is available in subsequent modifiers. In
860 particular, the message can be preserved by coding like this:
862 warn !verify = sender
863 set acl_m0 = $acl_verify_message
865 Previously, $acl_verify_message was set only while expanding "message"
866 and "log_message" when a very denied access.
868 PH/28 Modified OS/os.c-Linux with
870 -#ifndef OS_LOAD_AVERAGE
871 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
873 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
874 nomenclature these days.)
876 PH/29 Installed patch from the Sieve maintainer that adds the options
877 sieve_useraddress and sieve_subaddress to the redirect router.
879 PH/30 In these circumstances:
880 . Two addresses routed to the same list of hosts;
881 . First host does not offer TLS;
882 . First host accepts first address;
883 . First host gives temporary error to second address;
884 . Second host offers TLS and a TLS session is established;
885 . Second host accepts second address.
886 Exim incorrectly logged both deliveries with the TLS parameters (cipher
887 and peerdn, if requested) that were in fact used only for the second
890 PH/31 When doing a callout as part of verifying an address, Exim was not paying
891 attention to any local part prefix or suffix that was matched by the
892 router that accepted the address. It now behaves in the same way as it
893 does for delivery: the affixes are removed from the local part unless
894 rcpt_include_affixes is set on the transport.
896 PH/32 Add the sender address, as F=<...>, to the log line when logging a
897 timeout during the DATA phase of an incoming message.
899 PH/33 Sieve envelope tests were broken for match types other than :is. I have
900 applied a patch sanctioned by the Sieve maintainer.
902 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
903 the uid or gid is negative. A case of a negative gid caused this to be
904 noticed. The fix allows for either to be negative.
906 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
907 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
910 PH/36 The change PH/12 above was broken. Fixed it.
912 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
913 the grounds that routing the same address twice would always produce the
914 same answer. This might have been true once, but it is certainly no
915 longer true now. Routing a child address may depend on the previous
916 routing that produced that child. Some complicated redirection strategies
917 went wrong when messages had multiple recipients, and made Exim's
918 behaviour dependent on the order in which the addresses were given.
920 I have moved the duplicate checking until after the routing is complete.
921 Exim scans the addresses that are assigned to local and remote
922 transports, and removes any duplicates. This means that more work will be
923 done, as duplicates will always all be routed, but duplicates are
924 presumably rare, so I don't expect this is of any significance.
926 For deliveries to pipes, files, and autoreplies, the duplicate checking
927 still happens during the routing process, since they are not going to be
930 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
931 It corrects a timeout issue with spamd. This is Ian's comment: "The
932 background is that sometimes spamd either never reads data from a
933 connection it has accepted, or it never writes response data. The exiscan
934 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
935 blindly assumes that writes won't block so it may never time out."
937 PH/39 Allow G after quota size as well as K and M.
939 PH/40 The value set for $authenticated_id in an authenticator may not contain
940 binary zeroes or newlines because the value is written to log lines and
941 to spool files. There was no check on this. Now the value is run through
942 the string_printing() function so that such characters are converted to
943 printable escape sequences.
945 PH/41 $message_linecount is a new variable that contains the total number of
946 lines in the message. Compare $body_linecount, which is the count for the
949 PH/42 Exim no longer gives details of delivery errors for specific addresses in
950 bounce and delay warning messages, except in certain special cases, which
953 (a) An SMTP error message from a remote host;
954 (b) A message specified in a :fail: redirection;
955 (c) A message specified in a "fail" command in a system filter;
956 (d) A message specified in a FAIL return from the queryprogram router;
957 (e) A message specified by the cannot_route_message router option.
959 In these cases only, Exim does include the error details in bounce and
960 warning messages. There are also a few cases where bland messages such
961 as "unrouteable address" or "local delivery error" are given.
963 PH/43 $value is now also set for the "else" part of a ${run expansion.
965 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
966 being worked on, but at least Exim now implements the latest version to
969 PH/45 In a pipe transport, although a timeout while waiting for the pipe
970 process to complete was treated as a delivery failure, a timeout while
971 writing the message to the pipe was logged, but erroneously treated as a
972 successful delivery. Such timeouts include transport filter timeouts. For
973 consistency with the overall process timeout, these timeouts are now
974 treated as errors, giving rise to delivery failures by default. However,
975 there is now a new Boolean option for the pipe transport called
976 timeout_defer, which, if set TRUE, converts the failures into defers for
977 both kinds of timeout. A transport filter timeout is now identified in
980 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
981 systems where "make" and "gmake" are different, calling "gmake" at top
982 level broke things. I've arranged for the value of $(MAKE) to be passed
983 from the Makefile to this script so that it can call the same version of
987 A note about Exim versions 4.44 and 4.50
988 ----------------------------------------
990 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
991 changes of various kinds. As a consequence, a big documentation update was
992 needed. This delayed the release for rather longer than seemed good, especially
993 in the light of a couple of (minor) security issues. Therefore, the changes
994 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
995 release. So 4.44 and 4.50 are in effect two different branches that both start
998 I have left the 4.50 change log unchanged; it contains all the changes since
999 4.43. The change log for 4.44 is below; many of its items are identical to
1000 those for 4.50. This seems to be the most sensible way to preserve the
1001 historical information.
1007 1. Minor wording change to the doc/README.SIEVE file.
1009 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1010 computation of the current number of files was incorrect.
1012 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1013 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1016 4. Give more explanation in the error message when the command for a transport
1017 filter fails to execute.
1019 5. There are several places where Exim runs a non-Exim command in a
1020 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1021 was being done only for the command run by the queryprogram router. It is
1022 now done for all such subprocesses. The other cases are: ${run, transport
1023 filters, and the commands run by the lmtp and pipe transports.
1025 6. Added CONFIGURE_GROUP build-time option.
1027 7. Some older OS have a limit of 256 on the maximum number of file
1028 descriptors. Exim was using setrlimit() to set 1000 as a large value
1029 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1030 systems. I've change it so that if it can't get 1000, it tries for 256.
1032 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1033 was an oversight, and furthermore, ever since the addition of extra
1034 controls (e.g. 4.43/32), the checks on when to allow different forms of
1035 "control" were broken. There should now be diagnostics for all cases when a
1036 control that does not make sense is encountered.
1038 9. Added the /retain_sender option to "control=submission".
1040 10. $recipients is now available in the predata ACL (oversight).
1042 11. Tidy the search cache before the fork to do a delivery from a message
1043 received from the command line. Otherwise the child will trigger a lookup
1044 failure and thereby defer the delivery if it tries to use (for example) a
1045 cached ldap connection that the parent has called unbind on.
1047 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1048 of $address_data from the recipient verification was clobbered by the
1049 sender verification.
1051 13. The value of address_data from a sender verification is now available in
1052 $sender_address_data in subsequent conditions in the ACL statement.
1054 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1056 15. Added a new option "connect=<time>" to callout options, to set a different
1059 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1060 was its contents. (It was OK if the option was not defined at all.)
1062 17. A "Completed" log line is now written for messages that are removed from
1063 the spool by the -Mrm option.
1065 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1066 information about exactly what failed.
1068 19. Added -dd to debug only the daemon process.
1070 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1071 handles timeouts, both on the server side and network timeouts. Renamed the
1072 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1074 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1075 It is now set to "smtps".
1077 22. $host_address is now set to the target address during the checking of
1078 ignore_target_hosts.
1080 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1081 being passed; this would have caused $sender_host_name to have been used if
1082 matching the list had actually called for a host name (not very likely,
1083 since this list is usually IP addresses). A host name is now passed as
1086 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1087 code that specifies a non-restarting handler (typically sigaction() in
1088 modern systems) in an attempt to fix a rare and obscure crash bug.
1090 25. Narrowed the window for a race in the daemon that could cause it to ignore
1091 SIGCHLD signals. This is not a major problem, because they are used only to
1092 wake it up if nothing else does.
1094 26. A malformed maildirsize file could cause Exim to calculate negative values
1095 for the mailbox size or file count. Odd effects could occur as a result.
1096 The maildirsize information is now recalculated if the size or filecount
1099 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1100 support for a long time. Removed HAVE_SYS_VFS_H.
1102 28. Installed the latest version of exipick from John Jetmore.
1104 29. In an address list, if the pattern was not a regular expression, an empty
1105 subject address (from a bounce message) matched only if the pattern was an
1106 empty string. Non-empty patterns were not even tested. This was the wrong
1107 because it is perfectly reasonable to use an empty address as part of a
1108 database query. An empty address is now tested by patterns that are
1109 lookups. However, all the other forms of pattern expect the subject to
1110 contain a local part and a domain, and therefore, for them, an empty
1111 address still always fails if the pattern is not itself empty.
1113 30. Exim went into a mad DNS loop when attempting to do a callout where the
1114 host was specified on an smtp transport, and looking it up yielded more
1115 than one IP address.
1117 31. Re-factored the code for checking spool and log partition space into a
1118 function that finds that data and another that does the check. The former
1119 is then used to implement four new variables: $spool_space, $log_space,
1120 $spool_inodes, and $log_inodes.
1122 32. The RFC2047 encoding function was originally intended for short strings
1123 such as real names; it was not keeping to the 75-character limit for
1124 encoded words that the RFC imposes. It now respects the limit, and
1125 generates multiple encoded words if necessary. To be on the safe side, I
1126 have increased the buffer size for the ${rfc2047: expansion operator from
1129 33. It is now permitted to omit both strings after an "if" condition; if the
1130 condition is true, the result is "true". As before, when the second string
1131 is omitted, a false condition yields an empty string. This makes it less
1132 cumbersome to write custom ACL and router conditions.
1134 34. Failure to deliver a bounce message always caused it to be frozen, even if
1135 there was an errors_to setting on the router. The errors_to setting is now
1138 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1139 canonical form (fully expanded) before being placed in
1140 $sender_host_address.
1142 36. The table in the code that translates DNS record types into text (T_A to
1143 "A" for instance) was missing entries for NS and CNAME. It is just possible
1144 that this could have caused confusion if both these types were looked up
1145 for the same domain, because the text type is used as part of Exim's
1146 per-process caching. But the chance of anyone hitting this buglet seems
1149 37. The dnsdb lookup has been extended in a number of ways.
1151 (1) There is a new type, "zns", which walks up the domain tree until it
1152 finds some nameserver records. It should be used with care.
1154 (2) There is a new type, "mxh", which is like "mx" except that it returns
1155 just the host names, not the priorities.
1157 (3) It is now possible to give a list of domains (or IP addresses) to be
1158 looked up. The behaviour when one of the lookups defers can be
1159 controlled by a keyword.
1161 (4) It is now possible to specify the separator character for use when
1162 multiple records are returned.
1164 38. The dnslists ACL condition has been extended: it is now possible to supply
1165 a list of IP addresses and/or domains to be looked up in a particular DNS
1168 39. Added log_selector=+queue_time_overall.
1170 40. When running the queue in the test harness, wait just a tad after forking a
1171 delivery process, to get repeatability of debugging output.
1173 41. Include certificate and key file names in error message when GnuTLS fails
1174 to set them up, because the GnuTLS error message doesn't include the name
1175 of the failing file when there is a problem reading it.
1177 42. Allow both -bf and -bF in the same test run.
1179 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1181 44. The "Exiscan patch" is now merged into the mainline Exim source.
1183 45. Sometimes the final signoff response after QUIT could fail to get
1184 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1185 < 0 before doing a fflush(). This bug looks as though it goes back to the
1186 introduction of TLS in release 3.20, but "sometimes" must have been rare
1187 because the tests only now provoked it.
1189 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1190 (this can affect the format of dates).
1192 47. exim_tidydb, when checking for the continued existence of a message for
1193 which it has found a message-specific retry record, was not finding
1194 messages that were in split spool directories. Consequently, it was
1195 deleting retry records that should have stayed in existence.
1197 48. Steve fixed some bugs in eximstats.
1199 49. The SPA authentication driver was not abandoning authentication and moving
1200 on to the next authenticator when an expansion was forced to fail,
1201 contradicting the general specification for all authenticators. Instead it
1202 was generating a temporary error. It now behaves as specified.
1204 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1205 (the order specifies the preference for clients). The order is now AES256,
1206 AES128, 3DES, ARCFOUR128.
1208 51. Small patch to Sieve code - explicitly set From: when generating an
1211 52. Exim crashed if a remote delivery caused a very long error message to be
1212 recorded - for instance if somebody sent an entire SpamAssassin report back
1213 as a large number of 550 error lines. This bug was coincidentally fixed by
1214 increasing the size of one of Exim's internal buffers (big_buffer) that
1215 happened as part of the Exiscan merge. However, to be on the safe side, I
1216 have made the code more robust (and fixed the comments that describe what
1219 53. Now that there can be additional text after "Completed" in log lines (if
1220 the queue_time_overall log selector is set), a one-byte patch to exigrep
1221 was needed to allow it to recognize "Completed" as not the last thing in
1224 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1225 patch that reportedly fixes this has been added. I am not expert enough to
1226 create a test for it. This is what the patch creator wrote:
1228 "I found a little strange behaviour of ldap code when working with
1229 Windows 2003 AD Domain, where users was placed in more than one
1230 Organization Units. When I tried to give exim partial DN, the exit code
1231 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1232 But simultaneously result of request was absolutely normal ldap result,
1233 so I produce this patch..."
1235 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1236 REFERENCE, so I have modified the code to exclude the patch when that macro
1239 55. Some experimental protocols are using DNS PTR records for new purposes. The
1240 keys for these records are domain names, not reversed IP addresses. The
1241 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1242 leaves it alone. Component reversal etc. now happens only for IP addresses.
1245 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1247 57. Double the size of the debug message buffer (to 2048) so that more of very
1248 long debug lines gets shown.
1250 58. The exicyclog utility now does better if the number of log files to keep
1251 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1253 59. Two changes related to the smtp_active_hostname option:
1255 (1) $smtp_active_hostname is now available as a variable.
1256 (2) The default for smtp_banner uses $smtp_active_hostname instead
1257 of $primary_hostname.
1259 60. The host_aton() function is supposed to be passed a string that is known
1260 to be a valid IP address. However, in the case of IPv6 addresses, it was
1261 not checking this. This is a hostage to fortune. Exim now panics and dies
1262 if the condition is not met. A case was found where this could be provoked
1263 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1264 components; fortuitously, this particular loophole had already been fixed
1265 by change 4.50/55 above.
1267 If there are any other similar loopholes, the new check in host_aton()
1268 itself should stop them being exploited. The report I received stated that
1269 data on the command line could provoke the exploit when Exim was running as
1270 exim, but did not say which command line option was involved. All I could
1271 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1272 running as the user.
1275 61. There was a buffer overflow vulnerability in the SPA authentication code
1276 (which came originally from the Samba project). I have added a test to the
1277 spa_base64_to_bits() function which I hope fixes it.
1280 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1281 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1283 63. The daemon start-up calls getloadavg() while still root for those OS that
1284 need the first call to be done as root, but it missed one case: when
1285 deliver_queue_load_max is set with deliver_drop_privilege. This is
1286 necessary for the benefit of the queue runner, because there is no re-exec
1287 when deliver_drop_privilege is set.
1289 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1290 This has been fixed.
1292 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1293 was in use, was not putting the data itself into the right store pool;
1294 consequently, it could be overwritten for a subsequent message in the same
1295 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1298 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1300 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1301 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1302 the function were treating the return as a boolean value, which happened to
1303 work because 0=false and not-0=true, but is not correct code.
1305 68. The host_aton() function was not handling scoped IPv6 addresses (those
1306 with, for example, "%eth0" on the end) correctly.
1308 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1309 negated items (that is, ~something) in unsigned ints. Some compilers
1310 apparently mutter when there is no cast.
1312 70. If an address verification called from an ACL failed, and did not produce a
1313 user-specific message (i.e. there was only a "system" message), nothing was
1314 put in $acl_verify_message. In this situation, it now puts the system
1317 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1318 session; change 4.31/43 added the unwanted input to the log line - except
1319 that it did not do this in the start of session case. It now does.
1321 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1322 This is wrong and can cause the other end to generate a synchronization
1323 error if it is another Exim or anything else that does the synchronization
1324 check. A QUIT command is no longer sent after a timeout.
1326 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1327 during host lookups.
1329 74. The defer_ok option of callout verification was not working if it was used
1330 when verifying addresses in header lines, that is, for this case:
1332 verify = header_sender/callout=defer_ok
1334 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1335 those file descriptors could be used for SMTP connections. If anything
1336 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1337 could be sent to the SMTP client, causing chaos. The daemon now opens
1338 stdin, stdout, and stderr to /dev/null when it puts itself into the
1341 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1342 log by default. The user can override this with suitable Perl magic.
1344 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1345 the log message when discard triggers, was not working for the DATA ACL or
1346 for the non-SMTP ACL.
1348 78. Error message wording change in sieve.c.
1350 79. If smtp_accept_max_per_host was set, the number of connections could be
1351 restricted to fewer than expected, because the daemon was trying to set up
1352 a new connection before checking whether the processes handling previous
1353 connections had finished. The check for completed processes is now done
1354 earlier. On busy systems, this bug wouldn't be noticed because something
1355 else would have woken the daemon, and it would have reaped the completed
1358 80. If a message was submitted locally by a user whose login name contained one
1359 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1360 It caused a spool format error. I have fixed the spool reading code. A
1361 related problem was that the "from" clause in the Received: line became
1362 illegal because of the space(s). It is now covered by ${quote_local_part.
1364 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1367 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1369 83. Patch from Sieve maintainer to fix unterminated string problem in
1370 "vacation" handling.
1372 84. Some minor changes to the Linux configuration files to help with other
1373 OS variants using glibc.
1375 85. One more patch for Sieve to update vacation handling to latest spec.
1378 ----------------------------------------------------
1379 See the note above about the 4.44 and 4.50 releases.
1380 ----------------------------------------------------
1386 1. Change 4.43/35 introduced a bug that caused file counts to be
1387 incorrectly computed when quota_filecount was set in an appendfile
1390 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1391 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1394 3. Give more explanation in the error message when the command for a transport
1395 filter fails to execute.
1397 4. There are several places where Exim runs a non-Exim command in a
1398 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1399 was being done only for the command run by the queryprogram router. It is
1400 now done for all such subprocesses. The other cases are: ${run, transport
1401 filters, and the commands run by the lmtp and pipe transports.
1403 5. Some older OS have a limit of 256 on the maximum number of file
1404 descriptors. Exim was using setrlimit() to set 1000 as a large value
1405 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1406 systems. I've change it so that if it can't get 1000, it tries for 256.
1408 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1409 was an oversight, and furthermore, ever since the addition of extra
1410 controls (e.g. 4.43/32), the checks on when to allow different forms of
1411 "control" were broken. There should now be diagnostics for all cases when a
1412 control that does not make sense is encountered.
1414 7. $recipients is now available in the predata ACL (oversight).
1416 8. Tidy the search cache before the fork to do a delivery from a message
1417 received from the command line. Otherwise the child will trigger a lookup
1418 failure and thereby defer the delivery if it tries to use (for example) a
1419 cached ldap connection that the parent has called unbind on.
1421 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1422 of $address_data from the recipient verification was clobbered by the
1423 sender verification.
1425 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1426 was its contents. (It was OK if the option was not defined at all.)
1428 11. A "Completed" log line is now written for messages that are removed from
1429 the spool by the -Mrm option.
1431 12. $host_address is now set to the target address during the checking of
1432 ignore_target_hosts.
1434 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1435 being passed; this would have caused $sender_host_name to have been used if
1436 matching the list had actually called for a host name (not very likely,
1437 since this list is usually IP addresses). A host name is now passed as
1440 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1441 code that specifies a non-restarting handler (typically sigaction() in
1442 modern systems) in an attempt to fix a rare and obscure crash bug.
1444 15. Narrowed the window for a race in the daemon that could cause it to ignore
1445 SIGCHLD signals. This is not a major problem, because they are used only to
1446 wake it up if nothing else does.
1448 16. A malformed maildirsize file could cause Exim to calculate negative values
1449 for the mailbox size or file count. Odd effects could occur as a result.
1450 The maildirsize information is now recalculated if the size or filecount
1453 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1454 support for a long time. Removed HAVE_SYS_VFS_H.
1456 18. Updated exipick to current release from John Jetmore.
1458 19. Allow an empty sender to be matched against a lookup in an address list.
1459 Previously the only cases considered were a regular expression, or an
1462 20. Exim went into a mad DNS lookup loop when doing a callout where the
1463 host was specified on the transport, if the DNS lookup yielded more than
1466 21. The RFC2047 encoding function was originally intended for short strings
1467 such as real names; it was not keeping to the 75-character limit for
1468 encoded words that the RFC imposes. It now respects the limit, and
1469 generates multiple encoded words if necessary. To be on the safe side, I
1470 have increased the buffer size for the ${rfc2047: expansion operator from
1473 22. Failure to deliver a bounce message always caused it to be frozen, even if
1474 there was an errors_to setting on the router. The errors_to setting is now
1477 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1478 canonical form (fully expanded) before being placed in
1479 $sender_host_address.
1481 24. Updated eximstats to version 1.33
1483 25. Include certificate and key file names in error message when GnuTLS fails
1484 to set them up, because the GnuTLS error message doesn't include the name
1485 of the failing file when there is a problem reading it.
1487 26. Expand error message when OpenSSL has problems setting up cert/key files.
1490 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1491 (this can affect the format of dates).
1493 28. exim_tidydb, when checking for the continued existence of a message for
1494 which it has found a message-specific retry record, was not finding
1495 messages that were in split spool directories. Consequently, it was
1496 deleting retry records that should have stayed in existence.
1498 29. eximstats updated to version 1.35
1499 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1500 1.35 - bugfix such that pie charts by volume are generated correctly
1502 30. The SPA authentication driver was not abandoning authentication and moving
1503 on to the next authenticator when an expansion was forced to fail,
1504 contradicting the general specification for all authenticators. Instead it
1505 was generating a temporary error. It now behaves as specified.
1507 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1508 (the order specifies the preference for clients). The order is now AES256,
1509 AES128, 3DES, ARCFOUR128.
1511 31. Small patch to Sieve code - explicitly set From: when generating an
1514 32. Exim crashed if a remote delivery caused a very long error message to be
1515 recorded - for instance if somebody sent an entire SpamAssassin report back
1516 as a large number of 550 error lines. This bug was coincidentally fixed by
1517 increasing the size of one of Exim's internal buffers (big_buffer) that
1518 happened as part of the Exiscan merge. However, to be on the safe side, I
1519 have made the code more robust (and fixed the comments that describe what
1522 33. Some experimental protocols are using DNS PTR records for new purposes. The
1523 keys for these records are domain names, not reversed IP addresses. The
1524 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1525 leaves it alone. Component reversal etc. now happens only for IP addresses.
1528 34. The host_aton() function is supposed to be passed a string that is known
1529 to be a valid IP address. However, in the case of IPv6 addresses, it was
1530 not checking this. This is a hostage to fortune. Exim now panics and dies
1531 if the condition is not met. A case was found where this could be provoked
1532 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1533 components; fortuitously, this particular loophole had already been fixed
1534 by change 4.50/55 or 4.44/33 above.
1536 If there are any other similar loopholes, the new check in host_aton()
1537 itself should stop them being exploited. The report I received stated that
1538 data on the command line could provoke the exploit when Exim was running as
1539 exim, but did not say which command line option was involved. All I could
1540 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1541 running as the user.
1544 35. There was a buffer overflow vulnerability in the SPA authentication code
1545 (which came originally from the Samba project). I have added a test to the
1546 spa_base64_to_bits() function which I hope fixes it.
1549 36. The daemon start-up calls getloadavg() while still root for those OS that
1550 need the first call to be done as root, but it missed one case: when
1551 deliver_queue_load_max is set with deliver_drop_privilege. This is
1552 necessary for the benefit of the queue runner, because there is no re-exec
1553 when deliver_drop_privilege is set.
1555 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1556 was in use, was not putting the data itself into the right store pool;
1557 consequently, it could be overwritten for a subsequent message in the same
1558 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1561 38. Sometimes the final signoff response after QUIT could fail to get
1562 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1563 < 0 before doing a fflush(). This bug looks as though it goes back to the
1564 introduction of TLS in release 3.20, but "sometimes" must have been rare
1565 because the tests only now provoked it.
1571 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1572 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1573 Now it returns an integer. A number of calls were still expecting a T/F
1574 return. Fortuitously, in all cases, the tests worked in OK situations,
1575 which is the norm. However, things would have gone wrong on any write
1576 failures on the smtp file descriptor. This function is used when sending
1577 messages over SMTP and also when doing verify callouts.
1579 2. When Exim is called to do synchronous delivery of a locally submitted
1580 message (the -odf or -odi options), it no longer closes stderr before doing
1583 3. Implemented the mua_wrapper option.
1585 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1587 5. Implemented the functions header_remove(), header_testname(),
1588 header_add_at_position(), and receive_remove_recipient(), and exported them
1591 6. If an ACL "warn" statement specified the addition of headers, Exim already
1592 inserted X-ACL-Warn: at the start if there was no header name. However, it
1593 was not making this test for the second and subsequent header lines if
1594 there were newlines in the string. This meant that an invalid header could
1595 be inserted if Exim was badly configured.
1597 7. Allow an ACL "warn" statement to add header lines at the start or after all
1598 the Received: headers, as well as at the end.
1600 8. Added the rcpt_4xx retry error code.
1602 9. Added postmaster_mailfrom=xxx to callout verification option.
1604 10. Added mailfrom=xxxx to the callout verification option, for verify=
1607 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1608 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1610 12. Inserted some casts to stop certain compilers warning when using pointer
1611 differences as field lengths or precisions in printf-type calls (mostly
1612 affecting debugging statements).
1614 13. Added optional readline() support for -be (dynamically loaded).
1616 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1617 same clock tick as a message's arrival, so that its received time was the
1618 same as the "first fail" time on the retry record, and that message
1619 remained on the queue past the ultimate address timeout, every queue runner
1620 would try a delivery (because it was past the ultimate address timeout) but
1621 after another failure, the ultimate address timeout, which should have then
1622 bounced the address, did not kick in. This was a "< instead of <=" error;
1623 in most cases the first failure would have been in the next clock tick
1624 after the received time, and all would be well.
1626 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1627 being recognized when the domain list was tested by the match_domain
1628 condition in an expansion string.
1630 16. Added the ${str2b64: operator.
1632 17. Exim was always calling setrlimit() to set a large limit for the number of
1633 processes, without checking whether the existing limit was already
1634 adequate. (It did check for the limit on file descriptors.) Furthermore,
1635 errors from getrlimit() and setrlimit() were being ignored. Now they are
1636 logged to the main and panic logs, but Exim does carry on, to try to do its
1637 job under whatever limits there are.
1639 18. Imported PCRE 5.0.
1641 19. Trivial typo in log message " temporarily refused connection" (the leading
1644 20. If the log selector return_path_on_delivery was set and an address was
1645 redirected to /dev/null, the delivery process crashed because it assumed
1646 that a return path would always be set for a "successful" delivery. In this
1647 case, the whole delivery is bypassed as an optimization, and therefore no
1650 21. Internal re-arrangement: the function for sending a challenge and reading
1651 a response while authentication was assuming a zero-terminated challenge
1652 string. It's now changed to take a pointer and a length, to allow for
1653 binary data in such strings.
1655 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1657 23. Exim was not respecting finduser_retries when seeking the login of the
1658 uid under which it was called; it was always trying 10 times. (The default
1659 setting of finduser_retries is zero.) Also, it was sleeping after the final
1660 failure, which is pointless.
1662 24. Implemented tls_on_connect_ports.
1664 25. Implemented acl_smtp_predata.
1666 26. If the domain in control=submission is set empty, Exim assumes that the
1667 authenticated id is a complete email address when it generates From: or
1668 Sender: header lines.
1670 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1671 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1672 chown and chgrp in /bin and hostname in /usr/bin.
1674 28. Exim was keeping the "process log" file open after each use, just as it
1675 does for the main log. This opens the possibility of it remaining open for
1676 long periods when the USR1 signal hits a daemon. Occasional processlog
1677 errors were reported, that could have been caused by this. Anyway, it seems
1678 much more sensible not to leave this file open at all, so that is what now
1681 29. The long-running daemon process does not normally write to the log once it
1682 has entered its main loop, and it closes the log before doing so. This is
1683 so that log files can straightforwardly be renamed and moved. However,
1684 there are a couple of unusual error situations where the daemon does write
1685 log entries, and I had neglected to close the log afterwards.
1687 30. The text of an SMTP error response that was received during a remote
1688 delivery was being truncated at 512 bytes. This is too short for some of
1689 the long messages that one sometimes sees. I've increased the limit to
1692 31. It is now possible to make retry rules that apply only when a message has a
1693 specific sender, in particular, an empty sender.
1695 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1696 it possible to be selective about when SMTP synchronization is enforced.
1698 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1700 32. Implemented hosts_connection_nolog.
1702 33. Added an ACL for QUIT.
1704 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1707 35. Added mailbox_size and mailbox_filecount to appendfile.
1709 36. Added control = no_multiline_responses to ACLs.
1711 37. There was a bug in the logic of the code that waits for the clock to tick
1712 in the case where the clock went backwards by a substantial amount such
1713 that the microsecond fraction of "now" was more than the microsecond
1714 fraction of "then" (but the whole seconds number was less).
1716 38. Added support for the libradius Radius client library this is found on
1717 FreeBSD (previously only the radiusclient library was supported).
1723 1. When certain lookups returned multiple values in the form name=value, the
1724 quoting of the values was not always being done properly. Specifically:
1725 (a) If the value started with a double quote, but contained no whitespace,
1727 (b) If the value contained whitespace other than a space character (i.e.
1728 tabs or newlines or carriage returns) it was not quoted.
1729 This fix has been applied to the mysql and pgsql lookups by writing a
1730 separate quoting function and calling it from the lookup code. The fix
1731 should probably also be applied to nisplus, ibase and oracle lookups, but
1732 since I cannot test any of those, I have not disturbed their existing code.
1734 2. A hit in the callout cache for a specific address caused a log line with no
1735 reason for rejecting RCPT. Now it says "Previous (cached) callout
1736 verification failure".
1738 3. There was an off-by-one bug in the queryprogram router. An over-long
1739 return line was truncated at 256 instead of 255 characters, thereby
1740 overflowing its buffer with the terminating zero. As well as fixing this, I
1741 have increased the buffer size to 1024 (and made a note to document this).
1743 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1744 when Exim is waiting for an SMTP response from a remote server, Exim
1745 restarts its select() call on the socket, thereby resetting its timeout.
1746 This is not a problem when such interrupts are rare. Somebody set up a cron
1747 job to run exiwhat every 2 minutes, which is less than the normal select()
1748 timeout (5 or 10 minutes). This meant that the select() timeout never
1749 kicked in because it was always reset. I have fixed this by comparing the
1750 time when an interrupt arrives with the time at the start of the first call
1751 to select(). If more time than the timeout has elapsed, the interrupt is
1752 treated as a timeout.
1754 5. Some internal re-factoring in preparation for the addition of Sieve
1755 extensions (by MH). In particular, the "personal" test is moved to a
1756 separate function, and given an option for scanning Cc: and Bcc: (which is
1757 not set for Exim filters).
1759 6. When Exim created an email address using the login of the caller as the
1760 local part (e.g. when creating a From: or Sender: header line), it was not
1761 quoting the local part when it contained special characters such as @.
1763 7. Installed new OpenBSD configuration files.
1765 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1766 try to make them clearer.
1768 9. Callout options, other than the timeout value, were being ignored when
1769 verifying sender addresses in header lines. For example, when using
1771 verify = header_sender/callout=no_cache
1773 the cache was (incorrectly) being used.
1775 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1776 only the Cygwin environment.
1778 11. When return_path_on_delivery was set as a log selector, if different remote
1779 addresses in the same message used different return paths and parallel
1780 remote delivery occurred, the wrong values would sometimes be logged.
1781 (Whenever a remote delivery process finished, the return path value from
1782 the most recently started remote delivery process was logged.)
1784 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1785 lines when AUTH and/or TLS are in use. This is the "received protocol"
1786 field. Exim used to use "asmtp" for authenticated SMTP, without any
1787 indication (in the protocol name) for TLS use. Now it follows the RFC and
1788 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1789 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1790 names appear in log lines as well as in Received: header lines.
1792 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1793 extensions, and comparison tests, and to fix some bugs.
1795 14. Changes to the "personal" filter test:
1797 (1) The test was buggy in that it was just doing the equivalent of
1798 "contains" tests on header lines. For example, if a user's address was
1799 anne@some.where, the "personal" test would incorrectly be true for
1801 To: susanne@some.where
1803 This test is now done by extracting each address from the header in turn,
1804 and checking the entire address. Other tests that are part of "personal"
1805 are now done using regular expressions (for example, to check local parts
1806 of addresses in From: header lines).
1808 (2) The list of non-personal local parts in From: addresses has been
1809 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1810 taken from the Sieve specification recommendations.
1812 (3) If the message contains any header line starting with "List-" it is
1813 treated as non-personal.
1815 (4) The test for "circular" in the Subject: header line has been removed
1816 because it now seems ill-conceived.
1818 15. Minor typos in src/EDITME comments corrected.
1820 16. Installed latest exipick from John Jetmore.
1822 17. If headers_add on a router specified a text string that was too long for
1823 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1824 of string_sprintf() is now avoided.
1826 18. $message_body_size was not set (it was always zero) when running the DATA
1827 ACL and the local_scan() function.
1829 19. For the "mail" command in an Exim filter, no default was being set for
1830 the once_repeat time, causing a random time value to be used if "once" was
1831 specified. (If the value happened to be <= 0, no repeat happened.) The
1832 default is now 0s, meaning "never repeat". The "vacation" command was OK
1833 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1834 (I found it when inspecting the code).
1836 20. There is now an overall timeout for performing a callout verification. It
1837 defaults to 4 times the callout timeout, which applies to individual SMTP
1838 commands during the callout. The overall timeout applies when there is more
1839 than one host that can be tried. The timeout is checked before trying the
1840 next host. This prevents very long delays if there are a large number of
1841 hosts and all are timing out (e.g. when the network connections are timing
1842 out). The value of the overall timeout can be changed by specifying an
1843 additional sub-option for "callout", called "maxwait". For example:
1845 verify = sender/callout=5s,maxwait=20s
1847 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1848 to the end before writing, but this should make it even safer).
1850 22. Exim was forgetting that it had advertised PIPELINING for the second and
1851 subsequent messages on an SMTP connection. It was also not resetting its
1852 memory on STARTTLS and an internal HELO.
1854 23. When Exim logs an SMTP synchronization error within a session, it now
1855 records whether PIPELINING has been advertised or not.
1857 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1858 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1859 rather than long int.
1861 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1863 26. Added the never_mail option to autoreply.
1869 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1870 crash if the getsockname() call failed; this can happen if a connection is
1871 closed very soon after it is established. The problem was simply in the
1872 order in which certain operations were done, causing Exim to try to write
1873 to the SMTP stream before it had set up the file descriptor. The bug has
1874 been fixed by making things happen in the correct order.
1880 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1881 before the connection was closed, thus losing the rejection response.
1883 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1884 some early Solaris releases, but causes trouble in current releases where
1885 socklen_t is defined.
1887 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1890 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1893 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1894 that releasing the top of it at the end releases what was used for sub-
1895 expansions (unless the block got too big). However, discard this block if
1896 the first thing is a variable or header, so that we can use its block when
1897 it is dynamic (useful for very large $message_headers, for example).
1899 6. Lookups now cache *every* query, not just the most recent. A new, separate
1900 store pool is used for this. It can be recovered when all lookup caches are
1901 flushed. Lookups now release memory at the end of their result strings.
1902 This has involved some general refactoring of the lookup sources.
1904 7. Some code has been added to the store_xxx() functions to reduce the amount
1905 of flapping under certain conditions.
1907 8. log_incoming_interface used to affect only the <= reception log lines. Now
1908 it causes the local interface and port to be added to several more SMTP log
1909 lines, for example "SMTP connection from", and rejection lines.
1911 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1913 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1915 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1916 could be overwritten at the end of the current message (or the start of a
1917 new message if it was set in a HELO ACL). The value is now preserved for
1918 the duration of the SMTP connection.
1920 12. If a transport had a headers_rewrite setting, and a matching header line
1921 contained an unqualified address, that address was qualified, even if it
1922 did not match any rewriting rules. The underlying bug was that the values
1923 of the flags that permit the existence of unqualified sender and recipient
1924 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1925 non-local messages, and by -bnq for local messages) were not being
1926 preserved with the message after it was received.
1928 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1929 "next input=" as part of the text comprising the host identity instead of
1930 the correct text. The code was using the same buffer for two different
1931 strings. However, depending on which order the printing function evaluated
1932 its arguments, the bug did not always show up. Under Linux, for example, my
1933 test suite worked just fine.
1935 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1936 doesn't work with some older versions of Perl. It has been changed to "my",
1937 which in any case is probably the better facility to use.
1939 15. A really picky compiler found some instances of statements for creating
1940 error messages that either had too many or two few arguments for the format
1943 16. The size of the buffer for calls to the DNS resolver has been increased
1944 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1945 for addresses that have a lot of PTR records. This alleviates a problem; it
1946 does not fully solve it.
1948 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1949 buffer now truncates the list and logs the incident, which is the same
1950 action as happens when Exim is looking up a host name and its aliases.
1951 Previously in this situation something unpredictable would happen;
1952 sometimes it was "internal error: store_reset failed".
1954 18. If a server dropped the connection unexpectedly when an Exim client was
1955 using GnuTLS and trying to read a response, the client delivery process
1956 crashed while trying to generate an error log message.
1958 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1959 string, for example:
1961 warn message = H1: something\nH2: something
1963 the text was added as a single header line from Exim's point of view
1964 though it ended up OK in the delivered message. However, searching for the
1965 second and subsequent header lines using $h_h2: did not work. This has been
1966 fixed. Similarly, if a system filter added multiple headers in this way,
1967 the routers could not see them.
1969 20. Expanded the error message when iplsearch is called with an invalid key to
1970 suggest using net-iplsearch in a host list.
1972 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1973 ACLs are no longer actually imposed (and a message to that effect is
1976 22. If a "gecos" field in a passwd entry contained escaped characters, in
1977 particular, if it contained a \" sequence, Exim got it wrong when building
1978 a From: or a Sender: header from that name. A second bug also caused
1979 incorrect handling when an unquoted " was present following a character
1980 that needed quoting.
1982 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1983 was not being matched caselessly.
1985 24. Arranged for all hyphens in the exim.8 source to be escaped with
1988 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1989 properly. Recipient callout cache records were still being keyed to include
1990 the sender, even when use_sender was set false. This led to far more
1991 callouts that were necessary. The sender is no longer included in the key
1992 when use_sender is false.
1994 26. Added "control = submission" modifier to ACLs.
1996 27. Added the ${base62d: operator to decode base 62 numbers.
1998 28. dnsdb lookups can now access SRV records.
2000 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2001 the configuration file.
2003 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2004 (-v) mode. This makes the output for a verbose queue run more intelligible.
2006 31. Added a use_postmaster feature to recipient callouts.
2008 32. Added the $body_zerocount variable, containing the number of binary zero
2009 bytes in the message body.
2011 33. The time of last modification of the "new" subdirectory is now used as the
2012 "mailbox time last read" when there is a quota error for a maildir
2015 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2017 35. Added +ignore_unknown as a special item in host lists.
2019 36. Code for decoding IPv6 addresses in host lists is now included, even if
2020 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2021 address was recognized as an IP address, but was then not correctly decoded
2022 into binary, causing unexpected and incorrect effects when compared with
2029 1. Very minor rewording of debugging text in manualroute to say "list of
2030 hosts" instead of "hostlist".
2032 2. If verify=header_syntax was set, and a header line with an unqualified
2033 address (no domain) and a large number of spaces between the end of the
2034 name and the colon was received, the reception process suffered a buffer
2035 overflow, and (when I tested it) crashed. This was caused by some obsolete
2036 code that should have been removed. The fix is to remove it!
2038 3. When running in the test harness, delay a bit after writing a bounce
2039 message to get a bit more predictability in the log output.
2041 4. Added a call to search_tidyup() just before forking a reception process. In
2042 theory, someone could use a lookup in the expansion of smtp_accept_max_
2043 per_host which, without the tidyup, could leave open a database connection.
2045 5. Added the variables $recipient_data and $sender_data which get set from a
2046 lookup success in an ACL "recipients" or "senders" condition, or a router
2047 "senders" option, similar to $domain_data and $local_part_data.
2049 6. Moved the writing of debug_print from before to after the "senders" test
2052 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2053 problems for message scanning, either using a data ACL, or using
2054 local_scan() because the Received: header was not generated till after they
2055 were called (in order to set the time as the time of reception completion).
2056 I have revised the way this works. The header is now generated after the
2057 body is received, but before the ACL or local_scan() are called. After they
2058 are run, the timestamp in the header is updated.
2064 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2065 before starting a queue runner without re-exec. This happened only when
2066 deliver_drop_privilege was set or when the Exim user was set to root. The
2067 effect of the bug was that timeouts during subsequent deliveries caused
2068 crashes instead of being properly handled. The handler is now left at its
2069 default (and expected) setting.
2071 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2072 message, again when deliver_drop_privilege is set or Exim is run as root.
2073 The bug described in (1) was not present in this case, but the tidying up
2074 of the other signals was missing. I have made the two cases consistent.
2076 3. The ignore_target_hosts setting on a manualroute router was being ignored
2077 for hosts that were looked up using the /MX notation.
2079 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2082 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2083 operated on the sender address. After changing the $sender_address to <>
2084 for the sender address verify, Exim was re-instated it as the original
2085 (before rewriting) address, but remembering that it had rewritten it, so it
2086 wasn't rewriting it again. This bug also had the effect of breaking the
2087 sender address verification caching when the sender address was rewritten.
2089 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2090 This has been changed so that if the ip literal address matches
2091 ignore_target_hosts, the router declines.
2093 7. Added expansion conditions match_domain, match_address, and match_local_
2094 part (NOT match_host).
2096 8. The placeholder for the Received: header didn't have a length field set.
2098 9. Added code to Exim itself and to exim_lock to test for a specific race
2099 condition that could lead to file corruption when using MBX delivery. The
2100 issue is with the lockfile that is created in /tmp. If this file is removed
2101 after a process has opened it but before that process has acquired a lock,
2102 there is the potential for a second process to recreate the file and also
2103 acquire a lock. This could lead to two Exim processes writing to the file
2104 at the same time. The added code performs the same test as UW imapd; it
2105 checks after acquiring the lock that its file descriptor still refers to
2106 the same named file.
2108 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2109 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2112 11. Added the smtp_active_hostname option. If used, this will typically be made
2113 to depend on the incoming interface address. Because $interface_address is
2114 not set up until the daemon has forked a reception process, error responses
2115 that can happen earlier (such as "too many connections") no longer contain
2118 12. If an expansion in a condition on a "warn" statement fails because a lookup
2119 defers, the "warn" statement is abandoned, and the next ACL statement is
2120 processed. Previously this caused the whole ACL to be aborted.
2122 13. Added the iplsearch lookup type.
2124 14. Added ident_timeout as a log selector.
2126 15. Added tls_certificate_verified as a log selector.
2128 16. Added a global option tls_require_ciphers (compare the smtp transport
2129 option of the same name). This controls incoming TLS connections.
2131 17. I finally figured out how to make tls_require_ciphers do a similar thing
2132 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2133 before starting the TLS session.
2135 18. Tabs are now shown as \t in -bP output.
2137 19. If the log selector return_path_on_delivery was set, Exim crashed when
2138 bouncing a message because it had too many Received: header lines.
2140 20. If two routers both had headers_remove settings, and the first one included
2141 a superfluous trailing colon, the final name in the first list and the
2142 first name in the second list were incorrectly joined into one item (with a
2143 colon in the middle).
2149 1. Added -C and -D options to the exinext utility, mainly to make it easier
2150 to include in the automated testing, but these could be helpful when
2151 multiple configurations are in use.
2153 2. The exinext utility was not formatting the output nicely when there was
2154 an alternate port involved in the retry record key, nor when there was a
2155 message id as well (for retries that were specific to a specific message
2156 and a specific host). It was also confused by IPv6 addresses, because of
2157 the additional colons they contain. I have fixed the IPv4 problem, and
2158 patched it up to do a reasonable job for IPv6.
2160 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2161 delivery, the log line now contains "pipelined" if PIPELINING was used.
2163 4. An SMTP transport process used to panic and die if the bind() call to set
2164 an explicit outgoing interface failed. This has been changed; it is now
2165 treated in the same way as a connect() failure.
2167 5. A reference to $sender_host_name in the part of a conditional expansion
2168 that was being skipped was still causing a DNS lookup. This no longer
2171 6. The def: expansion condition was not recognizing references to header lines
2172 that used bh_ and bheader_.
2174 7. Added the _cache feature to named lists.
2176 8. The code for checking quota_filecount in the appendfile transport was
2177 allowing one more file than it should have been.
2179 9. For compatibility with Sendmail, the command line option
2187 and sets the incoming protocol and host name (for trusted callers). The
2188 host name and its colon can be omitted when only the protocol is to be set.
2189 Note the Exim already has two private options, -pd and -ps, that refer to
2190 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2191 "s", but I don't think that's a major issue.
2193 10. A number of refactoring changes to the code, none of which should affect
2196 (a) The number of logging options was getting close to filling up the
2197 32-bit word that was used as a bit map. I have split them into two classes:
2198 those that are passed in the argument to log_write(), and those that are
2199 only ever tested independently outside of that function. These are now in
2200 separate 32-bit words, so there is plenty of room for expansion again.
2201 There is no change in the user interface or the logging behaviour.
2203 (b) When building, for example, log lines, the code previously used a
2204 macro that called string_cat() twice, in order to add two strings. This is
2205 not really sufficiently general. Furthermore, there was one instance where
2206 it was actually wrong because one of the argument was used twice, and in
2207 one call a function was used. (As it happened, calling the function twice
2208 did not affect the overall behaviour.) The macro has been replaced by a
2209 function that can join an arbitrary number of extra strings onto a growing
2212 (c) The code for expansion conditions now uses a table and a binary chop
2213 instead of a serial search (which was left over from when there were very
2214 few conditions). Also, it now recognizes conditions like "pam" even when
2215 the relevant support is not compiled in: a suitably worded error message is
2216 given if an attempt is made to use such a condition.
2218 11. Added ${time_interval:xxxxx}.
2220 12. A bug was causing one of the ddress fields not to be passed back correctly
2221 from remote delivery subprocesses. The field in question was not being
2222 subsequently used, so this caused to problems in practice.
2224 13. Added new log selectors queue_time and deliver_time.
2226 14. Might have fixed a bug in maildirsizefile handling that threw up
2227 "unexpected character" debug warnings, and recalculated the data
2228 unnecessarily. In any case, I expanded the warning message to give more
2231 15. Added the message "Restricted characters in address" to the statements in
2232 the default ACL that block characters like @ and % in local parts.
2234 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2235 Three changes have been made:
2237 (a) There was a serious bug; a negative response to MAIL caused the whole
2238 recipient domain to be cached as invalid, thereby blocking all messages
2239 to all local parts at the same domain, from all senders. This bug has
2240 been fixed. The domain is no longer cached after a negative response to
2241 MAIL if the sender used is not empty.
2243 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2246 (c) A new callout option, "use_sender" has been added for people who want
2247 the modified behaviour.
2253 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2256 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2257 indeed breaks things for older releases.
2259 3. Added additional logging to the case where there is a problem reading data
2260 from a filter that is running in a subprocess using a pipe, in order to
2261 try to track down a specific problem.
2263 4. Testing facility fudge: when running in the test harness and attempting
2264 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2265 getting "No route to host". Convert this to a timeout.
2267 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2270 6. Some OS don't have socklen_t but use size_t instead. This affects the
2271 fifth argument of getsockopt() amongst other things. This is now
2272 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2273 can be set for individual OS. I have set it for SunOS5, OSF1, and
2274 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2275 some earlier ones do not.
2277 7. Change 4.30/15 was not doing the test caselessly.
2279 8. The standard form for an IPv6 address literal was being rejected by address
2280 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2281 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2282 this, as well as the form without the "IPv6" on the front (but only when
2283 address literals are enabled, of course).
2285 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2287 10. Exim crashed if a message with an empty sender address specified by -f
2288 encountered a router with an errors_to setting. This could be provoked only
2289 by a command such as
2293 where an empty string was supplied; "<>" did not hit this bug.
2295 11. Installed PCRE release 4.5.
2297 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2298 remained set. It is now erased.
2300 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2301 times from message ids (which are base 36 rather than the normal 62).
2303 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2304 were being counted as actual protocol errors, and logged if the log
2305 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2306 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2307 and DATA following a set of rejected RCPTs do not count as protocol errors.
2308 In other words, Exim assumes they were pipelined, though this may not
2309 actually be the case. Of course, in all cases the client gets an
2310 appropriate error code.
2312 15. If a lookup fails in an ACL condition, a message about the failure may
2313 be available; it is used if testing the ACL cannot continue, because most
2314 such messages specify what the cause of the deferral is. However, some
2315 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2316 that caused an old message to be retained and used if a later statement
2317 caused a defer, replacing the real cause of the deferral.
2319 16. If an IP address had so many PTR records that the DNS lookup buffer
2320 was not large enough to hold them, Exim could crash while trying to process
2321 the truncated data. It now detects and logs this case.
2323 17. Further to 4.21/58, another change has been made: if (and only if) the
2324 first line of a message (the first header line) ends with CRLF, a bare LF
2325 in a subsequent header line has a space inserted after it, so as not to
2326 terminate the header.
2328 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2329 unnecessarily, used atoi() instead of strtol(), and didn't check the
2330 termination when getting file sizes from file names by regex.
2332 19. Completely re-implemented the support for maildirsize files, in the light
2333 of a number of problems with the previous contributed implementation
2334 (4.30/29). In particular:
2336 . If the quota is zero, the maildirsize file is maintained, but no quota is
2339 . If the maildir directory does not exist, it is created before any attempt
2340 to write a maildirsize file.
2342 . The quota value in the file is just a cache; if the quota is changed in
2343 the transport, the new value overrides.
2345 . A regular expression is available for excluding directories from the
2348 20. The autoreply transport checks the characters in options that define the
2349 message's headers; it allows continued headers, but it was checking with
2350 isspace() after an embedded newline instead of explicitly looking for a
2353 21. If all the "regular" hosts to which an address was routed had passed their
2354 expiry times, and had not reached their retry times, the address was
2355 bounced, even if fallback hosts were defined. Now Exim should go on to try
2358 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2359 equivalent code in the SMTP transport. Some hosts send humungous responses
2360 to HELO/EHLO, more than 1024 it seems.
2362 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2363 gives compiler warnings in some environments. I've now done it "properly",
2366 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2367 (because of problems with the built-in one) was declared to return uschar *
2368 instead of char *, causing compiler failure.
2370 25. Fixed a file descriptor leak when processing alias/forward files.
2372 26. Fixed a minor format string issue in dbfn.c.
2374 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2376 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2377 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2380 29. When a sender address is verified, it is cached, to save repeating the test
2381 when there is more than one recipient in a message. However, when the
2382 verification involves a callout, it is possible for different callout
2383 options to be set for different recipients. It is too complicated to keep
2384 track of this in the cache, so now Exim always runs a verification when a
2385 callout is required, relying on the callout cache for the optimization.
2386 The overhead is duplication of the address routing, but this should not be
2389 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2390 to be verified with callout=postmaster, and the main callout worked but the
2391 postmaster check failed, the verification correctly failed. However, if a
2392 subsequent RCPT command asked for sender verification *without* the
2393 postmaster check, incorrect caching caused this verification also to fail,
2396 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2397 it was not caching the DNS options (qualify_single, search_parents) that
2398 were used when the lookup failed. A subsequent lookup with different
2399 options therefore always gave the same answer, though there were cases
2400 where it should not have. (Example: a "domains = !$mx_any" option on a
2401 dnslookup router: the "domains" option is always processed without any
2402 widening, but the router might have qualify_single set.) Now Exim uses the
2403 cached value only when the same options are set.
2405 32. Added John Jetmore's "exipick" utility to the distribution.
2407 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2408 than a timeout (e.g. a certificate is required, and is not provided), an
2409 Exim server now closes the connection immediately. Previously it waited for
2410 the client to close - but if the client is SSL, it seems that they each
2411 wait for each other, leading to a delay before one of them times out.
2413 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2414 maintained 0.8.x compatibility because I don't think many are using it, and
2415 it is clearly obsolete.
2417 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2420 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2421 client certificate was expired. A simple patch fixes this, though I don't
2422 understand the full logic of why the verify callback is called multiple
2425 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2426 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2427 which causes problems with some clients (such as the Certicom SSL Plus
2428 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2429 disables the coutermeasure allowing Eudora to connect."
2431 38. Exim was not checking that a write() to a log file succeeded. This could
2432 lead to Bad Things if a log got too big, in particular if it hit a file
2433 size limit. Exim now panics and dies if it cannot write to a log file, just
2434 as it does if it cannot open a log file.
2436 39. Modified OS/Makefile-Linux so that it now contains
2438 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2440 The two -D definitions ensure that Exim is compiled with large file
2441 support, which makes it possible to handle log files that are bigger than
2444 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2445 instance) a domain was checked against a named list that involved a lookup,
2446 causing $domain_data to be set, then another domain was checked against the
2447 same list, then the first domain was re-checked, the value of $domain_data
2448 after the final check could be wrong. In particular, if the second check
2449 failed, it could be set empty. This bug probably also applied to
2452 41. The strip_trailing_dot option was not being applied to the address given
2453 with the -f command-line option.
2455 42. The code for reading a message's header from the spool was incrementing
2456 $received_count, but never initializing it. This meant that the value was
2457 incorrect (doubled) while delivering a message in the same process in which
2458 it was received. In the most common configuration of Exim, this never
2459 happens - a fresh exec is done - but it can happen when
2460 deliver_drop_privilege is set.
2462 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2463 it now includes up to 150 characters of the unexpected data in the log
2466 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2467 and building data strings. The size of both of these buffers was 10 000
2468 bytes - far larger than anybody would *ever* want, thought I. Needless to
2469 say, somebody hit the limit. I have increased the maximum line length to
2470 20 000 and the maximum data length of concatenated lines to 100 000. I have
2471 also fixed two bugs, because there was no checking on these buffers. Tsk,
2472 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2473 buffer is too small.
2475 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2476 lsearch lookups. Now it does.
2478 46. When parsing a route_list item in a manualroute router, a fixed-length
2479 buffer was used for the list of hosts. I made this 1024 bytes long,
2480 thinking that nobody would ever have a list of hosts that long. Wrong.
2481 Somebody had a whole pile of complicated expansion conditions, and the
2482 string was silently truncated, leading to an expansion error. It turns out
2483 that it is easier to change to an unlimited length (owing to other changes
2484 that have happened since this code was originally written) than to build
2485 structure for giving a limitation error. The length of the item that
2486 expands into the list of hosts is now unlimited.
2488 47. The lsearch lookup could not handle data where the length of text line was
2489 more than 4095 characters. Such lines were truncated, leading to shortened
2490 data being returned. It should now handle lines of any length.
2492 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2493 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2496 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2497 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2498 no longer happens in comment lines. A list of replacements is now placed
2499 at the head of all of the source files, except those whose only change is
2500 to replace PERL_COMMAND in the very first #! line.
2502 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2503 messages on the queue, with a bottom-up merge sort, using code contributed
2504 by Michael Haardt. This should make operations like -bp somewhat faster on
2505 large queues. It won't affect queue runners, except when queue_run_in_order
2508 51. Installed eximstats 1.31 in the distribution.
2510 52. Added support for SRV lookups to the dnslookup router.
2512 53. If an ACL referred to $message_body or $message_body_end, the value was not
2513 reset for any messages that followed in the same SMTP session.
2515 54. The store-handling optimization for building very long strings was not
2516 differentiating between the different store pools. I don't think this
2517 actually made any difference in practice, but I've tidied it.
2519 55. While running the routers to verify a sender address, $sender_address
2520 was still set to the sender address. This is wrong, because when routing to
2521 send a bounce to the sender, it would be empty. Therefore, I have changed
2522 it so that, while verifying a sender address, $sender_address is set to <>.
2523 (There is no change to what happens when verifying a recipient address.)
2525 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2526 target A or AAAA records (if not already returned) without resetting the
2527 qualify_single or search_parents options of the DNS resolver. These are
2528 inappropriate in this case because the targets of MX and SRV records must
2529 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2530 target that, when qualified, matched something in the local domain. These
2531 two options are now turned off when doing these lookups.
2533 57. It seems that at least some releases of Reiserfs (which does not have the
2534 concept of a fixed number of inodes) returns zero and not -1 for the
2535 number of available inodes. This interacted badly with check_spool_inodes,
2536 which assumed that -1 was the "no such thing" setting. What I have done is
2537 to check that the total number of inodes is greater than zero before doing
2538 the test of how many are available.
2540 58. When a "warn" ACL statement has a log_message modifier, the message is
2541 remembered, and not repeated. This is to avoid a lot of repetition when a
2542 message has many recipients that cause the same warning to be written.
2543 Howewer, Exim was preserving the list of already written lines for an
2544 entire SMTP session, which doesn't seem right. The memory is now reset if a
2545 new message is started.
2547 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2548 debugging output unless log_rewrite was also set.
2550 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2551 of (int)(handle) when we know that handle contains (void *)(-1).
2553 61. The Exim daemon panic-logs an error return when it closes the incoming
2554 connection. However "connection reset by peer" seems to be common, and
2555 isn't really an error worthy of noting specially, so that particular error
2558 62. When Exim is trying to find all the local interfaces, it used to panic and
2559 die if the ioctl to get the interface flags failed. However, it seems that
2560 on at least one OS (Solaris 9) it is possible to have an interface that is
2561 included in the list of interfaces, but for which you get a failure error
2562 for this call. This happens when the interface is not "plumbed" into a
2563 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2564 failure of the "get flags" call assumes that the interface is down.
2566 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2567 makes life easier for people who are doing arithmetic on fields extracted
2568 from dates, where you often get leading zeros that should not be
2569 interpreted as octal.
2571 64. Added qualify_domain to the redirect router, to override the global
2574 65. If a pathologically long header line contained very many addresses (the
2575 report of this problem mentioned 10 000) and each of them was rewritten,
2576 Exim could use up a very large amount of memory. (It kept on making new
2577 copies of the header line as it rewrote, and never released the old ones.)
2578 At the expense of a bit more processing, the header rewriting function has
2579 been changed so that it no longer eats memory in this way.
2581 66. The generation of the Received: header has been moved from the time that a
2582 message starts to be received, to the time that it finishes. The timestamp
2583 in the Received: header should now be very close to that of the <= log
2584 line. There are two side-effects of this change:
2586 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2587 logged header lines no longer include the local Received: line, because
2588 it has not yet been created. The same applies to a copy of the message
2589 that is returned to a non-SMTP sender when a message is rejected.
2591 (b) When a filter file is tested using -bf, no additional Received: header
2592 is added to the test message. After some thought, I decided that this
2595 This change does not affect the value of $received_for. It is still set
2596 after address rewriting, but before local_scan() is called.
2598 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2600 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2601 gave an unhelpful panic error message, and a defer error. I have managed to
2602 change this behaviour so that it now rejects any supplied certificate,
2603 which seems right, as the list of acceptable certificates is empty.
2605 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2606 gave an unhelpful defer error. I have not managed to make this reject any
2607 supplied certificates, but the error message it gives is "no certificate
2608 supplied", which is not helpful.
2610 70. exigrep's output now also includes lines that are not associated with any
2611 message, but which match the given pattern. Implemented by a patch from
2612 Martin Sluka, which also tidied up the Perl a bit.
2614 71. Recipient callout verification, like sender verification, was using <> in
2615 the MAIL FROM command. This isn't really the right thing, since the actual
2616 sender may affect whether the remote host accepts the recipient or not. I
2617 have changed it to use the actual sender in the callout; this means that
2618 the cache record is now keyed on a recipient/sender pair, not just the
2619 recipient address. There doesn't seem to be a real danger of callout loops,
2620 since a callout by the remote host to check the sender would use <>.
2621 [SEE ABOVE: changed after hitting problems.]
2623 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2624 temporary errors. However, in the case of such a code being given after
2625 the end of a data transmission (i.e. after ".") Exim was failing to write
2626 a retry record for the message. (Yes, there was some broken host that was
2627 actually sending 8xx at this point.)
2629 73. An unknown lookup type in a host list could cause Exim to panic-die when
2630 the list was checked. (An example that provoked this was putting <; in the
2631 middle of a list instead of at the start.) If this happened during a DATA
2632 ACL check, a -D file could be left lying around. This kind of configuration
2633 error no longer causes Exim to die; instead it causes a defer errror. The
2634 incident is still logged to the main and panic logs.
2636 74. Buglet left over from Exim 3 conversion. The message "too many messages
2637 in one connection" was written to the rejectlog but not the mainlog, except
2638 when address rewriting (yes!) was being logged.
2640 75. Added write_rejectlog option.
2642 76. When a system filter was run not as root (that is, when system_filter_user
2643 was set), the values of the $n variables were not being returned to the
2644 main process; thus, they were not subsequently available in the $sn
2647 77. Added +return_path_on_delivery log selector.
2649 78. A connection timeout was being treated differently from recipients deferred
2650 when testing hosts_max_try with a message that was older than the host's
2651 retry timeout. (The host should not be counted, thus allowing all hosts to
2652 be tried at least once before bouncing.) This may have been the cause of an
2653 occasionally reported bug whereby a message would remain on the queue
2654 longer than the retry timeout, but would be bounced if a delivery was
2655 forced. I say "may" because I never totally pinned down the problem;
2656 setting up timeout/retry tests is difficult. See also the next item.
2658 79. The ultimate address timeout was not being applied to errors that involved
2659 a combination of host plus message (for example, a timeout on a MAIL
2660 command). When an address resolved to a number of possible hosts, and they
2661 were not all tried for each delivery (e.g. because of hosts_max_try), a
2662 message could remain on the queue longer than the retry timeout.
2664 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2667 81. Fixed an obscure SMTP outgoing bug which required at least the following
2668 conditions: (a) there was another message waiting for the same server;
2669 (b) the server returned 5xx to all RCPT commands in the first message so
2670 that the message was not completed; (c) the server dropped the connection
2671 or gave a negative response to the RSET that Exim sends to abort the
2672 transaction. The observed case was a dropped connection after DATA that had
2673 been sent in pipelining mode. That is, the server had advertised PIPELINING
2674 but was not implementing it correctly. The effect of the bug was incorrect
2675 behaviour, such as trying another host, and this could lead to a crash.
2681 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2682 and daemon.c were passed as pointers to ints; they should have been
2683 pointers to socklen_t variables (which are typically unsigned ints).
2685 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2688 3. Fixed a really odd bug that affected only the testing scheme; patching a
2689 certain fixed string in the binary changed the value of another string that
2690 happened to be identical to the end of the original first string.
2692 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2693 name", it returns that address as the IP address. On some operating
2694 systems (e.g. Solaris), it also passes back the IP address string as the
2695 "host name". However, on others (e.g. Linux), it passes back an empty
2696 string. Exim wasn't checking for this, and was changing the host name to an
2697 empty string, assuming it had been canonicized.
2699 5. Although rare, it is permitted to have more than one PTR record for a given
2700 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2701 all the names associated with an address, because they do in Solaris.
2702 However, it seems that they do not in Linux for data that comes from the
2703 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2704 I found this out when I moved to a new Linux workstation and tried to run
2705 the Exim test suite.
2707 To get round this problem I have changed the code so that it now does its
2708 own call to the DNS to look up PTR records when searching for a host name.
2709 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2710 addresses that are only in /etc/hosts are still found.
2712 This behaviour is, however, controlled by an option called host_lookup_
2713 order, which defaults to "bydns:byaddr". If people want to use the other
2714 order, or indeed, just use one or the other means of lookup, they can
2715 specify it in this variable.
2717 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2718 some operating systems, this comes back from gethostbyaddr() as an empty
2719 string, and this is what Exim used to test for. However, it seems that in
2720 other systems, "." is yielded. Exim now tests for this case too.
2722 7. The values of check_spool_space and check_log_space are now held internally
2723 as a number of kilobytes instead of an absolute number of bytes. If a
2724 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2725 kilobyte. This means that much larger values can be stored.
2727 8. Exim monitor: an attempt to get the action menu when not actually pointing
2728 at a message produces an empty menu entitled "No message selected". This
2729 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2730 no entries in it ("Shell widget menu has zero width and/or height"). So I
2731 have added a single, blank menu entry in this case.
2733 9. Added ${quote_local_part.
2735 10. MIME decoding is now applied to the contents of Subject: header lines when
2738 11. Now that a reference to $sender_host_address automatically causes a reverse
2739 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2740 host lookup before query-style lookups in lists that might use this
2741 variable. This has therefore been abolished, and the "net-" prefix is no
2742 longer necessary for query-style lookups.
2744 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2745 have been a typo for LFLAGS, so it has been changed.
2747 13. The install script calls Exim with "-C /dev/null" in order to find the
2748 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2749 to be output. Howeve, since Exim outputs its version number before the
2750 error, it didn't break the script. It just looked ugly. I fixed this by
2751 always allowing "-C /dev/null" if the caller is root.
2753 14. Ignore overlarge ACL variable number when reading spool file - insurance
2754 against a later release with more variables having written the file.
2756 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2757 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2758 this, as well as the form without the "IPv6" on the front.
2760 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2761 OS/Makefile-Darwin file.
2763 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2764 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2766 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2767 variable. However, some OS use ints for inodes. I've added cast to long int
2768 to get rid of the compiler warning.
2770 19. I had forgotten to lock out "/../" in configuration file names when
2771 ALT_CONFIG_PREFIX was set.
2773 20. Routers used for verification do not need to specify transports. However,
2774 if such a router generated a host list, and callout was configured, Exim
2775 crashed, because it could not find a port number from the (non-existent)
2776 transport. It now assumes port 25 in this circumstance.
2778 21. Added the -t option to exigrep.
2780 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2781 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2782 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2783 them is set, LOOKUP_LSEARCH is forced.
2785 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2787 24. Added sender and host information to the "rejected by local_scan()" log
2788 line; previously there was no indication of these.
2790 25. Added .include_if_exists.
2792 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2793 Linux. It turns out that not all file systems support this. Apparently some
2794 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2795 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2796 not supported on the file descriptor, is now ignored when Exim is trying to
2797 sync a directory. This applies only to Linux.
2799 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2801 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2802 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2805 29. Added support for maildirsize files from supplied patch (modified a bit).
2807 30. The use of :fail: followed by an empty string could lead Exim to respond to
2808 sender verification failures with (e.g.):
2810 550 Verification failed for <xxx>
2811 550 Sender verify failed
2813 where the first response line was missing the '-' that indicates it is not
2814 the final line of the response.
2816 31. The loop for finding the name of the user that called Exim had a hardwired
2817 limit of 10; it now uses the value of finduser_retries, which is used for
2818 all other user lookups.
2820 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2823 33. Exim was neglecting to zero errno before one call of strtol() when
2824 expanding a string and expecting an integer value. On some systems this
2825 resulted in spurious "integer overflow" errors. Also, it was casting the
2826 result into an int without checking.
2828 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2829 did not work. The code looks as if it has *never* worked, though it appears
2830 to have been documented since at least releast 1.62. I have made it work.
2832 35. The "timeout_DNS" error in retry rules, also documented since at least
2833 1.62, also never worked. As it isn't clear exactly what this means, and
2834 clearly it isn't a major issue, I have abolished the feature by treating it
2835 as "timeout", and writing a warning to the main and panic logs.
2837 36. The display of retry rules for -brt wasn't always showing the error code
2840 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2841 timeout_connect_A, timeout_connect_MX.
2843 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2844 to the empty sender.
2846 39. The daemon was not analysing the content of -oX till after it had closed
2847 stderr and disconnected from the controlling terminal. This meant that any
2848 syntax errors were only noted on the panic log, and the return code from
2849 the command was 0. By re-arranging the code a little, I've made the
2850 decoding happen first, so such errors now appear on stderr, and the return
2851 code is 1. However, the actual setting up of the sockets still happens in
2852 the disconnected process, so errors there are still only recorded on the
2855 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2856 connections (as happens on some IP stacks) was logged at start up time as
2857 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2858 it from "IPv6 and IPv4", which means that two separate sockets are being
2861 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2862 says whether AF_INET or AF_INET6 was passed as an argument.
2864 42. Exiwhat output was messed up when time zones were included in log
2867 43. Exiwhat now gives more information about the daemon's listening ports,
2868 and whether -tls-on-connect was used.
2870 44. The "port" option of the smtp transport is now expanded.
2872 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2873 silently ignored. Now an error message is written to the main and panic
2876 46. There's a new ACL modifier called "logwrite" which writes to a log file
2877 as soon as it is encountered.
2879 47. Added $local_user_uid and $local_user_gid at routing time.
2881 48. Exim crashed when trying to verify a sender address that was being
2884 49. Exim was recognizing only a space character after ".include". It now also
2885 recognizes a tab character.
2887 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2888 extracting the relevant information from the specification. The man page no
2889 longer contains scrambled data for the -d option, and I've added a section
2890 at the front about calling Exim under different names.
2892 51. Added "extra_headers" argument to the "mail" command in filter files.
2894 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2897 53. Installed eximstats 1.29.
2899 54. Added transport_filter_timeout as a generic transport option.
2901 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2902 Cc: header lines. This was required by RFC 822, but it not required by RFC
2905 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2906 incoming messages that did not have them. Now it does so only if the
2907 message originates locally, that is, if there is no associated remote host
2908 address. When Resent- header lines are present, this applies to the Resent-
2909 lines rather than the non-Resent- lines.
2911 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2912 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2914 58. Messages for configuration errors now include the name of the main
2915 configuration file - useful now that there may be more than one file in a
2916 list (.included file names were always shown).
2918 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2919 for those rare installations that do not start the daemon as root or run it
2920 setuid root. I've cut out the call to initgroups() if the daemon is not
2923 60. The Exim user and group can now be bound into the binary as text strings
2924 that are looked up at the start of Exim's processing.
2926 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2928 62. Added $mailstore_basename variable.
2930 63. Installed patch to sieve.c from Michael Haardt.
2932 64. When Exim failed to open the panic log after failing to open the main log,
2933 the original message it was trying to log was written to stderr and debug
2934 output, but if they were not available (the usual case in production), it
2935 was lost. Now it is written to syslog before the two lines that record the
2936 failures to open the logs.
2938 65. Users' Exim filters run in subprocesses under the user's uid. It is
2939 possible for a "deliver" command or an alias in a "personal" command to
2940 provoke an address rewrite. If logging of address rewriting is configured,
2941 this fails because the process is not running as root or exim. There may be
2942 a better way of dealing with this, but for the moment (because 4.30 needs
2943 to be released), I have disabled address rewrite logging when running a
2944 filter in a non-root, non-exim process.
2950 1. The buildconfig auxiliary program wasn't quoting the value set for
2951 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2952 not defined. This bug was masked in 4.22 by the effect that was fixed in
2955 2. Some messages that were rejected after a message id was allocated were
2956 shown as "incomplete" by exigrep. It no longer does this for messages that
2957 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2959 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2960 have allow_domain_literals set, the ID did not get logged in the <= line.
2961 Domain literals are now always recognized in Message-ID: header lines.
2963 4. The first argument for a ${extract expansion item is the key name or field
2964 number. Leading and trailing spaces in this item were not being ignored,
2965 causing some misleading effects.
2967 5. When deliver_drop_privilege was set, single queue runner processes started
2968 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2969 same command in the process it spins off) were not dropping privilege.
2971 6. When the daemon running as "exim" started a queue runner, it always
2972 re-executed Exim in the spun-off process. This is a waste of effort when
2973 deliver_drop_privilege is set. The new process now just calls the
2974 queue-runner function directly.
2980 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2983 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2984 ignored. Though the use of -oP was forcing the writing of a pid file, it
2985 was always written to the default place.
2987 3. If the message "no IP address found for host xxxx" is generated during
2988 incoming verification, it is now followed by identification of the incoming
2989 connection (so you can more easily find what provoked it).
2991 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2993 5. Added some features to "harden" Exim a bit more against certain attacks:
2995 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2996 be put in Local/Makefile. This is like the never_users runtime option,
2997 but it cannot be overridden. The default setting is "root".
2999 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3000 prefix string with which any file named in a -C command line option
3003 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3004 is retained for -C and -D only if the caller of Exim is root. Without
3005 it, the exim user may also use -C and -D and retain privilege.
3007 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3008 command line option is disabled.
3010 6. Macro names set by the -D option must start with an upper case letter, just
3011 like macro names defined in the configuration file.
3013 7. Added "dereference=" facility to LDAP.
3015 8. Two instances of the typo "uknown" in the source files are fixed.
3017 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3018 the Configure-Makefile script screwed up while processing it.
3020 10. Incorporated PCRE 4.4.
3022 11. The SMTP synchronization check was not operating right at the start of an
3023 SMTP session. For example, it could not catch a HELO sent before the client
3024 waited for the greeting. There is now a check for outstanding input at the
3025 point when the greeting is written. Because of the duplex, asynchronous
3026 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3027 way, but not yet received, when the check is performed.
3029 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3030 on TCP/IP sockets, because this apparently causes some broken clients to
3033 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3034 unchanged) from the Cygwin maintainer.
3036 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3037 was supported instead of testing for mbx. Effectively a typo.
3039 15. The spa authenticator server code was not checking that the input it
3040 received was valid base64.
3042 16. The debug output line for the "set" modifier in ACLs was not showing the
3043 name of the variable that was being set.
3045 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3047 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3048 lookup on its own. Something else was needed to trigger the lookup. For
3049 example, a match in host_lookup or the need for a host name in a host list.
3050 Now, if $sender_host_name is referenced and the host name has not yet been
3051 looked up, a lookup is performed. If the lookup fails, the variable remains
3052 empty, and $host_lookup_failed is set to "1".
3054 19. Added "eqi" as a case-independent comparison operator.
3056 20. The saslauthd authentication condition could segfault if neither service
3057 nor realm was specified.
3059 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3060 error message that was logged was misleading, and incoming SMTP
3061 connections were dropped. The message is now more accurate, and temporary
3062 errors are given to SMTP connections.
3064 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3065 (except RSET) with a 421 error, until QUIT is received. However, it was
3066 failing to send a response to QUIT.
3068 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3069 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3070 ACL is now run after the helo_try_verify_hosts code.
3072 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3073 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3074 case-independent, so other case variants are also recognized). Apparently
3075 some systems use these upper case variants.
3077 25. If more than two messages were waiting for the same host, and a transport
3078 filter was specified for the transport, Exim sent two messages over the
3079 same TCP/IP connection, and then failed with "socket operation on non-
3080 socket" when it tried to send the third.
3082 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3084 27. The extern definition of crypt16() in expand.c was not being excluded when
3085 the OS had its own crypt16() function.
3087 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3088 as a preferred synonym for return_size_limit, both as an option and as an
3091 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3093 30. Changed the default configuration ACL to relax the local part checking rule
3094 for addresses that are not in any local domains. For these addresses,
3095 slashes and pipe symbols are allowed within local parts, but the sequence
3096 /../ is explicitly forbidden.
3098 31. SPA server authentication was not clearing the challenge buffer before
3101 32. log_message in a "warn" ACL statement was writing to the reject log as
3102 well as to the main log, which contradicts the documentation and doesn't
3103 seem right (because no rejection is happening). So I have stopped it.
3105 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3106 However, I am unable to do any testing of this.
3108 34. Fixed an infelicity in the appendfile transport. When checking directories
3109 for a mailbox, to see if any needed to be created, it was accidentally
3110 using path names with one or more superfluous leading slashes; tracing
3111 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3113 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3114 contents are added to the log line that is written for every discarded
3115 recipient. (Previously a log_message setting was ignored.)
3117 36. The ${quote: operator now quotes the string if it is empty.
3119 37. The install script runs exim in order to find its version number. If for
3120 some reason other than non-existence or emptiness, which it checks, it
3121 could not run './exim', it was installing it with an empty version number,
3122 i.e. as "exim-". This error state is now caught, and the installation is
3125 38. An argument was missing from the function that creates an error message
3126 when Exim fails to connect to the socket for saslauthd authentication.
3127 This could cause Exim to crash, or give a corrupted message.
3129 39. Added isip, isip4, and isip6 to ${if conditions.
3131 40. The ACL variables $acl_xx are now saved with the message, and can be
3132 accessed later in routers, transports, and filters.
3134 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3135 strings in the file are not string-expanded.
3137 42. If a MAIL command specified a SIZE value that was too large to fit into an
3138 int variable, the check against message_size_limit failed. Such values are
3139 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3140 day this will have to be increased, but I don't think I want to be around
3141 when emails are that large.
3148 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3149 iconv() is not standard in FreeBSD.
3151 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3152 IPv6 enabled. The observed symptom was a segmentation fault on return from
3153 the function os_common_find_running_interfaces() in src/os.c.
3155 3. In the check_special_case() function in daemon.c I had used "errno" as an
3156 argument name, which causes warnings on some systems. This was basically a
3157 typo, since it was named "eno" in the comments!
3159 4. The code that waits for the clock to tick (at a resolution of some fraction
3160 of a second) so as to ensure message-id uniqueness was always waiting for
3161 at least one whole tick, when it could have waited for less. [This is
3162 almost certainly not relevant at current processor speeds, where it is
3163 unlikely to ever wait at all. But we try to future-proof.]
3165 5. The function that sleeps for a time interval that includes fractions of a
3166 second contained a race. It did not block SIGALRM between setting the
3167 timer, and suspending (a couple of lines later). If the interval was short
3168 and the sigsuspend() was delayed until after it had expired, the suspension
3169 never ended. On busy systems this could lead to processes getting stuck for
3172 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3173 process, before it forks any delivery processes. The open lookup caching
3174 mechanism meant that the open file or database connection was passed into
3175 the delivery process. The problem was that delivery processes always tidy
3176 up cached lookup data. This could cause a problem for the next delivery
3177 process started by the queue runner, because the external queue runner
3178 process does not know about the closure. So the next delivery process
3179 still has data in the lookup cache. In the case of a file lookup, there was
3180 no problem because closing a file descriptor in a subprocess doesn't affect
3181 the parent. However, if the lookup was caching a connection to a database,
3182 the connection was closed, and the second delivery process was likely to
3183 see errors such as "PGSQL: query failed: server closed the connection
3184 unexpectedly". The problem has been fixed by closing all cached lookups
3185 in a queue runner before running a delivery process.
3187 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3188 seem to have the "const" qualifier which it has on other OS. I've
3191 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3192 *of the same type* (client or server) with the same public name that an
3193 error should be diagnosed.
3195 9. When Exim looked up a host name for an IP address, but failed to find the
3196 original IP address when looking up the host name (a safety check), it
3197 output the message "<ip address> does not match any IP for NULL", which was
3198 confusing, to say the least. The bug was that the host name should have
3199 appeared instead of "NULL".
3201 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3202 user that is built into the binary, and the -C or -D options is used, root
3203 privilege is dropped before the configuration file is read. In addition,
3204 logging is switched to stderr instead of the normal log files. If the
3205 configuration then re-defines the Exim user, the unprivileged environment
3206 is probably not what is expected, so Exim logs a panic warning message (but
3209 However, if deliver_drop_privilege is set, the unprivileged state may well
3210 be exactly what is intended, so the warning has been cut out in that case,
3211 and Exim is allowed to try to write to its normal log files.
3217 1. smtp_return_error_details was not giving details for temporary sender
3218 or receiver verification errors.
3220 2. Diagnose a configuration error if two authenticators have the same public
3223 3. Exim used not to create the message log file for a message until the first
3224 delivery attempt. This could be confusing when incoming messages were held
3225 for policy or load reasons. The message log file is now created at the time
3226 the message is received, and an initial "Received" line is written to it.
3228 4. The automatically generated man page for command line options had a minor
3229 bug that caused no ill effects; however, a more serious problem was that
3230 the procedure for building the man page automatically didn't always
3231 operate. Consequently, release 4.20 contains an out-of-date version. This
3232 shouldn't happen again.
3234 5. When building Exim with embedded Perl support, the script that builds the
3235 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3236 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3238 6. The freeze_tell option was not being used for messages that were frozen on
3239 arrival, either by an ACL or by local_scan().
3241 7. Added the smtp_incomplete_transaction log selector.
3243 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3244 was accepting AUTH without a new EHLO.
3246 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3247 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3250 10. Logging of TCP/IP connections (when configured) now happens in the main
3251 daemon process instead of the child process, so that the TCP/IP connection
3252 count is more accurate (but it can never be perfect).
3254 11. The use of "drop" in a nested ACL was not being handled correctly in the
3255 outer ACL. Now, if condition failure induced by the nested "drop" causes
3256 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3257 or "require"), the connection is dropped.
3259 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3260 that yield "discard" can now be used with an "accept" or a "discard" verb,
3261 but an error is generated for any others (because I can't see a useful way
3262 to define what should happen).
3264 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3265 are now processed in the same way as lines in the Exim configuration file.
3266 In particular, continuation lines are supported.
3268 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3270 15. Added -ti meaning -t -i.
3272 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3273 domains, and warn by logging if others are found.
3275 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3276 returned from GIFCONF when Exim is trying to find the list of interfaces on
3277 a host. The code in os.c has been modified to copy each ifreq to an aligned
3278 structure in all cases.
3280 Also, in some cases, the returned ifreq's were being copied to a 'struct
3281 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3282 means the last couple of bytes of an IPv6 address could be chopped if the
3283 ifreq contained only a normal sockaddr (14 bytes storage).
3285 18. Named domain lists were not supported in the hosts_treat_as_local option.
3286 An entry such as +xxxx was not recognized, and was treated as a literal
3289 19. Ensure that header lines added by a DATA ACL are included in the reject log
3290 if the ACL subsequently rejects the message.
3292 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3293 MD5 (which is deprecated).
3295 21. When testing a filter file using -bf, Exim was writing a message when it
3296 took the sender from a "From " line in the message, but it was not doing so
3297 when it took $return_path from a Return-Path: header line. It now does.
3299 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3300 with a valid header line field name (a series of printing characters
3301 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3303 23. Changed "disc" in the source to "disk" to conform to the documentation and
3304 the book and for uniformity.
3306 24. Ignore Sendmail's -Ooption=value command line item.
3308 25. When execve() failed while trying to run a command in a pipe transport,
3309 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3310 could be confused with a return value of 69 from the command itself. This
3311 has been changed to 127, the value the shell returns if it is asked to run
3312 a non-existent command. The wording for the related log line suggests a
3313 non-existent command as the problem.
3315 26. If received_header_text expands to an empty string, do not add a Received:
3316 header line to the message. (Well, it adds a token one on the spool, but
3317 marks it "old" so that it doesn't get used or transmitted.)
3319 27. Installed eximstats 1.28 (addition of -nt option).
3321 28. There was no check for failure on the call to getsockname() in the daemon
3322 code. This can fail if there is a shortage of resources on the system, with
3323 ENOMEM, for example. A temporary error is now given on failure.
3325 29. Contrary to the C standard, it seems that in some environments, the
3326 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3327 program. Exim now does this explicitly; it affects the formatting of
3328 timestamps using strftime().
3330 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3331 complaints. I've now initialized all the variables, to avoid this.
3333 32. Header lines added by a system filter were not being "seen" during
3334 transport-time rewrites.
3336 33. The info_callback() function passed to OpenSSL is set up with type void
3337 (*)(SSL *, int, int), as described somewhere. However, when calling the
3338 function (actually a macro) that sets it up, the type void(*)() is
3339 expected. I've put in a cast to prevent warnings from picky compilers.
3341 34. If a DNS black list lookup found a CNAME record, but there were no A
3342 records associated with the domain it pointed at, Exim crashed.
3344 35. If a DNS black list lookup returned more than one A record, Exim ignored
3345 all but the first. It now scans all returned addresses if a particular IP
3346 value is being sought. In this situation, the contents of the
3347 $dnslist_value variable are a list of all the addresses, separated by a
3350 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3351 to accept a host name and all its aliases if the forward lookup for any of
3352 them yielded the IP address of the incoming connection. Now it accepts only
3353 those names whose forward lookup yields the correct IP address. Any other
3354 names are discarded. This closes a loophole whereby a rogue DNS
3355 administrator could create reverse DNS records to break through a
3356 wildcarded host restriction in an ACL.
3358 37. If a user filter or a system filter that ran in a subprocess used any of
3359 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3360 the wrong values were passed to the pipe command ($thisaddress had the
3361 value of $0, $0 had the value of $1, etc). This bug was introduced by
3362 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3364 38. Improved the line breaking for long SMTP error messages from ACLs.
3365 Previously, if there was no break point between 40 and 75 characters, Exim
3366 left the rest of the message alone. Two changes have been made: (a) I've
3367 reduced the minimum length to 35 characters; (b) if it can't find a break
3368 point between 35 and 75 characters, it looks ahead and uses the first one
3369 that it finds. This may give the occasional overlong line, but at least the
3370 remaining text gets split now.
3372 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3373 file descriptors might be low, and that setting 1000 would always raise it.
3374 It turns out that in some environments, the limit is already over 1000 and
3375 that lowering it causes trouble. So now Exim takes care not to decrease it.
3377 40. When delivering a message, the value of $return_path is set to $sender_
3378 address at the start of routing (routers may change the value). By an
3379 oversight, this default was not being set up when an address was tested by
3380 -bt or -bv, which affected the outcome if any router or filter referred to
3383 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3384 log only when "message" or "log_message" are set. However, if one of the
3385 conditions was an address verification, or a call to a nested ACL, the
3386 messages generated by the underlying test were being passed through. This
3387 no longer happens. The underlying message is available in $acl_verify_
3388 message for both "message" and "log_message" expansions, so it can be
3389 passed through if needed.
3391 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3392 new expansion $bh_ to give the encoded byte string without charset
3393 translation. Translation happens only if iconv() is available; HAVE_ICONV
3394 indicates this at build time. HEADERS_CHARSET gives the charset to
3395 translate to; headers_charset can change it in the configuration, and
3396 "headers charset" can change it in an individual filter file.
3398 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3399 that creates RFC 2047 encoded "words" labels them as that charset instead
3400 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3401 expansion operator; (ii) when Exim creates a From: line for a local
3402 message; (iii) when a header line is rewritten to include a "phrase" part.
3404 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3405 buggy, causing it to skip the first lines of messages whose message ID
3406 ended in 'D'. This would not have bitten before Exim release 4.14, because
3407 message IDs were unlikely to end in 'D' before then. The effect was to have
3408 incorrect size information for certain domains.
3410 45. #include "config.h" was missing at the start of the crypt16.c module. This
3411 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3414 46. If there was a timeout during a "random" callout check, Exim treated it as
3415 a failure of the random address, and carried on sending RSET and the real
3416 address. If the delay was just some slowness somewhere, the response to the
3417 original RCPT would be taken as a response to RSET and so on, causing
3418 mayhem of various kinds.
3420 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3421 when I implemented it. It didn't allow for the fact that some option values
3422 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3423 the right test for positive values.
3425 48. Domain names in DNS records are case-independent. Exim always looks them up
3426 in lower case. Some resolvers return domain names in exactly the case they
3427 appear in the zone file, that is, they may contain uppercase letters. Not
3428 all resolvers do this - some return always lower case. Exim was treating a
3429 change of case by a resolver as a change of domain, similar to a widening
3430 of a domain abbreviation. This triggered its re-routing code and so it was
3431 trying to route what was effectively the same domain again. This normally
3432 caused routing to fail (because the router wouldn't handle the domain
3433 twice). Now Exim checks for this case specially, and just changes the
3434 casing of the domain that it ultimately uses when it transmits the message
3437 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3440 50. If a filter generated a file delivery with a non-absolute name (possible if
3441 no home directory exists for the router), the forbid_file option was not
3444 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3445 the existing equality matching.
3447 52. Exim was using ints instead of ino_t variables in some places where it was
3448 dealing with inode numbers.
3450 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3451 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3452 called TMPDIR, and if it finds it is different, it changes its value.
3454 54. The smtp_printf() function is now made available to local_scan() so
3455 additional output lines can be written before returning. There is also an
3456 smtp_fflush() function to enable the detection of a dropped connection.
3457 The variables smtp_input and smtp_batched_input are exported to
3460 55. Changed the default runtime configuration: the message "Unknown user"
3461 has been removed from the ACL, and instead placed on the localuser router,
3462 using the cannot_route_message feature. This means that any verification
3463 failures that generate their own messages won't get overridden. Similarly,
3464 the "Unrouteable address" message that was in the ACL for unverifiable
3465 relay addresses has also been removed.
3467 56. Added hosts_avoid_esmtp to the smtp transport.
3469 57. The exicyclog script was not checking for the esoteric option
3470 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3471 will work only if exicyclog is run under the appropriate euid.
3473 58. Following a discussion on the list, the rules by which Exim recognises line
3474 endings on incoming messages have been changed. The -dropcr and drop_cr
3475 options are now no-ops, retained only for backwards compatibility. The
3476 following line terminators are recognized: LF CRLF CR. However, special
3477 processing applies to CR:
3479 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3480 nor a local message in the state where . is a terminator.
3482 (ii) If a bare CR is encountered in a header line, an extra space is added
3483 after the line terminator so as not to end the header. The reasoning
3484 behind this is that bare CRs in header lines are most likely either
3485 to be mistakes, or people trying to play silly games.
3487 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3488 was being incorrectly given as 18 bytes larger than it should have been.
3489 This is a VOB (very old bug).
3491 60. This may never have affected anything current, but just in case it has:
3492 When the local host is found other than at the start of a list of hosts,
3493 the local host, those with the same MX, and any that follow, are discarded.
3494 When the list in question was part of a longer list of hosts, the following
3495 hosts (not currently being processed) were also being discarded. This no
3496 longer happens. I'm not sure if this situation could ever has previously
3499 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3502 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3503 header. This is something that is recommended in a new Internet Draft, and
3504 is something that is documented as being done by Sendmail. There are two
3505 possible values. For messages generated by the autoreply transport, Exim
3508 Auto-Submitted: auto-replied
3510 whereas for all other generated messages (e.g. bounces) it adds
3512 Auto-Submitted: auto-generated
3514 63. The "personal" condition in filters now includes a test for the
3515 Auto-Submitted: header. If it contains the string "auto-" the message it
3516 not considered personal.
3518 64. Added rcpt_include_affixes as a generic transport option.
3520 65. Added queue_only_override (default true).
3522 66. Added the syslog_duplication option.
3524 67. If what should have been the first header line of a message consisted of
3525 a space followed by a colon, Exim was mis-interpreting it as a header line.
3526 It isn't of course - it is syntactically invalid and should therefore be
3527 treated as the start of the message body. The misbehaviour could have
3528 caused a number of strange effects, including loss of data in subsequent
3529 header lines, and spool format errors.
3531 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3532 client host had authenticated. This control can now be exercised by an ACL
3533 for more flexibility.
3535 69. By default, callouts do not happen when testing with -bh. There is now a
3536 variant, -bhc, which does actually run the callout code, including
3537 consulting and updating the callout cache.
3539 70. Added support for saslauthd authentication, courtesy of Alexander
3542 71. If statvfs() failed on the spool or log directories while checking their
3543 size for availability, Exim confusingly gave the error "space shortage".
3544 Furthermore, in debugging mode it crashed with a floating point exception.
3545 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3546 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3547 serious problem, Exim now writes to the main and panic logs when this
3548 happens, with details of the failure. It then refuses to accept the
3549 incoming message, giving the message "spool directory problem" or "log
3550 directory problem" with a 421 code for SMTP messages.
3552 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3553 0, 1, and 2 exist, because some OS complain for execs without them (see
3554 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3555 descriptors. However, the code omitted to check that the open succeeded,
3556 causing mysterious errors if for some reason the permissions on /dev/null
3557 got screwed. Now Exim writes a message to the main and panic logs, and
3558 bombs out if it can't open /dev/null.
3560 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3561 interact so that it is all more flexible. It is supposed to remain
3562 backwards compatible. Also added extra_local_interfaces.
3564 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3565 to bomb out with an assertion failure - to the client this appears as a
3566 connection drop. This problem occurs in the part of the code that was taken
3567 from the Samba project. Fortunately, the assertion is in a very simple
3568 function, so I have fixed this by reproducing the function inline in the
3569 one place where it is called, and arranging for authentication to fail
3570 instead of killing the process with assert().
3572 75. The SPA client code was not working when the server requested OEM rather
3573 than Unicode encoding.
3575 76. Added code to make require_files with a specific uid setting more usable in
3576 the case where statting the file as root fails - usually a non-root-mounted
3577 NFS file system. When this happens and the failure is EACCES, Exim now
3578 forks a subprocess and does the per-uid checking as the relevant uid.
3580 77. Added process_log_path.
3582 78. If log_file_path was not explicitly set, a setting of check_log_space or
3583 check_log_inodes was ignored.
3585 79. If a space check for the spool or log partitions fails, the incident is now
3586 logged. Of course, in the latter case the data may get lost...
3588 80. Added the %p formatting code to string_format() so that it can be used to
3589 print addresses in debug_print(). Adjusted all the address printing in the
3590 debugging in store.c to use %p rather than %d.
3592 81. There was a concern that a line of code in smtp_in.c could overflow a
3593 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3594 initially expressed, the concern was not well-founded, because trailing
3595 spaces are removed early. However, if the trailing spaces were followed by
3596 a NULL, they did not get removed, so the overflow was possible. Two fixes
3599 (a) I re-wrote the offending code in a cleaner fashion.
3600 (b) If an incoming SMTP command contains a NULL character, it is rejected
3603 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3604 runs initgroups(), so that if the Exim user is in any additional groups,
3605 they will be used during message reception.
3611 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob