1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.277 2005/12/15 15:44:46 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
20 PH/03 Added disable_ipv6.
22 PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
29 PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
32 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
38 PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
43 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
46 PH/09 Moved a debug statement in filter processing to avoid a race problem when
53 PH/01 Two changes to the default runtime configuration:
55 (1) Move the checks for relay_from_hosts and authenticated clients from
56 after to before the (commented out) DNS black list checks.
58 (2) Add control=submission to the relay_from_hosts and authenticated
59 clients checks, on the grounds that messages accepted by these
60 statements are most likely to be submissions.
62 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
64 (1) Generate an error if the third argument for the ${prvs expansion is
67 (2) Treat a missing third argument of ${prvscheck as if it were an empty
70 (3) Reset the variables that are obtained from the first argument of
71 ${prvscheck and used in the second argument before leaving the code,
72 because their memory is reclaimed, so using them afterwards may do
75 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
76 one (it's much easier than Tom thought :-).
78 (5) Because of (4), we can now allow for the use of $prvscheck_result
79 inside the third argument.
81 PH/03 For some reason, the default setting of PATH when running a command from
82 a pipe transport was just "/usr/bin". I have changed it to
85 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
86 anything to be listed in the output from -bV.
88 PH/05 When a filter generated an autoreply, the entire To: header line was
89 quoted in the delivery log line, like this:
91 => >A.N.Other <ano@some.domain> <original@ddress> ...
93 This has been changed so that it extracts the operative address. There
94 may be more than one such address. If so, they are comma-separated, like
97 => >ano@some.domain,ona@other.domain <original@ddress> ...
99 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
100 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
101 not being looked up in the rDNS to get a host name, Exim was showing the
102 IP address twice in Received: lines, even though the IP addresses were
103 identical. For example:
105 Received: from [1.2.3.4] (helo=[1.2.3.4])
107 However, if the real host name was known, it was omitting the HELO data
108 if it matched the actual IP address. This has been tidied up so that it
109 doesn't show the same IP address twice.
111 PH/07 When both +timestamp and +memory debugging was on, the value given by
112 $tod_xxx expansions could be wrong, because the tod_stamp() function was
113 called by the debug printing, thereby overwriting the timestamp buffer.
114 Debugging no longer uses the tod_stamp() function when +timestamp is set.
116 PH/08 When the original message was included in an autoreply transport, it
117 always said "this is a copy of the message, including all the headers",
118 even if body_only or headers_only was set. It now gives an appropriate
121 PH/09 Applied a patch from the Sieve maintainer which:
123 o fixes some comments
124 o adds the (disabled) notify extension core
125 o adds some debug output for the result of if/elsif tests
126 o points to the current vacation draft in the documentation
127 and documents the missing references header update
131 o fixes a bug in processing the envelope test (when testing
132 multiple envelope elements, the last element determinted the
135 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
136 Electronic Mail") by including:
138 Auto-submitted: auto-generated
140 in the messages that it generates (bounce messages and others, such as
141 warnings). In the case of bounce messages for non-SMTP mesages, there was
142 also a typo: it was using "Auto_submitted" (underscore instead of
143 hyphen). Since every message generated by Exim is necessarily in response
144 to another message, thes have all been changed to:
146 Auto-Submitted: auto-replied
148 in accordance with these statements in the RFC:
150 The auto-replied keyword:
152 - SHOULD be used on messages sent in direct response to another
153 message by an automatic process,
155 - MUST NOT be used on manually-generated messages,
157 - MAY be used on Delivery Status Notifications (DSNs) and Message
158 Disposition Notifications (MDNs),
160 - MUST NOT be used on messages generated by automatic or periodic
161 processes, except for messages which are automatic responses to
164 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
165 to the default Received: header definition.
167 PH/12 Added log selector acl_warn_skipped (default on).
169 PH/13 After a successful wildlsearch lookup, discard the values of numeric
170 variables because (a) they are in the wrong storage pool and (b) even if
171 they were copied, it wouldn't work properly because of the caching.
173 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
174 checking when decoding. Apparently there are clients that generate
175 overlong encoded strings. Why am I not surprised?
177 PH/15 If the first argument of "${if match_address" was not empty, but did not
178 contain an "@" character, Exim crashed. Now it writes a panic log message
179 and treats the condition as false.
181 PH/16 In autoreply, treat an empty string for "once" the same as unset.
183 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
184 extension "envelope-auth". The code is finished and in agreement with
185 other implementations, but there is no documentation so far and in fact,
186 nobody wrote the draft yet. This extension is currently #undef'ed, thus
187 not changing the active code.
189 Print executed "if" and "elsif" statements when debugging is used. This
190 helps a great deal to understand what a filter does.
192 Document more things not specified clearly in RFC3028. I had all this
193 sorted out, when out of a sudden new issues came to my mind. Oops."
195 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
198 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
199 canonical form (as documented). However, after a host name lookup from
200 the IP address, check_host() was doing a simple string comparison with
201 addresses acquired from the DNS when checking that the found name did
202 have the original IP as one of its addresses. Since any found IPv6
203 addresses are likely to be in abbreviated form, the comparison could
204 fail. Luckily, there already exists a function for doing the comparison
205 by converting both addresses to binary, so now that is used instead of
208 PH/20 There was another similar case to PH/19, when a complete host name was
209 given in a host list; looking up its IP address could give an abbreviated
210 form, whereas the current host's name might or might not be abbreviated.
211 The same fix has been applied.
217 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
218 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
221 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
222 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
224 PH/03 Typo: missing ".o" in src/pcre/Makefile.
226 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
227 header line, restrict the check to what is listed in RFCs 2369 and 2929.
228 Also, for "Auto-Submitted", treat anything other than "no" as
229 non-personal, in accordance with RFC 3834. (Previously it treated
230 anything starting "auto-" as non-personal.)
232 TF/01 The control=submission/name=... option had a problem with syntax
233 errors if the name included a slash character. The /name= option
234 now slurps the rest of the string, so it can include any characters
235 but it must come last in the list of options (after /sender_retain
238 PH/05 Some modifications to the interface to the fake nameserver for the new
241 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
242 into the default Received: header string.
249 TK/01 Added the "success_on_redirect" address verification option. See
250 NewStuff for rationale and an example.
252 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
254 PH/02 Patch to exigrep to allow it to work on syslog lines.
256 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
257 fread() to skip over the body file's header line, because in Cygwin the
258 header line is locked and is inaccessible.
260 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
261 co-exist for some time) to make it clear that it is the Exim ID that is
262 referenced, not the Message-ID: header line.
264 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
265 string_format() function, because snprintf() does not exist on all
268 PH/06 The use of forbid_filter_existstest now also locks out the use of the
269 ${stat: expansion item.
271 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
272 protocol synchronization error", to keep the pedants happy.
274 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
275 well as for IRIX systems, when gcc is being used. See the host.c source
278 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
280 PH/10 Named domain lists were not working if used in a queue_smtp_domains
283 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
284 transport and to the smtp transport in LMTP mode.
286 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
288 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
289 run a filter in a subprocess. This could lead to confusion in subsequent
290 lookups in the parent process. There should also be a search_tidyup() at
291 the end of the subprocess.
293 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
294 only if the host matched helo_try_verify_hosts, which caused the
295 verification to occur when the EHLO/HELO command was issued. The ACL just
296 tested the remembered result. Now, if a previous verification attempt has
297 not happened, "verify = helo" does it there and then.
299 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
301 TK/03 Fix log output including CR from clamd.
303 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
304 exist provoked a memory error which could cause a segfault.
306 PH/15 Installed PCRE 6.2
308 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
310 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
311 of the problem. Specifically, suggested +O2 rather than +O1 for the
314 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
316 PH/20 If a delivery was routed to a non-standard port by means of an SRV
317 record, the port was not correctly logged when the outgoing_port log
318 selector was set (it logged the transort's default port).
320 PH/21 Added support for host-specific ports to manualroute, queryprogram,
321 fallback_hosts, and "hosts" in the smtp transport.
323 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
324 host errors such as "Connection refused".
326 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
327 authentication with radiusclient 0.4.9:
329 - Error returned from rc_read_config was caught wrongly
330 - Username/password not passed on to radius server due to wrong length.
332 The presumption is that some radiusclient API changes for 4.51/PH/17
333 were not taken care of correctly. The code is still untested by me (my
334 Linux distribution still has 0.3.2 of radiusclient), but it was
335 contributed by a Radius user.
337 PH/24 When doing a callout, the value of $domain wasn't set correctly when
338 expanding the "port" option of the smtp transport.
340 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
341 while reading a MIME header. Thanks to Tom Hughes for a patch.
343 PH/24 Include config.h inside local_scan.h so that configuration settings are
346 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
347 that in an ACL for RCPT (for example), you can examine exactly what was
350 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
351 commands, but it was not correctly comparing the address with the actual
352 client host address. Thus, it would show the EHLO address in Received:
353 header lines when this was not necessary.
355 PH/27 Added the % operator to ${eval:}.
357 PH/28 Exim tries to create and chdir to its spool directory when it starts;
358 it should be ignoring failures (because with -C, for example, it has lost
359 privilege). It wasn't ignoring creation failures other than "already
362 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
365 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
366 because an input line was too long, either on its own, or by virtue of
367 too many continuations, the temporary file was not being removed, and the
368 return code was incorrect.
370 PH/31 Missing "BOOL" in function definition in filtertest.c.
372 PH/32 Applied Sieve patches from the maintainer.
374 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
376 PH/33 Added "verify = not_blind".
378 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
379 Local/Makefile (with some defaults set). These are used in built scripts
380 such as exicyclog, but they have never been used in the exim_install
381 script (though there are many overriding facilities there). I have
382 arranged that the exim_install script now takes note of these two
385 PH/35 Installed configuration files for Dragonfly.
387 PH/36 When a locally submitted message by a trusted user did not contain a
388 From: header, and the sender address was obtained from -f or from an SMTP
389 MAIL command, and the trusted user did not use -F to supply a sender
390 name, $originator_name was incorrectly used when constructing a From:
391 header. Furthermore, $originator_name was used for submission mode
392 messages from external hosts without From: headers in a similar way,
393 which is clearly wrong.
395 PH/37 Added control=suppress_local_fixups.
397 PH/38 When log_selector = +received_sender was set, and the addition of the
398 sender made the log line's construction buffer exactly full, or one byte
399 less than full, an overflow happened when the terminating "\n" was
402 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
403 when the result of a list match is failure because a DNS lookup failed.
405 PH/40 RM_COMMAND is now used in the building process.
407 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
408 the "build-* directories that it finds.
410 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
411 address in a domain literal was a prefix of an interface address.
413 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
414 when verifying a sender address, unless rewrite_headers is false.
416 PH/44 Wrote a long comment about why errors_to addresses are verified as
417 recipients, not senders.
419 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
420 the ratelimit ACL was added.
422 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
424 PH/46 Added extra information about PostgreSQL errors to the error string.
426 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
427 avoiding the need to install special zones in a real server. This is
428 backwards compatible; if it can't find the fake resolver, it drops back.
429 Thus, both old and new test suites can be run.
431 TF/02 Added util/ratelimit.pl
433 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
434 clock is set back in time.
436 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
437 Candler <B.Candler@pobox.com>.
439 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
440 OK for addresses that are the result of redirections.
442 PH/48 A number of further additions for the benefit of the new test suite,
443 including a fake gethostbyname() that interfaces to the fake DNS resolver
446 TF/06 The fix for widen_domains has also been applied to qualify_single and
447 search_parents which are the other dnslookup options that can cause
450 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
453 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
455 TF/07 Exim produced the error message "an SRV record indicated no SMTP
456 service" if it encountered an MX record with an empty target hostname.
457 The message is now "an MX or SRV record indicated no SMTP service".
459 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
460 if the DNS of the sending site is misconfigured. This is quite a
461 common situation. This change restores the behaviour of treating a
462 helo verification defer as a failure.
464 PH/51 If self=fail was set on a router, the bounce message did not include the
465 actual error message.
471 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
473 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
474 command itself ended in error, the underlying message about the transport
475 filter timeout was being overwritten with the pipe command error. Now the
476 underlying error message should be appended to the second error message.
478 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
480 PH/02 Reduce the amount of output that "make" produces by default. Full output
481 can still be requested.
483 PH/03 The warning log line about a condition test deferring for a "warn" verb
484 was being output only once per connection, rather than after each
485 occurrence (because it was using the same function as for successful
486 "warn" verbs). This seems wrong, so I have changed it.
488 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
489 it should not have, which might have caused a crash in the right
490 circumstances, but probably never did.
492 PH/04 Installed a modified version of Tony Finch's patch to make submission
493 mode fix the return path as well as the Sender: header line, and to
494 add a /name= option so that you can make the user's friendly name appear
497 TF/03 Added the control = fakedefer ACL modifier.
499 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
500 Mark Lowes for thorough testing.
502 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
504 TK/03 Merged latest SRS patch from Miles Wilton.
506 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
507 with the definition in sysexits.h (which is #included earlier).
508 Fortunately, Exim does not actually use EX_OK. The code used to try to
509 preserve the sysexits.h value, by assumimg that macro definitions were
510 scanned for macro replacements. I have been disabused of this notion,
511 so now the code just undefines EX_OK before #including unistd.h.
513 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
514 in the smtp transport. When a block could not be written in a single
515 write() function, the timeout was being re-applied to each part-write.
516 This seems wrong - if the receiver was accepting one byte at a time it
517 would take for ever. The timeout is now adjusted when this happens. It
518 doesn't have to be particularly precise.
520 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
521 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
523 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
524 without a domain if the check to <postmaster@domain> fails.
526 SC/01 Eximstats: added -xls and the ability to specify output files
527 (patch written by Frank Heydlauf).
529 SC/02 Eximstats: use FileHandles for outputing results.
531 SC/03 Eximstats: allow any combination of xls, txt, and html output.
533 SC/04 Eximstats: fixed display of large numbers with -nvr option
535 SC/05 Eximstats: fixed merging of reports with empty tables.
537 SC/06 Eximstats: added the -include_original_destination flag
539 SC/07 Eximstats: removed tabs and trailing whitespace.
541 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
543 TK/06 MBOX spool code: Add real "From " MBOX separator line
544 so the .eml file is really in mbox format (even though
545 most programs do not really care). Patch from Alex Miller.
547 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
548 The latter is generated from $received_to and is only set if the
549 message has one envelope recipient. SA can use these headers,
550 obviously out-of-the-box. Patch from Alex Miller.
552 PH/08 The ${def test on a variable was returning false if the variable's
553 value was "0", contrary to what the specification has always said!
554 The result should be true unless the variable is empty.
556 PH/09 The syntax error of a character other than { following "${if
557 def:variable_name" (after optional whitespace) was not being diagnosed.
558 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
559 accidental colon was present, for example, could give incorrect results.
561 PH/10 Tidied the code in a number of places where the st_size field of a stat()
562 result is used (not including appendfile, where other changes are about
565 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
566 This involved changing a lot of size variables from int to off_t. It
567 should work with maildirs and everything.
569 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
570 spamd dying while we are connected to it.
572 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
575 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
576 function for BDB 4.3.
578 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
580 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
581 the sources for PCRE have been re-arranged and more files are now
584 PH/15 The code I had for printing potentially long long variables in PH/11
585 above was not the best (it lost precision). The length of off_t variables
586 is now inspected at build time, and an appropriate printing format (%ld
587 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
588 to be "long long int" or "long int". This is needed for the internal
589 formatting function string_vformat().
591 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
592 the configuration file to be ":syslog", then the script "guesses" where
593 the logs files are, rather than using the compiled in default. In our
594 case the guess is not the same as the compiled default, so the script
595 suddenly stopped working when I started to use syslog. The patch checks
596 to see if log_file_path is "". If so, it attempts to read it from exim
597 with no configuration file to get the compiled in version, before it
598 falls back to the previous guessing code."
600 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
601 implementing BATV in an Exim configuration. See NewStuff for the gory
604 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
605 Makefile that are specific to HP-UX.
607 PH/18 If the "use_postmaster" option was set for a recipient callout together
608 with the "random" option, the postmaster address was used as the MAIL
609 FROM address for the random test, but not for the subsequent recipient
610 test. It is now used for both.
612 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
613 patch removes a few documentation additions to RFC 3028, because the
614 latest draft now contains them. It adds the new en;ascii-case comparator
615 and a new error check for 8bit text in MIME parts. Comparator and
616 require names are now matched exactly. I enabled the subaddress
617 extension, but it is not well tested yet (read: it works for me)."
619 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
620 rework some of the code of TK/09 above to avoid the hardwired use of
621 "%lld" and "long long". Replaced the call to snprintf() with a call to
624 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
625 records point to non-existent hosts", "retry timeout exceeded", and
626 "retry time not reached for any host after a long failure period".
628 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
629 experimental DomainKeys support:
631 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
632 (2) On an error such as an illegally used "control", the wrong name for
633 the control was given.
635 These problems did NOT occur unless DomainKeys support was compiled.
637 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
639 PH/24 Added ${if match_ip condition.
641 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
642 databases so that it will be absolutely obvious if a crash occurs in the
643 DB library. This is a regular occurrence (often caused by mis-matched
646 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
647 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
648 picked up on a user's system that detects such things. There doesn't seem
649 to be a gcc warning option for this - only an attribute that has to be
650 put on the function's prototype. It seems that in Fedora Core 4 they have
651 set this on a number of new functions. No doubt there will be more in due
654 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
655 specify a transport. However, if an address that was verified by such a
656 router was the subject of a callout, Exim crashed because it tried to
657 read the rcpt_include_affixes from the non-existent transport. Now it
658 just assumes that the setting of that option is false. This bug was
659 introduced by 4.51/PH/31.
661 PH/28 Changed -d+all to exclude +memory, because that information is very
662 rarely of interest, but it makes the output a lot bigger. People tend to
663 do -d+all out of habit.
665 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
666 code in os-type was giving problems when libc.so lives in lib64, like on
669 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
670 aren't the modern standard, and it seems that some systems' include files
671 don't always have them. Exim was already checking for some of the newer
672 ones like T_AAAA, and defining it itself. I've added checks for all the
673 record types that Exim uses.
675 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
676 not automatically generating a new one, as it is supposed to. This
677 prevented TLS from working. If the file did exist, but contained invalid
678 data, a new version was generated, as expected. It was only the case of a
679 non-existent file that was broken.
681 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
682 with a change in libdomainkeys > 0.64.
684 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
685 from DNS. If the selector record carries the flag, it now has
686 precedence over the domain-wide flag.
688 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
690 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
691 the use of a port name that isn't defined in /etc/services) occurred, the
692 message was deferred as in a normal delivery, and thus remained on the
693 spool, instead of being failed because of the mua_wrapper setting. This
694 is now fixed, and I tidied up some of the mua_wrapper messages at the
697 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
698 the messages in a hash of arrays rather than using individual hashes.
699 This is a bit cleaner and results in dramatic memory savings, albeit
700 at a slight CPU cost.
702 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
703 as requested by Marc Sherman.
705 SC/10 Eximstats: added histograms for user specified patterns as requested
708 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
710 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
711 fopen() in the content-scanning modules that did not already have it.
717 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
718 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
720 TK/02 Fix ACL "control" statement not being available in MIME ACL.
722 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
724 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
725 to test Sieve filters that use "vacation".
727 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
728 that changes the way the GnuTLS parameters are stored in the cache file.
729 The new format can be generated externally. For backward compatibility,
730 if the data in the cache doesn't make sense, Exim assumes it has read an
731 old-format file, and it generates new data and writes a new file. This
732 means that you can't go back to an older release without removing the
735 PH/03 A redirect router that has both "unseen" and "one_time" set does not
736 work if there are any delivery delays because "one_time" forces the
737 parent to be marked "delivered", so its unseen clone is never tried
738 again. For this reason, Exim now forbids the simultaneous setting of
741 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
742 redirected to themselves ("homonym" addresses). Read the long ChangeLog
743 entry if you want to know the details. The fix, however, neglected to
744 consider the case when local delivery batching is involved. The test for
745 "previously delivered" was not happening when checking to see if an
746 address could be batched with a previous (undelivered) one; under
747 certain circumstances this could lead to multiple deliveries to the same
750 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
751 in its include files, and this causes problems building Exim.
753 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
754 header_syntax) but Exim was just ignoring anything given after a slash.
755 In particular, this caused confusion with an attempt to use "verify =
756 reverse_host_lookup/defer_ok". An error is now given when options are
757 supplied for verify items that do not have them. (Maybe reverse_host_
758 lookup should have a defer_ok option, but that's a different point.)
760 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
761 defined by RFC 821) to 2048, because there were problems with some AUTH
762 commands, and RFC 1869 says the size should be increased for extended
763 SMTP commands that take arguments.
765 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
768 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
769 "unknown" error; now it says that the functionality isn't in the binary.
771 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
772 an address' error message when a string expansion fails (syntax or
773 whatever). Otherwise the password may appear in the log. Following change
774 PH/42 below, there is no longer a chance of it appearing in a bounce
777 PH/11 Installed exipick version 20050225.0 from John Jetmore.
779 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
780 of its addresses was ever tried. (Bugzilla bug #2.)
782 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
783 the result incorrectly in the debug output. (It correctly added a newline
784 to what was transported.)
786 TF/01 Added $received_time.
788 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
789 commented out examples of how to interface to a virus scanner and to
790 SpamAssassin. Also added commented examples of av_scanner and
791 spamd_address settings.
793 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
794 and controls are allowed in which ACLs. There were a couple of minor
795 errors. Some of the entries in the conditions table (which is a table of
796 where they are NOT allowed) were getting very unwieldy; rewrote them as a
797 negation of where the condition IS allowed.
799 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
801 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
802 header file does not have a version number, so I've had to invent a new
803 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
804 API. The code is untested by me (my Linux distribution still has 0.3.2 of
805 radiusclient), but it was contributed by a Radius user.
807 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
808 files or directories, for OpenSSL.
810 PH/19 When an Exim process that is running as root has to create an Exim log
811 file, it does so in a subprocess that runs as exim:exim so as to get the
812 ownership right at creation (otherwise, other Exim processes might see
813 the file with the wrong ownership). There was no test for failure of this
814 fork() call, which would lead to the process getting stuck as it waited
815 for a non-existent subprocess. Forks do occasionally fail when resources
816 run out. I reviewed all the other calls to fork(); they all seem to check
819 PH/20 When checking for unexpected SMTP input at connect time (before writing
820 the banner), Exim was not dealing correctly with a non-positive return
821 from the read() function. If the client had disconnected by this time,
822 the result was a log entry for a synchronization error with an empty
823 string after "input=" when read() returned zero. If read() returned -1
824 (an event I could not check), uninitialized data bytes were printed.
825 There were reports of junk text (parts of files, etc) appearing after
828 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
830 PH/22 Added support for macro redefinition, and (re)definition in between
831 driver and ACL definitions.
833 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
834 forgetting to use the resulting value; it was using the unexpanded value.
836 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
837 hadn't been configured. The fix is from Juergen Kreileder, who
838 understands it better than I do:
840 "Here's what I see happening with three configured cyrus_sasl
841 authenticators configured (plain, login, cram-md5):
843 On startup auth_cyrus_sasl_init() gets called for each of these.
844 This means three calls to sasl_listmech() without a specified mech_list.
845 => SASL tests which mechs of all available mechs actually work
846 => three warnings about OTP not working
847 => the returned list contains: plain, login, cram-md5, digest-md5, ...
849 With the patch, sasl_listmech() also gets called three times. But now
850 SASL's mech_list option is set to the server_mech specified in the the
851 authenticator. Or in other words, the answer from sasl_listmech()
852 gets limited to just the mech you're testing for (which is different
854 => the return list contains just 'plain' or 'login', 'cram-md5' or
855 nothing depending on the value of ob->server_mech.
857 I've just tested the patch: Authentication still works fine,
858 unavailable mechs specified in the exim configuration are still
859 caught, and the auth.log warnings about OTP are gone."
861 PH/25 When debugging is enabled, the contents of the command line are added
862 to the debugging output, even when log_selector=+arguments is not
865 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
866 answer is "GNU", and only if the return is "GNU/something" is the answer
869 PH/27 $acl_verify_message is now set immediately after the failure of a
870 verification in an ACL, and so is available in subsequent modifiers. In
871 particular, the message can be preserved by coding like this:
873 warn !verify = sender
874 set acl_m0 = $acl_verify_message
876 Previously, $acl_verify_message was set only while expanding "message"
877 and "log_message" when a very denied access.
879 PH/28 Modified OS/os.c-Linux with
881 -#ifndef OS_LOAD_AVERAGE
882 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
884 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
885 nomenclature these days.)
887 PH/29 Installed patch from the Sieve maintainer that adds the options
888 sieve_useraddress and sieve_subaddress to the redirect router.
890 PH/30 In these circumstances:
891 . Two addresses routed to the same list of hosts;
892 . First host does not offer TLS;
893 . First host accepts first address;
894 . First host gives temporary error to second address;
895 . Second host offers TLS and a TLS session is established;
896 . Second host accepts second address.
897 Exim incorrectly logged both deliveries with the TLS parameters (cipher
898 and peerdn, if requested) that were in fact used only for the second
901 PH/31 When doing a callout as part of verifying an address, Exim was not paying
902 attention to any local part prefix or suffix that was matched by the
903 router that accepted the address. It now behaves in the same way as it
904 does for delivery: the affixes are removed from the local part unless
905 rcpt_include_affixes is set on the transport.
907 PH/32 Add the sender address, as F=<...>, to the log line when logging a
908 timeout during the DATA phase of an incoming message.
910 PH/33 Sieve envelope tests were broken for match types other than :is. I have
911 applied a patch sanctioned by the Sieve maintainer.
913 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
914 the uid or gid is negative. A case of a negative gid caused this to be
915 noticed. The fix allows for either to be negative.
917 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
918 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
921 PH/36 The change PH/12 above was broken. Fixed it.
923 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
924 the grounds that routing the same address twice would always produce the
925 same answer. This might have been true once, but it is certainly no
926 longer true now. Routing a child address may depend on the previous
927 routing that produced that child. Some complicated redirection strategies
928 went wrong when messages had multiple recipients, and made Exim's
929 behaviour dependent on the order in which the addresses were given.
931 I have moved the duplicate checking until after the routing is complete.
932 Exim scans the addresses that are assigned to local and remote
933 transports, and removes any duplicates. This means that more work will be
934 done, as duplicates will always all be routed, but duplicates are
935 presumably rare, so I don't expect this is of any significance.
937 For deliveries to pipes, files, and autoreplies, the duplicate checking
938 still happens during the routing process, since they are not going to be
941 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
942 It corrects a timeout issue with spamd. This is Ian's comment: "The
943 background is that sometimes spamd either never reads data from a
944 connection it has accepted, or it never writes response data. The exiscan
945 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
946 blindly assumes that writes won't block so it may never time out."
948 PH/39 Allow G after quota size as well as K and M.
950 PH/40 The value set for $authenticated_id in an authenticator may not contain
951 binary zeroes or newlines because the value is written to log lines and
952 to spool files. There was no check on this. Now the value is run through
953 the string_printing() function so that such characters are converted to
954 printable escape sequences.
956 PH/41 $message_linecount is a new variable that contains the total number of
957 lines in the message. Compare $body_linecount, which is the count for the
960 PH/42 Exim no longer gives details of delivery errors for specific addresses in
961 bounce and delay warning messages, except in certain special cases, which
964 (a) An SMTP error message from a remote host;
965 (b) A message specified in a :fail: redirection;
966 (c) A message specified in a "fail" command in a system filter;
967 (d) A message specified in a FAIL return from the queryprogram router;
968 (e) A message specified by the cannot_route_message router option.
970 In these cases only, Exim does include the error details in bounce and
971 warning messages. There are also a few cases where bland messages such
972 as "unrouteable address" or "local delivery error" are given.
974 PH/43 $value is now also set for the "else" part of a ${run expansion.
976 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
977 being worked on, but at least Exim now implements the latest version to
980 PH/45 In a pipe transport, although a timeout while waiting for the pipe
981 process to complete was treated as a delivery failure, a timeout while
982 writing the message to the pipe was logged, but erroneously treated as a
983 successful delivery. Such timeouts include transport filter timeouts. For
984 consistency with the overall process timeout, these timeouts are now
985 treated as errors, giving rise to delivery failures by default. However,
986 there is now a new Boolean option for the pipe transport called
987 timeout_defer, which, if set TRUE, converts the failures into defers for
988 both kinds of timeout. A transport filter timeout is now identified in
991 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
992 systems where "make" and "gmake" are different, calling "gmake" at top
993 level broke things. I've arranged for the value of $(MAKE) to be passed
994 from the Makefile to this script so that it can call the same version of
998 A note about Exim versions 4.44 and 4.50
999 ----------------------------------------
1001 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1002 changes of various kinds. As a consequence, a big documentation update was
1003 needed. This delayed the release for rather longer than seemed good, especially
1004 in the light of a couple of (minor) security issues. Therefore, the changes
1005 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1006 release. So 4.44 and 4.50 are in effect two different branches that both start
1009 I have left the 4.50 change log unchanged; it contains all the changes since
1010 4.43. The change log for 4.44 is below; many of its items are identical to
1011 those for 4.50. This seems to be the most sensible way to preserve the
1012 historical information.
1018 1. Minor wording change to the doc/README.SIEVE file.
1020 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1021 computation of the current number of files was incorrect.
1023 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1024 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1027 4. Give more explanation in the error message when the command for a transport
1028 filter fails to execute.
1030 5. There are several places where Exim runs a non-Exim command in a
1031 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1032 was being done only for the command run by the queryprogram router. It is
1033 now done for all such subprocesses. The other cases are: ${run, transport
1034 filters, and the commands run by the lmtp and pipe transports.
1036 6. Added CONFIGURE_GROUP build-time option.
1038 7. Some older OS have a limit of 256 on the maximum number of file
1039 descriptors. Exim was using setrlimit() to set 1000 as a large value
1040 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1041 systems. I've change it so that if it can't get 1000, it tries for 256.
1043 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1044 was an oversight, and furthermore, ever since the addition of extra
1045 controls (e.g. 4.43/32), the checks on when to allow different forms of
1046 "control" were broken. There should now be diagnostics for all cases when a
1047 control that does not make sense is encountered.
1049 9. Added the /retain_sender option to "control=submission".
1051 10. $recipients is now available in the predata ACL (oversight).
1053 11. Tidy the search cache before the fork to do a delivery from a message
1054 received from the command line. Otherwise the child will trigger a lookup
1055 failure and thereby defer the delivery if it tries to use (for example) a
1056 cached ldap connection that the parent has called unbind on.
1058 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1059 of $address_data from the recipient verification was clobbered by the
1060 sender verification.
1062 13. The value of address_data from a sender verification is now available in
1063 $sender_address_data in subsequent conditions in the ACL statement.
1065 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1067 15. Added a new option "connect=<time>" to callout options, to set a different
1070 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1071 was its contents. (It was OK if the option was not defined at all.)
1073 17. A "Completed" log line is now written for messages that are removed from
1074 the spool by the -Mrm option.
1076 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1077 information about exactly what failed.
1079 19. Added -dd to debug only the daemon process.
1081 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1082 handles timeouts, both on the server side and network timeouts. Renamed the
1083 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1085 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1086 It is now set to "smtps".
1088 22. $host_address is now set to the target address during the checking of
1089 ignore_target_hosts.
1091 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1092 being passed; this would have caused $sender_host_name to have been used if
1093 matching the list had actually called for a host name (not very likely,
1094 since this list is usually IP addresses). A host name is now passed as
1097 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1098 code that specifies a non-restarting handler (typically sigaction() in
1099 modern systems) in an attempt to fix a rare and obscure crash bug.
1101 25. Narrowed the window for a race in the daemon that could cause it to ignore
1102 SIGCHLD signals. This is not a major problem, because they are used only to
1103 wake it up if nothing else does.
1105 26. A malformed maildirsize file could cause Exim to calculate negative values
1106 for the mailbox size or file count. Odd effects could occur as a result.
1107 The maildirsize information is now recalculated if the size or filecount
1110 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1111 support for a long time. Removed HAVE_SYS_VFS_H.
1113 28. Installed the latest version of exipick from John Jetmore.
1115 29. In an address list, if the pattern was not a regular expression, an empty
1116 subject address (from a bounce message) matched only if the pattern was an
1117 empty string. Non-empty patterns were not even tested. This was the wrong
1118 because it is perfectly reasonable to use an empty address as part of a
1119 database query. An empty address is now tested by patterns that are
1120 lookups. However, all the other forms of pattern expect the subject to
1121 contain a local part and a domain, and therefore, for them, an empty
1122 address still always fails if the pattern is not itself empty.
1124 30. Exim went into a mad DNS loop when attempting to do a callout where the
1125 host was specified on an smtp transport, and looking it up yielded more
1126 than one IP address.
1128 31. Re-factored the code for checking spool and log partition space into a
1129 function that finds that data and another that does the check. The former
1130 is then used to implement four new variables: $spool_space, $log_space,
1131 $spool_inodes, and $log_inodes.
1133 32. The RFC2047 encoding function was originally intended for short strings
1134 such as real names; it was not keeping to the 75-character limit for
1135 encoded words that the RFC imposes. It now respects the limit, and
1136 generates multiple encoded words if necessary. To be on the safe side, I
1137 have increased the buffer size for the ${rfc2047: expansion operator from
1140 33. It is now permitted to omit both strings after an "if" condition; if the
1141 condition is true, the result is "true". As before, when the second string
1142 is omitted, a false condition yields an empty string. This makes it less
1143 cumbersome to write custom ACL and router conditions.
1145 34. Failure to deliver a bounce message always caused it to be frozen, even if
1146 there was an errors_to setting on the router. The errors_to setting is now
1149 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1150 canonical form (fully expanded) before being placed in
1151 $sender_host_address.
1153 36. The table in the code that translates DNS record types into text (T_A to
1154 "A" for instance) was missing entries for NS and CNAME. It is just possible
1155 that this could have caused confusion if both these types were looked up
1156 for the same domain, because the text type is used as part of Exim's
1157 per-process caching. But the chance of anyone hitting this buglet seems
1160 37. The dnsdb lookup has been extended in a number of ways.
1162 (1) There is a new type, "zns", which walks up the domain tree until it
1163 finds some nameserver records. It should be used with care.
1165 (2) There is a new type, "mxh", which is like "mx" except that it returns
1166 just the host names, not the priorities.
1168 (3) It is now possible to give a list of domains (or IP addresses) to be
1169 looked up. The behaviour when one of the lookups defers can be
1170 controlled by a keyword.
1172 (4) It is now possible to specify the separator character for use when
1173 multiple records are returned.
1175 38. The dnslists ACL condition has been extended: it is now possible to supply
1176 a list of IP addresses and/or domains to be looked up in a particular DNS
1179 39. Added log_selector=+queue_time_overall.
1181 40. When running the queue in the test harness, wait just a tad after forking a
1182 delivery process, to get repeatability of debugging output.
1184 41. Include certificate and key file names in error message when GnuTLS fails
1185 to set them up, because the GnuTLS error message doesn't include the name
1186 of the failing file when there is a problem reading it.
1188 42. Allow both -bf and -bF in the same test run.
1190 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1192 44. The "Exiscan patch" is now merged into the mainline Exim source.
1194 45. Sometimes the final signoff response after QUIT could fail to get
1195 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1196 < 0 before doing a fflush(). This bug looks as though it goes back to the
1197 introduction of TLS in release 3.20, but "sometimes" must have been rare
1198 because the tests only now provoked it.
1200 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1201 (this can affect the format of dates).
1203 47. exim_tidydb, when checking for the continued existence of a message for
1204 which it has found a message-specific retry record, was not finding
1205 messages that were in split spool directories. Consequently, it was
1206 deleting retry records that should have stayed in existence.
1208 48. Steve fixed some bugs in eximstats.
1210 49. The SPA authentication driver was not abandoning authentication and moving
1211 on to the next authenticator when an expansion was forced to fail,
1212 contradicting the general specification for all authenticators. Instead it
1213 was generating a temporary error. It now behaves as specified.
1215 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1216 (the order specifies the preference for clients). The order is now AES256,
1217 AES128, 3DES, ARCFOUR128.
1219 51. Small patch to Sieve code - explicitly set From: when generating an
1222 52. Exim crashed if a remote delivery caused a very long error message to be
1223 recorded - for instance if somebody sent an entire SpamAssassin report back
1224 as a large number of 550 error lines. This bug was coincidentally fixed by
1225 increasing the size of one of Exim's internal buffers (big_buffer) that
1226 happened as part of the Exiscan merge. However, to be on the safe side, I
1227 have made the code more robust (and fixed the comments that describe what
1230 53. Now that there can be additional text after "Completed" in log lines (if
1231 the queue_time_overall log selector is set), a one-byte patch to exigrep
1232 was needed to allow it to recognize "Completed" as not the last thing in
1235 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1236 patch that reportedly fixes this has been added. I am not expert enough to
1237 create a test for it. This is what the patch creator wrote:
1239 "I found a little strange behaviour of ldap code when working with
1240 Windows 2003 AD Domain, where users was placed in more than one
1241 Organization Units. When I tried to give exim partial DN, the exit code
1242 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1243 But simultaneously result of request was absolutely normal ldap result,
1244 so I produce this patch..."
1246 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1247 REFERENCE, so I have modified the code to exclude the patch when that macro
1250 55. Some experimental protocols are using DNS PTR records for new purposes. The
1251 keys for these records are domain names, not reversed IP addresses. The
1252 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1253 leaves it alone. Component reversal etc. now happens only for IP addresses.
1256 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1258 57. Double the size of the debug message buffer (to 2048) so that more of very
1259 long debug lines gets shown.
1261 58. The exicyclog utility now does better if the number of log files to keep
1262 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1264 59. Two changes related to the smtp_active_hostname option:
1266 (1) $smtp_active_hostname is now available as a variable.
1267 (2) The default for smtp_banner uses $smtp_active_hostname instead
1268 of $primary_hostname.
1270 60. The host_aton() function is supposed to be passed a string that is known
1271 to be a valid IP address. However, in the case of IPv6 addresses, it was
1272 not checking this. This is a hostage to fortune. Exim now panics and dies
1273 if the condition is not met. A case was found where this could be provoked
1274 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1275 components; fortuitously, this particular loophole had already been fixed
1276 by change 4.50/55 above.
1278 If there are any other similar loopholes, the new check in host_aton()
1279 itself should stop them being exploited. The report I received stated that
1280 data on the command line could provoke the exploit when Exim was running as
1281 exim, but did not say which command line option was involved. All I could
1282 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1283 running as the user.
1286 61. There was a buffer overflow vulnerability in the SPA authentication code
1287 (which came originally from the Samba project). I have added a test to the
1288 spa_base64_to_bits() function which I hope fixes it.
1291 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1292 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1294 63. The daemon start-up calls getloadavg() while still root for those OS that
1295 need the first call to be done as root, but it missed one case: when
1296 deliver_queue_load_max is set with deliver_drop_privilege. This is
1297 necessary for the benefit of the queue runner, because there is no re-exec
1298 when deliver_drop_privilege is set.
1300 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1301 This has been fixed.
1303 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1304 was in use, was not putting the data itself into the right store pool;
1305 consequently, it could be overwritten for a subsequent message in the same
1306 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1309 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1311 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1312 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1313 the function were treating the return as a boolean value, which happened to
1314 work because 0=false and not-0=true, but is not correct code.
1316 68. The host_aton() function was not handling scoped IPv6 addresses (those
1317 with, for example, "%eth0" on the end) correctly.
1319 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1320 negated items (that is, ~something) in unsigned ints. Some compilers
1321 apparently mutter when there is no cast.
1323 70. If an address verification called from an ACL failed, and did not produce a
1324 user-specific message (i.e. there was only a "system" message), nothing was
1325 put in $acl_verify_message. In this situation, it now puts the system
1328 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1329 session; change 4.31/43 added the unwanted input to the log line - except
1330 that it did not do this in the start of session case. It now does.
1332 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1333 This is wrong and can cause the other end to generate a synchronization
1334 error if it is another Exim or anything else that does the synchronization
1335 check. A QUIT command is no longer sent after a timeout.
1337 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1338 during host lookups.
1340 74. The defer_ok option of callout verification was not working if it was used
1341 when verifying addresses in header lines, that is, for this case:
1343 verify = header_sender/callout=defer_ok
1345 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1346 those file descriptors could be used for SMTP connections. If anything
1347 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1348 could be sent to the SMTP client, causing chaos. The daemon now opens
1349 stdin, stdout, and stderr to /dev/null when it puts itself into the
1352 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1353 log by default. The user can override this with suitable Perl magic.
1355 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1356 the log message when discard triggers, was not working for the DATA ACL or
1357 for the non-SMTP ACL.
1359 78. Error message wording change in sieve.c.
1361 79. If smtp_accept_max_per_host was set, the number of connections could be
1362 restricted to fewer than expected, because the daemon was trying to set up
1363 a new connection before checking whether the processes handling previous
1364 connections had finished. The check for completed processes is now done
1365 earlier. On busy systems, this bug wouldn't be noticed because something
1366 else would have woken the daemon, and it would have reaped the completed
1369 80. If a message was submitted locally by a user whose login name contained one
1370 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1371 It caused a spool format error. I have fixed the spool reading code. A
1372 related problem was that the "from" clause in the Received: line became
1373 illegal because of the space(s). It is now covered by ${quote_local_part.
1375 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1378 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1380 83. Patch from Sieve maintainer to fix unterminated string problem in
1381 "vacation" handling.
1383 84. Some minor changes to the Linux configuration files to help with other
1384 OS variants using glibc.
1386 85. One more patch for Sieve to update vacation handling to latest spec.
1389 ----------------------------------------------------
1390 See the note above about the 4.44 and 4.50 releases.
1391 ----------------------------------------------------
1397 1. Change 4.43/35 introduced a bug that caused file counts to be
1398 incorrectly computed when quota_filecount was set in an appendfile
1401 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1402 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1405 3. Give more explanation in the error message when the command for a transport
1406 filter fails to execute.
1408 4. There are several places where Exim runs a non-Exim command in a
1409 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1410 was being done only for the command run by the queryprogram router. It is
1411 now done for all such subprocesses. The other cases are: ${run, transport
1412 filters, and the commands run by the lmtp and pipe transports.
1414 5. Some older OS have a limit of 256 on the maximum number of file
1415 descriptors. Exim was using setrlimit() to set 1000 as a large value
1416 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1417 systems. I've change it so that if it can't get 1000, it tries for 256.
1419 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1420 was an oversight, and furthermore, ever since the addition of extra
1421 controls (e.g. 4.43/32), the checks on when to allow different forms of
1422 "control" were broken. There should now be diagnostics for all cases when a
1423 control that does not make sense is encountered.
1425 7. $recipients is now available in the predata ACL (oversight).
1427 8. Tidy the search cache before the fork to do a delivery from a message
1428 received from the command line. Otherwise the child will trigger a lookup
1429 failure and thereby defer the delivery if it tries to use (for example) a
1430 cached ldap connection that the parent has called unbind on.
1432 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1433 of $address_data from the recipient verification was clobbered by the
1434 sender verification.
1436 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1437 was its contents. (It was OK if the option was not defined at all.)
1439 11. A "Completed" log line is now written for messages that are removed from
1440 the spool by the -Mrm option.
1442 12. $host_address is now set to the target address during the checking of
1443 ignore_target_hosts.
1445 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1446 being passed; this would have caused $sender_host_name to have been used if
1447 matching the list had actually called for a host name (not very likely,
1448 since this list is usually IP addresses). A host name is now passed as
1451 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1452 code that specifies a non-restarting handler (typically sigaction() in
1453 modern systems) in an attempt to fix a rare and obscure crash bug.
1455 15. Narrowed the window for a race in the daemon that could cause it to ignore
1456 SIGCHLD signals. This is not a major problem, because they are used only to
1457 wake it up if nothing else does.
1459 16. A malformed maildirsize file could cause Exim to calculate negative values
1460 for the mailbox size or file count. Odd effects could occur as a result.
1461 The maildirsize information is now recalculated if the size or filecount
1464 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1465 support for a long time. Removed HAVE_SYS_VFS_H.
1467 18. Updated exipick to current release from John Jetmore.
1469 19. Allow an empty sender to be matched against a lookup in an address list.
1470 Previously the only cases considered were a regular expression, or an
1473 20. Exim went into a mad DNS lookup loop when doing a callout where the
1474 host was specified on the transport, if the DNS lookup yielded more than
1477 21. The RFC2047 encoding function was originally intended for short strings
1478 such as real names; it was not keeping to the 75-character limit for
1479 encoded words that the RFC imposes. It now respects the limit, and
1480 generates multiple encoded words if necessary. To be on the safe side, I
1481 have increased the buffer size for the ${rfc2047: expansion operator from
1484 22. Failure to deliver a bounce message always caused it to be frozen, even if
1485 there was an errors_to setting on the router. The errors_to setting is now
1488 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1489 canonical form (fully expanded) before being placed in
1490 $sender_host_address.
1492 24. Updated eximstats to version 1.33
1494 25. Include certificate and key file names in error message when GnuTLS fails
1495 to set them up, because the GnuTLS error message doesn't include the name
1496 of the failing file when there is a problem reading it.
1498 26. Expand error message when OpenSSL has problems setting up cert/key files.
1501 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1502 (this can affect the format of dates).
1504 28. exim_tidydb, when checking for the continued existence of a message for
1505 which it has found a message-specific retry record, was not finding
1506 messages that were in split spool directories. Consequently, it was
1507 deleting retry records that should have stayed in existence.
1509 29. eximstats updated to version 1.35
1510 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1511 1.35 - bugfix such that pie charts by volume are generated correctly
1513 30. The SPA authentication driver was not abandoning authentication and moving
1514 on to the next authenticator when an expansion was forced to fail,
1515 contradicting the general specification for all authenticators. Instead it
1516 was generating a temporary error. It now behaves as specified.
1518 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1519 (the order specifies the preference for clients). The order is now AES256,
1520 AES128, 3DES, ARCFOUR128.
1522 31. Small patch to Sieve code - explicitly set From: when generating an
1525 32. Exim crashed if a remote delivery caused a very long error message to be
1526 recorded - for instance if somebody sent an entire SpamAssassin report back
1527 as a large number of 550 error lines. This bug was coincidentally fixed by
1528 increasing the size of one of Exim's internal buffers (big_buffer) that
1529 happened as part of the Exiscan merge. However, to be on the safe side, I
1530 have made the code more robust (and fixed the comments that describe what
1533 33. Some experimental protocols are using DNS PTR records for new purposes. The
1534 keys for these records are domain names, not reversed IP addresses. The
1535 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1536 leaves it alone. Component reversal etc. now happens only for IP addresses.
1539 34. The host_aton() function is supposed to be passed a string that is known
1540 to be a valid IP address. However, in the case of IPv6 addresses, it was
1541 not checking this. This is a hostage to fortune. Exim now panics and dies
1542 if the condition is not met. A case was found where this could be provoked
1543 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1544 components; fortuitously, this particular loophole had already been fixed
1545 by change 4.50/55 or 4.44/33 above.
1547 If there are any other similar loopholes, the new check in host_aton()
1548 itself should stop them being exploited. The report I received stated that
1549 data on the command line could provoke the exploit when Exim was running as
1550 exim, but did not say which command line option was involved. All I could
1551 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1552 running as the user.
1555 35. There was a buffer overflow vulnerability in the SPA authentication code
1556 (which came originally from the Samba project). I have added a test to the
1557 spa_base64_to_bits() function which I hope fixes it.
1560 36. The daemon start-up calls getloadavg() while still root for those OS that
1561 need the first call to be done as root, but it missed one case: when
1562 deliver_queue_load_max is set with deliver_drop_privilege. This is
1563 necessary for the benefit of the queue runner, because there is no re-exec
1564 when deliver_drop_privilege is set.
1566 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1567 was in use, was not putting the data itself into the right store pool;
1568 consequently, it could be overwritten for a subsequent message in the same
1569 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1572 38. Sometimes the final signoff response after QUIT could fail to get
1573 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1574 < 0 before doing a fflush(). This bug looks as though it goes back to the
1575 introduction of TLS in release 3.20, but "sometimes" must have been rare
1576 because the tests only now provoked it.
1582 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1583 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1584 Now it returns an integer. A number of calls were still expecting a T/F
1585 return. Fortuitously, in all cases, the tests worked in OK situations,
1586 which is the norm. However, things would have gone wrong on any write
1587 failures on the smtp file descriptor. This function is used when sending
1588 messages over SMTP and also when doing verify callouts.
1590 2. When Exim is called to do synchronous delivery of a locally submitted
1591 message (the -odf or -odi options), it no longer closes stderr before doing
1594 3. Implemented the mua_wrapper option.
1596 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1598 5. Implemented the functions header_remove(), header_testname(),
1599 header_add_at_position(), and receive_remove_recipient(), and exported them
1602 6. If an ACL "warn" statement specified the addition of headers, Exim already
1603 inserted X-ACL-Warn: at the start if there was no header name. However, it
1604 was not making this test for the second and subsequent header lines if
1605 there were newlines in the string. This meant that an invalid header could
1606 be inserted if Exim was badly configured.
1608 7. Allow an ACL "warn" statement to add header lines at the start or after all
1609 the Received: headers, as well as at the end.
1611 8. Added the rcpt_4xx retry error code.
1613 9. Added postmaster_mailfrom=xxx to callout verification option.
1615 10. Added mailfrom=xxxx to the callout verification option, for verify=
1618 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1619 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1621 12. Inserted some casts to stop certain compilers warning when using pointer
1622 differences as field lengths or precisions in printf-type calls (mostly
1623 affecting debugging statements).
1625 13. Added optional readline() support for -be (dynamically loaded).
1627 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1628 same clock tick as a message's arrival, so that its received time was the
1629 same as the "first fail" time on the retry record, and that message
1630 remained on the queue past the ultimate address timeout, every queue runner
1631 would try a delivery (because it was past the ultimate address timeout) but
1632 after another failure, the ultimate address timeout, which should have then
1633 bounced the address, did not kick in. This was a "< instead of <=" error;
1634 in most cases the first failure would have been in the next clock tick
1635 after the received time, and all would be well.
1637 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1638 being recognized when the domain list was tested by the match_domain
1639 condition in an expansion string.
1641 16. Added the ${str2b64: operator.
1643 17. Exim was always calling setrlimit() to set a large limit for the number of
1644 processes, without checking whether the existing limit was already
1645 adequate. (It did check for the limit on file descriptors.) Furthermore,
1646 errors from getrlimit() and setrlimit() were being ignored. Now they are
1647 logged to the main and panic logs, but Exim does carry on, to try to do its
1648 job under whatever limits there are.
1650 18. Imported PCRE 5.0.
1652 19. Trivial typo in log message " temporarily refused connection" (the leading
1655 20. If the log selector return_path_on_delivery was set and an address was
1656 redirected to /dev/null, the delivery process crashed because it assumed
1657 that a return path would always be set for a "successful" delivery. In this
1658 case, the whole delivery is bypassed as an optimization, and therefore no
1661 21. Internal re-arrangement: the function for sending a challenge and reading
1662 a response while authentication was assuming a zero-terminated challenge
1663 string. It's now changed to take a pointer and a length, to allow for
1664 binary data in such strings.
1666 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1668 23. Exim was not respecting finduser_retries when seeking the login of the
1669 uid under which it was called; it was always trying 10 times. (The default
1670 setting of finduser_retries is zero.) Also, it was sleeping after the final
1671 failure, which is pointless.
1673 24. Implemented tls_on_connect_ports.
1675 25. Implemented acl_smtp_predata.
1677 26. If the domain in control=submission is set empty, Exim assumes that the
1678 authenticated id is a complete email address when it generates From: or
1679 Sender: header lines.
1681 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1682 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1683 chown and chgrp in /bin and hostname in /usr/bin.
1685 28. Exim was keeping the "process log" file open after each use, just as it
1686 does for the main log. This opens the possibility of it remaining open for
1687 long periods when the USR1 signal hits a daemon. Occasional processlog
1688 errors were reported, that could have been caused by this. Anyway, it seems
1689 much more sensible not to leave this file open at all, so that is what now
1692 29. The long-running daemon process does not normally write to the log once it
1693 has entered its main loop, and it closes the log before doing so. This is
1694 so that log files can straightforwardly be renamed and moved. However,
1695 there are a couple of unusual error situations where the daemon does write
1696 log entries, and I had neglected to close the log afterwards.
1698 30. The text of an SMTP error response that was received during a remote
1699 delivery was being truncated at 512 bytes. This is too short for some of
1700 the long messages that one sometimes sees. I've increased the limit to
1703 31. It is now possible to make retry rules that apply only when a message has a
1704 specific sender, in particular, an empty sender.
1706 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1707 it possible to be selective about when SMTP synchronization is enforced.
1709 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1711 32. Implemented hosts_connection_nolog.
1713 33. Added an ACL for QUIT.
1715 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1718 35. Added mailbox_size and mailbox_filecount to appendfile.
1720 36. Added control = no_multiline_responses to ACLs.
1722 37. There was a bug in the logic of the code that waits for the clock to tick
1723 in the case where the clock went backwards by a substantial amount such
1724 that the microsecond fraction of "now" was more than the microsecond
1725 fraction of "then" (but the whole seconds number was less).
1727 38. Added support for the libradius Radius client library this is found on
1728 FreeBSD (previously only the radiusclient library was supported).
1734 1. When certain lookups returned multiple values in the form name=value, the
1735 quoting of the values was not always being done properly. Specifically:
1736 (a) If the value started with a double quote, but contained no whitespace,
1738 (b) If the value contained whitespace other than a space character (i.e.
1739 tabs or newlines or carriage returns) it was not quoted.
1740 This fix has been applied to the mysql and pgsql lookups by writing a
1741 separate quoting function and calling it from the lookup code. The fix
1742 should probably also be applied to nisplus, ibase and oracle lookups, but
1743 since I cannot test any of those, I have not disturbed their existing code.
1745 2. A hit in the callout cache for a specific address caused a log line with no
1746 reason for rejecting RCPT. Now it says "Previous (cached) callout
1747 verification failure".
1749 3. There was an off-by-one bug in the queryprogram router. An over-long
1750 return line was truncated at 256 instead of 255 characters, thereby
1751 overflowing its buffer with the terminating zero. As well as fixing this, I
1752 have increased the buffer size to 1024 (and made a note to document this).
1754 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1755 when Exim is waiting for an SMTP response from a remote server, Exim
1756 restarts its select() call on the socket, thereby resetting its timeout.
1757 This is not a problem when such interrupts are rare. Somebody set up a cron
1758 job to run exiwhat every 2 minutes, which is less than the normal select()
1759 timeout (5 or 10 minutes). This meant that the select() timeout never
1760 kicked in because it was always reset. I have fixed this by comparing the
1761 time when an interrupt arrives with the time at the start of the first call
1762 to select(). If more time than the timeout has elapsed, the interrupt is
1763 treated as a timeout.
1765 5. Some internal re-factoring in preparation for the addition of Sieve
1766 extensions (by MH). In particular, the "personal" test is moved to a
1767 separate function, and given an option for scanning Cc: and Bcc: (which is
1768 not set for Exim filters).
1770 6. When Exim created an email address using the login of the caller as the
1771 local part (e.g. when creating a From: or Sender: header line), it was not
1772 quoting the local part when it contained special characters such as @.
1774 7. Installed new OpenBSD configuration files.
1776 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1777 try to make them clearer.
1779 9. Callout options, other than the timeout value, were being ignored when
1780 verifying sender addresses in header lines. For example, when using
1782 verify = header_sender/callout=no_cache
1784 the cache was (incorrectly) being used.
1786 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1787 only the Cygwin environment.
1789 11. When return_path_on_delivery was set as a log selector, if different remote
1790 addresses in the same message used different return paths and parallel
1791 remote delivery occurred, the wrong values would sometimes be logged.
1792 (Whenever a remote delivery process finished, the return path value from
1793 the most recently started remote delivery process was logged.)
1795 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1796 lines when AUTH and/or TLS are in use. This is the "received protocol"
1797 field. Exim used to use "asmtp" for authenticated SMTP, without any
1798 indication (in the protocol name) for TLS use. Now it follows the RFC and
1799 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1800 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1801 names appear in log lines as well as in Received: header lines.
1803 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1804 extensions, and comparison tests, and to fix some bugs.
1806 14. Changes to the "personal" filter test:
1808 (1) The test was buggy in that it was just doing the equivalent of
1809 "contains" tests on header lines. For example, if a user's address was
1810 anne@some.where, the "personal" test would incorrectly be true for
1812 To: susanne@some.where
1814 This test is now done by extracting each address from the header in turn,
1815 and checking the entire address. Other tests that are part of "personal"
1816 are now done using regular expressions (for example, to check local parts
1817 of addresses in From: header lines).
1819 (2) The list of non-personal local parts in From: addresses has been
1820 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1821 taken from the Sieve specification recommendations.
1823 (3) If the message contains any header line starting with "List-" it is
1824 treated as non-personal.
1826 (4) The test for "circular" in the Subject: header line has been removed
1827 because it now seems ill-conceived.
1829 15. Minor typos in src/EDITME comments corrected.
1831 16. Installed latest exipick from John Jetmore.
1833 17. If headers_add on a router specified a text string that was too long for
1834 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1835 of string_sprintf() is now avoided.
1837 18. $message_body_size was not set (it was always zero) when running the DATA
1838 ACL and the local_scan() function.
1840 19. For the "mail" command in an Exim filter, no default was being set for
1841 the once_repeat time, causing a random time value to be used if "once" was
1842 specified. (If the value happened to be <= 0, no repeat happened.) The
1843 default is now 0s, meaning "never repeat". The "vacation" command was OK
1844 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1845 (I found it when inspecting the code).
1847 20. There is now an overall timeout for performing a callout verification. It
1848 defaults to 4 times the callout timeout, which applies to individual SMTP
1849 commands during the callout. The overall timeout applies when there is more
1850 than one host that can be tried. The timeout is checked before trying the
1851 next host. This prevents very long delays if there are a large number of
1852 hosts and all are timing out (e.g. when the network connections are timing
1853 out). The value of the overall timeout can be changed by specifying an
1854 additional sub-option for "callout", called "maxwait". For example:
1856 verify = sender/callout=5s,maxwait=20s
1858 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1859 to the end before writing, but this should make it even safer).
1861 22. Exim was forgetting that it had advertised PIPELINING for the second and
1862 subsequent messages on an SMTP connection. It was also not resetting its
1863 memory on STARTTLS and an internal HELO.
1865 23. When Exim logs an SMTP synchronization error within a session, it now
1866 records whether PIPELINING has been advertised or not.
1868 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1869 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1870 rather than long int.
1872 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1874 26. Added the never_mail option to autoreply.
1880 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1881 crash if the getsockname() call failed; this can happen if a connection is
1882 closed very soon after it is established. The problem was simply in the
1883 order in which certain operations were done, causing Exim to try to write
1884 to the SMTP stream before it had set up the file descriptor. The bug has
1885 been fixed by making things happen in the correct order.
1891 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1892 before the connection was closed, thus losing the rejection response.
1894 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1895 some early Solaris releases, but causes trouble in current releases where
1896 socklen_t is defined.
1898 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1901 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1904 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1905 that releasing the top of it at the end releases what was used for sub-
1906 expansions (unless the block got too big). However, discard this block if
1907 the first thing is a variable or header, so that we can use its block when
1908 it is dynamic (useful for very large $message_headers, for example).
1910 6. Lookups now cache *every* query, not just the most recent. A new, separate
1911 store pool is used for this. It can be recovered when all lookup caches are
1912 flushed. Lookups now release memory at the end of their result strings.
1913 This has involved some general refactoring of the lookup sources.
1915 7. Some code has been added to the store_xxx() functions to reduce the amount
1916 of flapping under certain conditions.
1918 8. log_incoming_interface used to affect only the <= reception log lines. Now
1919 it causes the local interface and port to be added to several more SMTP log
1920 lines, for example "SMTP connection from", and rejection lines.
1922 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1924 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1926 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1927 could be overwritten at the end of the current message (or the start of a
1928 new message if it was set in a HELO ACL). The value is now preserved for
1929 the duration of the SMTP connection.
1931 12. If a transport had a headers_rewrite setting, and a matching header line
1932 contained an unqualified address, that address was qualified, even if it
1933 did not match any rewriting rules. The underlying bug was that the values
1934 of the flags that permit the existence of unqualified sender and recipient
1935 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1936 non-local messages, and by -bnq for local messages) were not being
1937 preserved with the message after it was received.
1939 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1940 "next input=" as part of the text comprising the host identity instead of
1941 the correct text. The code was using the same buffer for two different
1942 strings. However, depending on which order the printing function evaluated
1943 its arguments, the bug did not always show up. Under Linux, for example, my
1944 test suite worked just fine.
1946 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1947 doesn't work with some older versions of Perl. It has been changed to "my",
1948 which in any case is probably the better facility to use.
1950 15. A really picky compiler found some instances of statements for creating
1951 error messages that either had too many or two few arguments for the format
1954 16. The size of the buffer for calls to the DNS resolver has been increased
1955 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1956 for addresses that have a lot of PTR records. This alleviates a problem; it
1957 does not fully solve it.
1959 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1960 buffer now truncates the list and logs the incident, which is the same
1961 action as happens when Exim is looking up a host name and its aliases.
1962 Previously in this situation something unpredictable would happen;
1963 sometimes it was "internal error: store_reset failed".
1965 18. If a server dropped the connection unexpectedly when an Exim client was
1966 using GnuTLS and trying to read a response, the client delivery process
1967 crashed while trying to generate an error log message.
1969 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1970 string, for example:
1972 warn message = H1: something\nH2: something
1974 the text was added as a single header line from Exim's point of view
1975 though it ended up OK in the delivered message. However, searching for the
1976 second and subsequent header lines using $h_h2: did not work. This has been
1977 fixed. Similarly, if a system filter added multiple headers in this way,
1978 the routers could not see them.
1980 20. Expanded the error message when iplsearch is called with an invalid key to
1981 suggest using net-iplsearch in a host list.
1983 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1984 ACLs are no longer actually imposed (and a message to that effect is
1987 22. If a "gecos" field in a passwd entry contained escaped characters, in
1988 particular, if it contained a \" sequence, Exim got it wrong when building
1989 a From: or a Sender: header from that name. A second bug also caused
1990 incorrect handling when an unquoted " was present following a character
1991 that needed quoting.
1993 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1994 was not being matched caselessly.
1996 24. Arranged for all hyphens in the exim.8 source to be escaped with
1999 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2000 properly. Recipient callout cache records were still being keyed to include
2001 the sender, even when use_sender was set false. This led to far more
2002 callouts that were necessary. The sender is no longer included in the key
2003 when use_sender is false.
2005 26. Added "control = submission" modifier to ACLs.
2007 27. Added the ${base62d: operator to decode base 62 numbers.
2009 28. dnsdb lookups can now access SRV records.
2011 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2012 the configuration file.
2014 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2015 (-v) mode. This makes the output for a verbose queue run more intelligible.
2017 31. Added a use_postmaster feature to recipient callouts.
2019 32. Added the $body_zerocount variable, containing the number of binary zero
2020 bytes in the message body.
2022 33. The time of last modification of the "new" subdirectory is now used as the
2023 "mailbox time last read" when there is a quota error for a maildir
2026 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2028 35. Added +ignore_unknown as a special item in host lists.
2030 36. Code for decoding IPv6 addresses in host lists is now included, even if
2031 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2032 address was recognized as an IP address, but was then not correctly decoded
2033 into binary, causing unexpected and incorrect effects when compared with
2040 1. Very minor rewording of debugging text in manualroute to say "list of
2041 hosts" instead of "hostlist".
2043 2. If verify=header_syntax was set, and a header line with an unqualified
2044 address (no domain) and a large number of spaces between the end of the
2045 name and the colon was received, the reception process suffered a buffer
2046 overflow, and (when I tested it) crashed. This was caused by some obsolete
2047 code that should have been removed. The fix is to remove it!
2049 3. When running in the test harness, delay a bit after writing a bounce
2050 message to get a bit more predictability in the log output.
2052 4. Added a call to search_tidyup() just before forking a reception process. In
2053 theory, someone could use a lookup in the expansion of smtp_accept_max_
2054 per_host which, without the tidyup, could leave open a database connection.
2056 5. Added the variables $recipient_data and $sender_data which get set from a
2057 lookup success in an ACL "recipients" or "senders" condition, or a router
2058 "senders" option, similar to $domain_data and $local_part_data.
2060 6. Moved the writing of debug_print from before to after the "senders" test
2063 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2064 problems for message scanning, either using a data ACL, or using
2065 local_scan() because the Received: header was not generated till after they
2066 were called (in order to set the time as the time of reception completion).
2067 I have revised the way this works. The header is now generated after the
2068 body is received, but before the ACL or local_scan() are called. After they
2069 are run, the timestamp in the header is updated.
2075 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2076 before starting a queue runner without re-exec. This happened only when
2077 deliver_drop_privilege was set or when the Exim user was set to root. The
2078 effect of the bug was that timeouts during subsequent deliveries caused
2079 crashes instead of being properly handled. The handler is now left at its
2080 default (and expected) setting.
2082 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2083 message, again when deliver_drop_privilege is set or Exim is run as root.
2084 The bug described in (1) was not present in this case, but the tidying up
2085 of the other signals was missing. I have made the two cases consistent.
2087 3. The ignore_target_hosts setting on a manualroute router was being ignored
2088 for hosts that were looked up using the /MX notation.
2090 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2093 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2094 operated on the sender address. After changing the $sender_address to <>
2095 for the sender address verify, Exim was re-instated it as the original
2096 (before rewriting) address, but remembering that it had rewritten it, so it
2097 wasn't rewriting it again. This bug also had the effect of breaking the
2098 sender address verification caching when the sender address was rewritten.
2100 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2101 This has been changed so that if the ip literal address matches
2102 ignore_target_hosts, the router declines.
2104 7. Added expansion conditions match_domain, match_address, and match_local_
2105 part (NOT match_host).
2107 8. The placeholder for the Received: header didn't have a length field set.
2109 9. Added code to Exim itself and to exim_lock to test for a specific race
2110 condition that could lead to file corruption when using MBX delivery. The
2111 issue is with the lockfile that is created in /tmp. If this file is removed
2112 after a process has opened it but before that process has acquired a lock,
2113 there is the potential for a second process to recreate the file and also
2114 acquire a lock. This could lead to two Exim processes writing to the file
2115 at the same time. The added code performs the same test as UW imapd; it
2116 checks after acquiring the lock that its file descriptor still refers to
2117 the same named file.
2119 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2120 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2123 11. Added the smtp_active_hostname option. If used, this will typically be made
2124 to depend on the incoming interface address. Because $interface_address is
2125 not set up until the daemon has forked a reception process, error responses
2126 that can happen earlier (such as "too many connections") no longer contain
2129 12. If an expansion in a condition on a "warn" statement fails because a lookup
2130 defers, the "warn" statement is abandoned, and the next ACL statement is
2131 processed. Previously this caused the whole ACL to be aborted.
2133 13. Added the iplsearch lookup type.
2135 14. Added ident_timeout as a log selector.
2137 15. Added tls_certificate_verified as a log selector.
2139 16. Added a global option tls_require_ciphers (compare the smtp transport
2140 option of the same name). This controls incoming TLS connections.
2142 17. I finally figured out how to make tls_require_ciphers do a similar thing
2143 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2144 before starting the TLS session.
2146 18. Tabs are now shown as \t in -bP output.
2148 19. If the log selector return_path_on_delivery was set, Exim crashed when
2149 bouncing a message because it had too many Received: header lines.
2151 20. If two routers both had headers_remove settings, and the first one included
2152 a superfluous trailing colon, the final name in the first list and the
2153 first name in the second list were incorrectly joined into one item (with a
2154 colon in the middle).
2160 1. Added -C and -D options to the exinext utility, mainly to make it easier
2161 to include in the automated testing, but these could be helpful when
2162 multiple configurations are in use.
2164 2. The exinext utility was not formatting the output nicely when there was
2165 an alternate port involved in the retry record key, nor when there was a
2166 message id as well (for retries that were specific to a specific message
2167 and a specific host). It was also confused by IPv6 addresses, because of
2168 the additional colons they contain. I have fixed the IPv4 problem, and
2169 patched it up to do a reasonable job for IPv6.
2171 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2172 delivery, the log line now contains "pipelined" if PIPELINING was used.
2174 4. An SMTP transport process used to panic and die if the bind() call to set
2175 an explicit outgoing interface failed. This has been changed; it is now
2176 treated in the same way as a connect() failure.
2178 5. A reference to $sender_host_name in the part of a conditional expansion
2179 that was being skipped was still causing a DNS lookup. This no longer
2182 6. The def: expansion condition was not recognizing references to header lines
2183 that used bh_ and bheader_.
2185 7. Added the _cache feature to named lists.
2187 8. The code for checking quota_filecount in the appendfile transport was
2188 allowing one more file than it should have been.
2190 9. For compatibility with Sendmail, the command line option
2198 and sets the incoming protocol and host name (for trusted callers). The
2199 host name and its colon can be omitted when only the protocol is to be set.
2200 Note the Exim already has two private options, -pd and -ps, that refer to
2201 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2202 "s", but I don't think that's a major issue.
2204 10. A number of refactoring changes to the code, none of which should affect
2207 (a) The number of logging options was getting close to filling up the
2208 32-bit word that was used as a bit map. I have split them into two classes:
2209 those that are passed in the argument to log_write(), and those that are
2210 only ever tested independently outside of that function. These are now in
2211 separate 32-bit words, so there is plenty of room for expansion again.
2212 There is no change in the user interface or the logging behaviour.
2214 (b) When building, for example, log lines, the code previously used a
2215 macro that called string_cat() twice, in order to add two strings. This is
2216 not really sufficiently general. Furthermore, there was one instance where
2217 it was actually wrong because one of the argument was used twice, and in
2218 one call a function was used. (As it happened, calling the function twice
2219 did not affect the overall behaviour.) The macro has been replaced by a
2220 function that can join an arbitrary number of extra strings onto a growing
2223 (c) The code for expansion conditions now uses a table and a binary chop
2224 instead of a serial search (which was left over from when there were very
2225 few conditions). Also, it now recognizes conditions like "pam" even when
2226 the relevant support is not compiled in: a suitably worded error message is
2227 given if an attempt is made to use such a condition.
2229 11. Added ${time_interval:xxxxx}.
2231 12. A bug was causing one of the ddress fields not to be passed back correctly
2232 from remote delivery subprocesses. The field in question was not being
2233 subsequently used, so this caused to problems in practice.
2235 13. Added new log selectors queue_time and deliver_time.
2237 14. Might have fixed a bug in maildirsizefile handling that threw up
2238 "unexpected character" debug warnings, and recalculated the data
2239 unnecessarily. In any case, I expanded the warning message to give more
2242 15. Added the message "Restricted characters in address" to the statements in
2243 the default ACL that block characters like @ and % in local parts.
2245 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2246 Three changes have been made:
2248 (a) There was a serious bug; a negative response to MAIL caused the whole
2249 recipient domain to be cached as invalid, thereby blocking all messages
2250 to all local parts at the same domain, from all senders. This bug has
2251 been fixed. The domain is no longer cached after a negative response to
2252 MAIL if the sender used is not empty.
2254 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2257 (c) A new callout option, "use_sender" has been added for people who want
2258 the modified behaviour.
2264 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2267 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2268 indeed breaks things for older releases.
2270 3. Added additional logging to the case where there is a problem reading data
2271 from a filter that is running in a subprocess using a pipe, in order to
2272 try to track down a specific problem.
2274 4. Testing facility fudge: when running in the test harness and attempting
2275 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2276 getting "No route to host". Convert this to a timeout.
2278 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2281 6. Some OS don't have socklen_t but use size_t instead. This affects the
2282 fifth argument of getsockopt() amongst other things. This is now
2283 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2284 can be set for individual OS. I have set it for SunOS5, OSF1, and
2285 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2286 some earlier ones do not.
2288 7. Change 4.30/15 was not doing the test caselessly.
2290 8. The standard form for an IPv6 address literal was being rejected by address
2291 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2292 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2293 this, as well as the form without the "IPv6" on the front (but only when
2294 address literals are enabled, of course).
2296 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2298 10. Exim crashed if a message with an empty sender address specified by -f
2299 encountered a router with an errors_to setting. This could be provoked only
2300 by a command such as
2304 where an empty string was supplied; "<>" did not hit this bug.
2306 11. Installed PCRE release 4.5.
2308 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2309 remained set. It is now erased.
2311 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2312 times from message ids (which are base 36 rather than the normal 62).
2314 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2315 were being counted as actual protocol errors, and logged if the log
2316 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2317 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2318 and DATA following a set of rejected RCPTs do not count as protocol errors.
2319 In other words, Exim assumes they were pipelined, though this may not
2320 actually be the case. Of course, in all cases the client gets an
2321 appropriate error code.
2323 15. If a lookup fails in an ACL condition, a message about the failure may
2324 be available; it is used if testing the ACL cannot continue, because most
2325 such messages specify what the cause of the deferral is. However, some
2326 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2327 that caused an old message to be retained and used if a later statement
2328 caused a defer, replacing the real cause of the deferral.
2330 16. If an IP address had so many PTR records that the DNS lookup buffer
2331 was not large enough to hold them, Exim could crash while trying to process
2332 the truncated data. It now detects and logs this case.
2334 17. Further to 4.21/58, another change has been made: if (and only if) the
2335 first line of a message (the first header line) ends with CRLF, a bare LF
2336 in a subsequent header line has a space inserted after it, so as not to
2337 terminate the header.
2339 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2340 unnecessarily, used atoi() instead of strtol(), and didn't check the
2341 termination when getting file sizes from file names by regex.
2343 19. Completely re-implemented the support for maildirsize files, in the light
2344 of a number of problems with the previous contributed implementation
2345 (4.30/29). In particular:
2347 . If the quota is zero, the maildirsize file is maintained, but no quota is
2350 . If the maildir directory does not exist, it is created before any attempt
2351 to write a maildirsize file.
2353 . The quota value in the file is just a cache; if the quota is changed in
2354 the transport, the new value overrides.
2356 . A regular expression is available for excluding directories from the
2359 20. The autoreply transport checks the characters in options that define the
2360 message's headers; it allows continued headers, but it was checking with
2361 isspace() after an embedded newline instead of explicitly looking for a
2364 21. If all the "regular" hosts to which an address was routed had passed their
2365 expiry times, and had not reached their retry times, the address was
2366 bounced, even if fallback hosts were defined. Now Exim should go on to try
2369 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2370 equivalent code in the SMTP transport. Some hosts send humungous responses
2371 to HELO/EHLO, more than 1024 it seems.
2373 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2374 gives compiler warnings in some environments. I've now done it "properly",
2377 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2378 (because of problems with the built-in one) was declared to return uschar *
2379 instead of char *, causing compiler failure.
2381 25. Fixed a file descriptor leak when processing alias/forward files.
2383 26. Fixed a minor format string issue in dbfn.c.
2385 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2387 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2388 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2391 29. When a sender address is verified, it is cached, to save repeating the test
2392 when there is more than one recipient in a message. However, when the
2393 verification involves a callout, it is possible for different callout
2394 options to be set for different recipients. It is too complicated to keep
2395 track of this in the cache, so now Exim always runs a verification when a
2396 callout is required, relying on the callout cache for the optimization.
2397 The overhead is duplication of the address routing, but this should not be
2400 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2401 to be verified with callout=postmaster, and the main callout worked but the
2402 postmaster check failed, the verification correctly failed. However, if a
2403 subsequent RCPT command asked for sender verification *without* the
2404 postmaster check, incorrect caching caused this verification also to fail,
2407 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2408 it was not caching the DNS options (qualify_single, search_parents) that
2409 were used when the lookup failed. A subsequent lookup with different
2410 options therefore always gave the same answer, though there were cases
2411 where it should not have. (Example: a "domains = !$mx_any" option on a
2412 dnslookup router: the "domains" option is always processed without any
2413 widening, but the router might have qualify_single set.) Now Exim uses the
2414 cached value only when the same options are set.
2416 32. Added John Jetmore's "exipick" utility to the distribution.
2418 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2419 than a timeout (e.g. a certificate is required, and is not provided), an
2420 Exim server now closes the connection immediately. Previously it waited for
2421 the client to close - but if the client is SSL, it seems that they each
2422 wait for each other, leading to a delay before one of them times out.
2424 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2425 maintained 0.8.x compatibility because I don't think many are using it, and
2426 it is clearly obsolete.
2428 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2431 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2432 client certificate was expired. A simple patch fixes this, though I don't
2433 understand the full logic of why the verify callback is called multiple
2436 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2437 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2438 which causes problems with some clients (such as the Certicom SSL Plus
2439 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2440 disables the coutermeasure allowing Eudora to connect."
2442 38. Exim was not checking that a write() to a log file succeeded. This could
2443 lead to Bad Things if a log got too big, in particular if it hit a file
2444 size limit. Exim now panics and dies if it cannot write to a log file, just
2445 as it does if it cannot open a log file.
2447 39. Modified OS/Makefile-Linux so that it now contains
2449 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2451 The two -D definitions ensure that Exim is compiled with large file
2452 support, which makes it possible to handle log files that are bigger than
2455 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2456 instance) a domain was checked against a named list that involved a lookup,
2457 causing $domain_data to be set, then another domain was checked against the
2458 same list, then the first domain was re-checked, the value of $domain_data
2459 after the final check could be wrong. In particular, if the second check
2460 failed, it could be set empty. This bug probably also applied to
2463 41. The strip_trailing_dot option was not being applied to the address given
2464 with the -f command-line option.
2466 42. The code for reading a message's header from the spool was incrementing
2467 $received_count, but never initializing it. This meant that the value was
2468 incorrect (doubled) while delivering a message in the same process in which
2469 it was received. In the most common configuration of Exim, this never
2470 happens - a fresh exec is done - but it can happen when
2471 deliver_drop_privilege is set.
2473 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2474 it now includes up to 150 characters of the unexpected data in the log
2477 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2478 and building data strings. The size of both of these buffers was 10 000
2479 bytes - far larger than anybody would *ever* want, thought I. Needless to
2480 say, somebody hit the limit. I have increased the maximum line length to
2481 20 000 and the maximum data length of concatenated lines to 100 000. I have
2482 also fixed two bugs, because there was no checking on these buffers. Tsk,
2483 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2484 buffer is too small.
2486 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2487 lsearch lookups. Now it does.
2489 46. When parsing a route_list item in a manualroute router, a fixed-length
2490 buffer was used for the list of hosts. I made this 1024 bytes long,
2491 thinking that nobody would ever have a list of hosts that long. Wrong.
2492 Somebody had a whole pile of complicated expansion conditions, and the
2493 string was silently truncated, leading to an expansion error. It turns out
2494 that it is easier to change to an unlimited length (owing to other changes
2495 that have happened since this code was originally written) than to build
2496 structure for giving a limitation error. The length of the item that
2497 expands into the list of hosts is now unlimited.
2499 47. The lsearch lookup could not handle data where the length of text line was
2500 more than 4095 characters. Such lines were truncated, leading to shortened
2501 data being returned. It should now handle lines of any length.
2503 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2504 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2507 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2508 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2509 no longer happens in comment lines. A list of replacements is now placed
2510 at the head of all of the source files, except those whose only change is
2511 to replace PERL_COMMAND in the very first #! line.
2513 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2514 messages on the queue, with a bottom-up merge sort, using code contributed
2515 by Michael Haardt. This should make operations like -bp somewhat faster on
2516 large queues. It won't affect queue runners, except when queue_run_in_order
2519 51. Installed eximstats 1.31 in the distribution.
2521 52. Added support for SRV lookups to the dnslookup router.
2523 53. If an ACL referred to $message_body or $message_body_end, the value was not
2524 reset for any messages that followed in the same SMTP session.
2526 54. The store-handling optimization for building very long strings was not
2527 differentiating between the different store pools. I don't think this
2528 actually made any difference in practice, but I've tidied it.
2530 55. While running the routers to verify a sender address, $sender_address
2531 was still set to the sender address. This is wrong, because when routing to
2532 send a bounce to the sender, it would be empty. Therefore, I have changed
2533 it so that, while verifying a sender address, $sender_address is set to <>.
2534 (There is no change to what happens when verifying a recipient address.)
2536 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2537 target A or AAAA records (if not already returned) without resetting the
2538 qualify_single or search_parents options of the DNS resolver. These are
2539 inappropriate in this case because the targets of MX and SRV records must
2540 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2541 target that, when qualified, matched something in the local domain. These
2542 two options are now turned off when doing these lookups.
2544 57. It seems that at least some releases of Reiserfs (which does not have the
2545 concept of a fixed number of inodes) returns zero and not -1 for the
2546 number of available inodes. This interacted badly with check_spool_inodes,
2547 which assumed that -1 was the "no such thing" setting. What I have done is
2548 to check that the total number of inodes is greater than zero before doing
2549 the test of how many are available.
2551 58. When a "warn" ACL statement has a log_message modifier, the message is
2552 remembered, and not repeated. This is to avoid a lot of repetition when a
2553 message has many recipients that cause the same warning to be written.
2554 Howewer, Exim was preserving the list of already written lines for an
2555 entire SMTP session, which doesn't seem right. The memory is now reset if a
2556 new message is started.
2558 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2559 debugging output unless log_rewrite was also set.
2561 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2562 of (int)(handle) when we know that handle contains (void *)(-1).
2564 61. The Exim daemon panic-logs an error return when it closes the incoming
2565 connection. However "connection reset by peer" seems to be common, and
2566 isn't really an error worthy of noting specially, so that particular error
2569 62. When Exim is trying to find all the local interfaces, it used to panic and
2570 die if the ioctl to get the interface flags failed. However, it seems that
2571 on at least one OS (Solaris 9) it is possible to have an interface that is
2572 included in the list of interfaces, but for which you get a failure error
2573 for this call. This happens when the interface is not "plumbed" into a
2574 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2575 failure of the "get flags" call assumes that the interface is down.
2577 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2578 makes life easier for people who are doing arithmetic on fields extracted
2579 from dates, where you often get leading zeros that should not be
2580 interpreted as octal.
2582 64. Added qualify_domain to the redirect router, to override the global
2585 65. If a pathologically long header line contained very many addresses (the
2586 report of this problem mentioned 10 000) and each of them was rewritten,
2587 Exim could use up a very large amount of memory. (It kept on making new
2588 copies of the header line as it rewrote, and never released the old ones.)
2589 At the expense of a bit more processing, the header rewriting function has
2590 been changed so that it no longer eats memory in this way.
2592 66. The generation of the Received: header has been moved from the time that a
2593 message starts to be received, to the time that it finishes. The timestamp
2594 in the Received: header should now be very close to that of the <= log
2595 line. There are two side-effects of this change:
2597 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2598 logged header lines no longer include the local Received: line, because
2599 it has not yet been created. The same applies to a copy of the message
2600 that is returned to a non-SMTP sender when a message is rejected.
2602 (b) When a filter file is tested using -bf, no additional Received: header
2603 is added to the test message. After some thought, I decided that this
2606 This change does not affect the value of $received_for. It is still set
2607 after address rewriting, but before local_scan() is called.
2609 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2611 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2612 gave an unhelpful panic error message, and a defer error. I have managed to
2613 change this behaviour so that it now rejects any supplied certificate,
2614 which seems right, as the list of acceptable certificates is empty.
2616 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2617 gave an unhelpful defer error. I have not managed to make this reject any
2618 supplied certificates, but the error message it gives is "no certificate
2619 supplied", which is not helpful.
2621 70. exigrep's output now also includes lines that are not associated with any
2622 message, but which match the given pattern. Implemented by a patch from
2623 Martin Sluka, which also tidied up the Perl a bit.
2625 71. Recipient callout verification, like sender verification, was using <> in
2626 the MAIL FROM command. This isn't really the right thing, since the actual
2627 sender may affect whether the remote host accepts the recipient or not. I
2628 have changed it to use the actual sender in the callout; this means that
2629 the cache record is now keyed on a recipient/sender pair, not just the
2630 recipient address. There doesn't seem to be a real danger of callout loops,
2631 since a callout by the remote host to check the sender would use <>.
2632 [SEE ABOVE: changed after hitting problems.]
2634 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2635 temporary errors. However, in the case of such a code being given after
2636 the end of a data transmission (i.e. after ".") Exim was failing to write
2637 a retry record for the message. (Yes, there was some broken host that was
2638 actually sending 8xx at this point.)
2640 73. An unknown lookup type in a host list could cause Exim to panic-die when
2641 the list was checked. (An example that provoked this was putting <; in the
2642 middle of a list instead of at the start.) If this happened during a DATA
2643 ACL check, a -D file could be left lying around. This kind of configuration
2644 error no longer causes Exim to die; instead it causes a defer errror. The
2645 incident is still logged to the main and panic logs.
2647 74. Buglet left over from Exim 3 conversion. The message "too many messages
2648 in one connection" was written to the rejectlog but not the mainlog, except
2649 when address rewriting (yes!) was being logged.
2651 75. Added write_rejectlog option.
2653 76. When a system filter was run not as root (that is, when system_filter_user
2654 was set), the values of the $n variables were not being returned to the
2655 main process; thus, they were not subsequently available in the $sn
2658 77. Added +return_path_on_delivery log selector.
2660 78. A connection timeout was being treated differently from recipients deferred
2661 when testing hosts_max_try with a message that was older than the host's
2662 retry timeout. (The host should not be counted, thus allowing all hosts to
2663 be tried at least once before bouncing.) This may have been the cause of an
2664 occasionally reported bug whereby a message would remain on the queue
2665 longer than the retry timeout, but would be bounced if a delivery was
2666 forced. I say "may" because I never totally pinned down the problem;
2667 setting up timeout/retry tests is difficult. See also the next item.
2669 79. The ultimate address timeout was not being applied to errors that involved
2670 a combination of host plus message (for example, a timeout on a MAIL
2671 command). When an address resolved to a number of possible hosts, and they
2672 were not all tried for each delivery (e.g. because of hosts_max_try), a
2673 message could remain on the queue longer than the retry timeout.
2675 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2678 81. Fixed an obscure SMTP outgoing bug which required at least the following
2679 conditions: (a) there was another message waiting for the same server;
2680 (b) the server returned 5xx to all RCPT commands in the first message so
2681 that the message was not completed; (c) the server dropped the connection
2682 or gave a negative response to the RSET that Exim sends to abort the
2683 transaction. The observed case was a dropped connection after DATA that had
2684 been sent in pipelining mode. That is, the server had advertised PIPELINING
2685 but was not implementing it correctly. The effect of the bug was incorrect
2686 behaviour, such as trying another host, and this could lead to a crash.
2692 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2693 and daemon.c were passed as pointers to ints; they should have been
2694 pointers to socklen_t variables (which are typically unsigned ints).
2696 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2699 3. Fixed a really odd bug that affected only the testing scheme; patching a
2700 certain fixed string in the binary changed the value of another string that
2701 happened to be identical to the end of the original first string.
2703 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2704 name", it returns that address as the IP address. On some operating
2705 systems (e.g. Solaris), it also passes back the IP address string as the
2706 "host name". However, on others (e.g. Linux), it passes back an empty
2707 string. Exim wasn't checking for this, and was changing the host name to an
2708 empty string, assuming it had been canonicized.
2710 5. Although rare, it is permitted to have more than one PTR record for a given
2711 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2712 all the names associated with an address, because they do in Solaris.
2713 However, it seems that they do not in Linux for data that comes from the
2714 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2715 I found this out when I moved to a new Linux workstation and tried to run
2716 the Exim test suite.
2718 To get round this problem I have changed the code so that it now does its
2719 own call to the DNS to look up PTR records when searching for a host name.
2720 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2721 addresses that are only in /etc/hosts are still found.
2723 This behaviour is, however, controlled by an option called host_lookup_
2724 order, which defaults to "bydns:byaddr". If people want to use the other
2725 order, or indeed, just use one or the other means of lookup, they can
2726 specify it in this variable.
2728 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2729 some operating systems, this comes back from gethostbyaddr() as an empty
2730 string, and this is what Exim used to test for. However, it seems that in
2731 other systems, "." is yielded. Exim now tests for this case too.
2733 7. The values of check_spool_space and check_log_space are now held internally
2734 as a number of kilobytes instead of an absolute number of bytes. If a
2735 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2736 kilobyte. This means that much larger values can be stored.
2738 8. Exim monitor: an attempt to get the action menu when not actually pointing
2739 at a message produces an empty menu entitled "No message selected". This
2740 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2741 no entries in it ("Shell widget menu has zero width and/or height"). So I
2742 have added a single, blank menu entry in this case.
2744 9. Added ${quote_local_part.
2746 10. MIME decoding is now applied to the contents of Subject: header lines when
2749 11. Now that a reference to $sender_host_address automatically causes a reverse
2750 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2751 host lookup before query-style lookups in lists that might use this
2752 variable. This has therefore been abolished, and the "net-" prefix is no
2753 longer necessary for query-style lookups.
2755 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2756 have been a typo for LFLAGS, so it has been changed.
2758 13. The install script calls Exim with "-C /dev/null" in order to find the
2759 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2760 to be output. Howeve, since Exim outputs its version number before the
2761 error, it didn't break the script. It just looked ugly. I fixed this by
2762 always allowing "-C /dev/null" if the caller is root.
2764 14. Ignore overlarge ACL variable number when reading spool file - insurance
2765 against a later release with more variables having written the file.
2767 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2768 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2769 this, as well as the form without the "IPv6" on the front.
2771 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2772 OS/Makefile-Darwin file.
2774 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2775 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2777 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2778 variable. However, some OS use ints for inodes. I've added cast to long int
2779 to get rid of the compiler warning.
2781 19. I had forgotten to lock out "/../" in configuration file names when
2782 ALT_CONFIG_PREFIX was set.
2784 20. Routers used for verification do not need to specify transports. However,
2785 if such a router generated a host list, and callout was configured, Exim
2786 crashed, because it could not find a port number from the (non-existent)
2787 transport. It now assumes port 25 in this circumstance.
2789 21. Added the -t option to exigrep.
2791 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2792 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2793 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2794 them is set, LOOKUP_LSEARCH is forced.
2796 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2798 24. Added sender and host information to the "rejected by local_scan()" log
2799 line; previously there was no indication of these.
2801 25. Added .include_if_exists.
2803 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2804 Linux. It turns out that not all file systems support this. Apparently some
2805 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2806 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2807 not supported on the file descriptor, is now ignored when Exim is trying to
2808 sync a directory. This applies only to Linux.
2810 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2812 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2813 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2816 29. Added support for maildirsize files from supplied patch (modified a bit).
2818 30. The use of :fail: followed by an empty string could lead Exim to respond to
2819 sender verification failures with (e.g.):
2821 550 Verification failed for <xxx>
2822 550 Sender verify failed
2824 where the first response line was missing the '-' that indicates it is not
2825 the final line of the response.
2827 31. The loop for finding the name of the user that called Exim had a hardwired
2828 limit of 10; it now uses the value of finduser_retries, which is used for
2829 all other user lookups.
2831 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2834 33. Exim was neglecting to zero errno before one call of strtol() when
2835 expanding a string and expecting an integer value. On some systems this
2836 resulted in spurious "integer overflow" errors. Also, it was casting the
2837 result into an int without checking.
2839 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2840 did not work. The code looks as if it has *never* worked, though it appears
2841 to have been documented since at least releast 1.62. I have made it work.
2843 35. The "timeout_DNS" error in retry rules, also documented since at least
2844 1.62, also never worked. As it isn't clear exactly what this means, and
2845 clearly it isn't a major issue, I have abolished the feature by treating it
2846 as "timeout", and writing a warning to the main and panic logs.
2848 36. The display of retry rules for -brt wasn't always showing the error code
2851 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2852 timeout_connect_A, timeout_connect_MX.
2854 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2855 to the empty sender.
2857 39. The daemon was not analysing the content of -oX till after it had closed
2858 stderr and disconnected from the controlling terminal. This meant that any
2859 syntax errors were only noted on the panic log, and the return code from
2860 the command was 0. By re-arranging the code a little, I've made the
2861 decoding happen first, so such errors now appear on stderr, and the return
2862 code is 1. However, the actual setting up of the sockets still happens in
2863 the disconnected process, so errors there are still only recorded on the
2866 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2867 connections (as happens on some IP stacks) was logged at start up time as
2868 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2869 it from "IPv6 and IPv4", which means that two separate sockets are being
2872 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2873 says whether AF_INET or AF_INET6 was passed as an argument.
2875 42. Exiwhat output was messed up when time zones were included in log
2878 43. Exiwhat now gives more information about the daemon's listening ports,
2879 and whether -tls-on-connect was used.
2881 44. The "port" option of the smtp transport is now expanded.
2883 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2884 silently ignored. Now an error message is written to the main and panic
2887 46. There's a new ACL modifier called "logwrite" which writes to a log file
2888 as soon as it is encountered.
2890 47. Added $local_user_uid and $local_user_gid at routing time.
2892 48. Exim crashed when trying to verify a sender address that was being
2895 49. Exim was recognizing only a space character after ".include". It now also
2896 recognizes a tab character.
2898 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2899 extracting the relevant information from the specification. The man page no
2900 longer contains scrambled data for the -d option, and I've added a section
2901 at the front about calling Exim under different names.
2903 51. Added "extra_headers" argument to the "mail" command in filter files.
2905 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2908 53. Installed eximstats 1.29.
2910 54. Added transport_filter_timeout as a generic transport option.
2912 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2913 Cc: header lines. This was required by RFC 822, but it not required by RFC
2916 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2917 incoming messages that did not have them. Now it does so only if the
2918 message originates locally, that is, if there is no associated remote host
2919 address. When Resent- header lines are present, this applies to the Resent-
2920 lines rather than the non-Resent- lines.
2922 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2923 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2925 58. Messages for configuration errors now include the name of the main
2926 configuration file - useful now that there may be more than one file in a
2927 list (.included file names were always shown).
2929 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2930 for those rare installations that do not start the daemon as root or run it
2931 setuid root. I've cut out the call to initgroups() if the daemon is not
2934 60. The Exim user and group can now be bound into the binary as text strings
2935 that are looked up at the start of Exim's processing.
2937 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2939 62. Added $mailstore_basename variable.
2941 63. Installed patch to sieve.c from Michael Haardt.
2943 64. When Exim failed to open the panic log after failing to open the main log,
2944 the original message it was trying to log was written to stderr and debug
2945 output, but if they were not available (the usual case in production), it
2946 was lost. Now it is written to syslog before the two lines that record the
2947 failures to open the logs.
2949 65. Users' Exim filters run in subprocesses under the user's uid. It is
2950 possible for a "deliver" command or an alias in a "personal" command to
2951 provoke an address rewrite. If logging of address rewriting is configured,
2952 this fails because the process is not running as root or exim. There may be
2953 a better way of dealing with this, but for the moment (because 4.30 needs
2954 to be released), I have disabled address rewrite logging when running a
2955 filter in a non-root, non-exim process.
2961 1. The buildconfig auxiliary program wasn't quoting the value set for
2962 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2963 not defined. This bug was masked in 4.22 by the effect that was fixed in
2966 2. Some messages that were rejected after a message id was allocated were
2967 shown as "incomplete" by exigrep. It no longer does this for messages that
2968 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2970 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2971 have allow_domain_literals set, the ID did not get logged in the <= line.
2972 Domain literals are now always recognized in Message-ID: header lines.
2974 4. The first argument for a ${extract expansion item is the key name or field
2975 number. Leading and trailing spaces in this item were not being ignored,
2976 causing some misleading effects.
2978 5. When deliver_drop_privilege was set, single queue runner processes started
2979 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2980 same command in the process it spins off) were not dropping privilege.
2982 6. When the daemon running as "exim" started a queue runner, it always
2983 re-executed Exim in the spun-off process. This is a waste of effort when
2984 deliver_drop_privilege is set. The new process now just calls the
2985 queue-runner function directly.
2991 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2994 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2995 ignored. Though the use of -oP was forcing the writing of a pid file, it
2996 was always written to the default place.
2998 3. If the message "no IP address found for host xxxx" is generated during
2999 incoming verification, it is now followed by identification of the incoming
3000 connection (so you can more easily find what provoked it).
3002 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3004 5. Added some features to "harden" Exim a bit more against certain attacks:
3006 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3007 be put in Local/Makefile. This is like the never_users runtime option,
3008 but it cannot be overridden. The default setting is "root".
3010 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3011 prefix string with which any file named in a -C command line option
3014 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3015 is retained for -C and -D only if the caller of Exim is root. Without
3016 it, the exim user may also use -C and -D and retain privilege.
3018 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3019 command line option is disabled.
3021 6. Macro names set by the -D option must start with an upper case letter, just
3022 like macro names defined in the configuration file.
3024 7. Added "dereference=" facility to LDAP.
3026 8. Two instances of the typo "uknown" in the source files are fixed.
3028 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3029 the Configure-Makefile script screwed up while processing it.
3031 10. Incorporated PCRE 4.4.
3033 11. The SMTP synchronization check was not operating right at the start of an
3034 SMTP session. For example, it could not catch a HELO sent before the client
3035 waited for the greeting. There is now a check for outstanding input at the
3036 point when the greeting is written. Because of the duplex, asynchronous
3037 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3038 way, but not yet received, when the check is performed.
3040 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3041 on TCP/IP sockets, because this apparently causes some broken clients to
3044 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3045 unchanged) from the Cygwin maintainer.
3047 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3048 was supported instead of testing for mbx. Effectively a typo.
3050 15. The spa authenticator server code was not checking that the input it
3051 received was valid base64.
3053 16. The debug output line for the "set" modifier in ACLs was not showing the
3054 name of the variable that was being set.
3056 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3058 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3059 lookup on its own. Something else was needed to trigger the lookup. For
3060 example, a match in host_lookup or the need for a host name in a host list.
3061 Now, if $sender_host_name is referenced and the host name has not yet been
3062 looked up, a lookup is performed. If the lookup fails, the variable remains
3063 empty, and $host_lookup_failed is set to "1".
3065 19. Added "eqi" as a case-independent comparison operator.
3067 20. The saslauthd authentication condition could segfault if neither service
3068 nor realm was specified.
3070 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3071 error message that was logged was misleading, and incoming SMTP
3072 connections were dropped. The message is now more accurate, and temporary
3073 errors are given to SMTP connections.
3075 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3076 (except RSET) with a 421 error, until QUIT is received. However, it was
3077 failing to send a response to QUIT.
3079 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3080 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3081 ACL is now run after the helo_try_verify_hosts code.
3083 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3084 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3085 case-independent, so other case variants are also recognized). Apparently
3086 some systems use these upper case variants.
3088 25. If more than two messages were waiting for the same host, and a transport
3089 filter was specified for the transport, Exim sent two messages over the
3090 same TCP/IP connection, and then failed with "socket operation on non-
3091 socket" when it tried to send the third.
3093 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3095 27. The extern definition of crypt16() in expand.c was not being excluded when
3096 the OS had its own crypt16() function.
3098 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3099 as a preferred synonym for return_size_limit, both as an option and as an
3102 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3104 30. Changed the default configuration ACL to relax the local part checking rule
3105 for addresses that are not in any local domains. For these addresses,
3106 slashes and pipe symbols are allowed within local parts, but the sequence
3107 /../ is explicitly forbidden.
3109 31. SPA server authentication was not clearing the challenge buffer before
3112 32. log_message in a "warn" ACL statement was writing to the reject log as
3113 well as to the main log, which contradicts the documentation and doesn't
3114 seem right (because no rejection is happening). So I have stopped it.
3116 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3117 However, I am unable to do any testing of this.
3119 34. Fixed an infelicity in the appendfile transport. When checking directories
3120 for a mailbox, to see if any needed to be created, it was accidentally
3121 using path names with one or more superfluous leading slashes; tracing
3122 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3124 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3125 contents are added to the log line that is written for every discarded
3126 recipient. (Previously a log_message setting was ignored.)
3128 36. The ${quote: operator now quotes the string if it is empty.
3130 37. The install script runs exim in order to find its version number. If for
3131 some reason other than non-existence or emptiness, which it checks, it
3132 could not run './exim', it was installing it with an empty version number,
3133 i.e. as "exim-". This error state is now caught, and the installation is
3136 38. An argument was missing from the function that creates an error message
3137 when Exim fails to connect to the socket for saslauthd authentication.
3138 This could cause Exim to crash, or give a corrupted message.
3140 39. Added isip, isip4, and isip6 to ${if conditions.
3142 40. The ACL variables $acl_xx are now saved with the message, and can be
3143 accessed later in routers, transports, and filters.
3145 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3146 strings in the file are not string-expanded.
3148 42. If a MAIL command specified a SIZE value that was too large to fit into an
3149 int variable, the check against message_size_limit failed. Such values are
3150 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3151 day this will have to be increased, but I don't think I want to be around
3152 when emails are that large.
3159 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3160 iconv() is not standard in FreeBSD.
3162 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3163 IPv6 enabled. The observed symptom was a segmentation fault on return from
3164 the function os_common_find_running_interfaces() in src/os.c.
3166 3. In the check_special_case() function in daemon.c I had used "errno" as an
3167 argument name, which causes warnings on some systems. This was basically a
3168 typo, since it was named "eno" in the comments!
3170 4. The code that waits for the clock to tick (at a resolution of some fraction
3171 of a second) so as to ensure message-id uniqueness was always waiting for
3172 at least one whole tick, when it could have waited for less. [This is
3173 almost certainly not relevant at current processor speeds, where it is
3174 unlikely to ever wait at all. But we try to future-proof.]
3176 5. The function that sleeps for a time interval that includes fractions of a
3177 second contained a race. It did not block SIGALRM between setting the
3178 timer, and suspending (a couple of lines later). If the interval was short
3179 and the sigsuspend() was delayed until after it had expired, the suspension
3180 never ended. On busy systems this could lead to processes getting stuck for
3183 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3184 process, before it forks any delivery processes. The open lookup caching
3185 mechanism meant that the open file or database connection was passed into
3186 the delivery process. The problem was that delivery processes always tidy
3187 up cached lookup data. This could cause a problem for the next delivery
3188 process started by the queue runner, because the external queue runner
3189 process does not know about the closure. So the next delivery process
3190 still has data in the lookup cache. In the case of a file lookup, there was
3191 no problem because closing a file descriptor in a subprocess doesn't affect
3192 the parent. However, if the lookup was caching a connection to a database,
3193 the connection was closed, and the second delivery process was likely to
3194 see errors such as "PGSQL: query failed: server closed the connection
3195 unexpectedly". The problem has been fixed by closing all cached lookups
3196 in a queue runner before running a delivery process.
3198 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3199 seem to have the "const" qualifier which it has on other OS. I've
3202 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3203 *of the same type* (client or server) with the same public name that an
3204 error should be diagnosed.
3206 9. When Exim looked up a host name for an IP address, but failed to find the
3207 original IP address when looking up the host name (a safety check), it
3208 output the message "<ip address> does not match any IP for NULL", which was
3209 confusing, to say the least. The bug was that the host name should have
3210 appeared instead of "NULL".
3212 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3213 user that is built into the binary, and the -C or -D options is used, root
3214 privilege is dropped before the configuration file is read. In addition,
3215 logging is switched to stderr instead of the normal log files. If the
3216 configuration then re-defines the Exim user, the unprivileged environment
3217 is probably not what is expected, so Exim logs a panic warning message (but
3220 However, if deliver_drop_privilege is set, the unprivileged state may well
3221 be exactly what is intended, so the warning has been cut out in that case,
3222 and Exim is allowed to try to write to its normal log files.
3228 1. smtp_return_error_details was not giving details for temporary sender
3229 or receiver verification errors.
3231 2. Diagnose a configuration error if two authenticators have the same public
3234 3. Exim used not to create the message log file for a message until the first
3235 delivery attempt. This could be confusing when incoming messages were held
3236 for policy or load reasons. The message log file is now created at the time
3237 the message is received, and an initial "Received" line is written to it.
3239 4. The automatically generated man page for command line options had a minor
3240 bug that caused no ill effects; however, a more serious problem was that
3241 the procedure for building the man page automatically didn't always
3242 operate. Consequently, release 4.20 contains an out-of-date version. This
3243 shouldn't happen again.
3245 5. When building Exim with embedded Perl support, the script that builds the
3246 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3247 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3249 6. The freeze_tell option was not being used for messages that were frozen on
3250 arrival, either by an ACL or by local_scan().
3252 7. Added the smtp_incomplete_transaction log selector.
3254 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3255 was accepting AUTH without a new EHLO.
3257 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3258 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3261 10. Logging of TCP/IP connections (when configured) now happens in the main
3262 daemon process instead of the child process, so that the TCP/IP connection
3263 count is more accurate (but it can never be perfect).
3265 11. The use of "drop" in a nested ACL was not being handled correctly in the
3266 outer ACL. Now, if condition failure induced by the nested "drop" causes
3267 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3268 or "require"), the connection is dropped.
3270 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3271 that yield "discard" can now be used with an "accept" or a "discard" verb,
3272 but an error is generated for any others (because I can't see a useful way
3273 to define what should happen).
3275 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3276 are now processed in the same way as lines in the Exim configuration file.
3277 In particular, continuation lines are supported.
3279 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3281 15. Added -ti meaning -t -i.
3283 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3284 domains, and warn by logging if others are found.
3286 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3287 returned from GIFCONF when Exim is trying to find the list of interfaces on
3288 a host. The code in os.c has been modified to copy each ifreq to an aligned
3289 structure in all cases.
3291 Also, in some cases, the returned ifreq's were being copied to a 'struct
3292 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3293 means the last couple of bytes of an IPv6 address could be chopped if the
3294 ifreq contained only a normal sockaddr (14 bytes storage).
3296 18. Named domain lists were not supported in the hosts_treat_as_local option.
3297 An entry such as +xxxx was not recognized, and was treated as a literal
3300 19. Ensure that header lines added by a DATA ACL are included in the reject log
3301 if the ACL subsequently rejects the message.
3303 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3304 MD5 (which is deprecated).
3306 21. When testing a filter file using -bf, Exim was writing a message when it
3307 took the sender from a "From " line in the message, but it was not doing so
3308 when it took $return_path from a Return-Path: header line. It now does.
3310 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3311 with a valid header line field name (a series of printing characters
3312 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3314 23. Changed "disc" in the source to "disk" to conform to the documentation and
3315 the book and for uniformity.
3317 24. Ignore Sendmail's -Ooption=value command line item.
3319 25. When execve() failed while trying to run a command in a pipe transport,
3320 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3321 could be confused with a return value of 69 from the command itself. This
3322 has been changed to 127, the value the shell returns if it is asked to run
3323 a non-existent command. The wording for the related log line suggests a
3324 non-existent command as the problem.
3326 26. If received_header_text expands to an empty string, do not add a Received:
3327 header line to the message. (Well, it adds a token one on the spool, but
3328 marks it "old" so that it doesn't get used or transmitted.)
3330 27. Installed eximstats 1.28 (addition of -nt option).
3332 28. There was no check for failure on the call to getsockname() in the daemon
3333 code. This can fail if there is a shortage of resources on the system, with
3334 ENOMEM, for example. A temporary error is now given on failure.
3336 29. Contrary to the C standard, it seems that in some environments, the
3337 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3338 program. Exim now does this explicitly; it affects the formatting of
3339 timestamps using strftime().
3341 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3342 complaints. I've now initialized all the variables, to avoid this.
3344 32. Header lines added by a system filter were not being "seen" during
3345 transport-time rewrites.
3347 33. The info_callback() function passed to OpenSSL is set up with type void
3348 (*)(SSL *, int, int), as described somewhere. However, when calling the
3349 function (actually a macro) that sets it up, the type void(*)() is
3350 expected. I've put in a cast to prevent warnings from picky compilers.
3352 34. If a DNS black list lookup found a CNAME record, but there were no A
3353 records associated with the domain it pointed at, Exim crashed.
3355 35. If a DNS black list lookup returned more than one A record, Exim ignored
3356 all but the first. It now scans all returned addresses if a particular IP
3357 value is being sought. In this situation, the contents of the
3358 $dnslist_value variable are a list of all the addresses, separated by a
3361 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3362 to accept a host name and all its aliases if the forward lookup for any of
3363 them yielded the IP address of the incoming connection. Now it accepts only
3364 those names whose forward lookup yields the correct IP address. Any other
3365 names are discarded. This closes a loophole whereby a rogue DNS
3366 administrator could create reverse DNS records to break through a
3367 wildcarded host restriction in an ACL.
3369 37. If a user filter or a system filter that ran in a subprocess used any of
3370 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3371 the wrong values were passed to the pipe command ($thisaddress had the
3372 value of $0, $0 had the value of $1, etc). This bug was introduced by
3373 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3375 38. Improved the line breaking for long SMTP error messages from ACLs.
3376 Previously, if there was no break point between 40 and 75 characters, Exim
3377 left the rest of the message alone. Two changes have been made: (a) I've
3378 reduced the minimum length to 35 characters; (b) if it can't find a break
3379 point between 35 and 75 characters, it looks ahead and uses the first one
3380 that it finds. This may give the occasional overlong line, but at least the
3381 remaining text gets split now.
3383 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3384 file descriptors might be low, and that setting 1000 would always raise it.
3385 It turns out that in some environments, the limit is already over 1000 and
3386 that lowering it causes trouble. So now Exim takes care not to decrease it.
3388 40. When delivering a message, the value of $return_path is set to $sender_
3389 address at the start of routing (routers may change the value). By an
3390 oversight, this default was not being set up when an address was tested by
3391 -bt or -bv, which affected the outcome if any router or filter referred to
3394 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3395 log only when "message" or "log_message" are set. However, if one of the
3396 conditions was an address verification, or a call to a nested ACL, the
3397 messages generated by the underlying test were being passed through. This
3398 no longer happens. The underlying message is available in $acl_verify_
3399 message for both "message" and "log_message" expansions, so it can be
3400 passed through if needed.
3402 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3403 new expansion $bh_ to give the encoded byte string without charset
3404 translation. Translation happens only if iconv() is available; HAVE_ICONV
3405 indicates this at build time. HEADERS_CHARSET gives the charset to
3406 translate to; headers_charset can change it in the configuration, and
3407 "headers charset" can change it in an individual filter file.
3409 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3410 that creates RFC 2047 encoded "words" labels them as that charset instead
3411 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3412 expansion operator; (ii) when Exim creates a From: line for a local
3413 message; (iii) when a header line is rewritten to include a "phrase" part.
3415 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3416 buggy, causing it to skip the first lines of messages whose message ID
3417 ended in 'D'. This would not have bitten before Exim release 4.14, because
3418 message IDs were unlikely to end in 'D' before then. The effect was to have
3419 incorrect size information for certain domains.
3421 45. #include "config.h" was missing at the start of the crypt16.c module. This
3422 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3425 46. If there was a timeout during a "random" callout check, Exim treated it as
3426 a failure of the random address, and carried on sending RSET and the real
3427 address. If the delay was just some slowness somewhere, the response to the
3428 original RCPT would be taken as a response to RSET and so on, causing
3429 mayhem of various kinds.
3431 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3432 when I implemented it. It didn't allow for the fact that some option values
3433 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3434 the right test for positive values.
3436 48. Domain names in DNS records are case-independent. Exim always looks them up
3437 in lower case. Some resolvers return domain names in exactly the case they
3438 appear in the zone file, that is, they may contain uppercase letters. Not
3439 all resolvers do this - some return always lower case. Exim was treating a
3440 change of case by a resolver as a change of domain, similar to a widening
3441 of a domain abbreviation. This triggered its re-routing code and so it was
3442 trying to route what was effectively the same domain again. This normally
3443 caused routing to fail (because the router wouldn't handle the domain
3444 twice). Now Exim checks for this case specially, and just changes the
3445 casing of the domain that it ultimately uses when it transmits the message
3448 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3451 50. If a filter generated a file delivery with a non-absolute name (possible if
3452 no home directory exists for the router), the forbid_file option was not
3455 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3456 the existing equality matching.
3458 52. Exim was using ints instead of ino_t variables in some places where it was
3459 dealing with inode numbers.
3461 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3462 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3463 called TMPDIR, and if it finds it is different, it changes its value.
3465 54. The smtp_printf() function is now made available to local_scan() so
3466 additional output lines can be written before returning. There is also an
3467 smtp_fflush() function to enable the detection of a dropped connection.
3468 The variables smtp_input and smtp_batched_input are exported to
3471 55. Changed the default runtime configuration: the message "Unknown user"
3472 has been removed from the ACL, and instead placed on the localuser router,
3473 using the cannot_route_message feature. This means that any verification
3474 failures that generate their own messages won't get overridden. Similarly,
3475 the "Unrouteable address" message that was in the ACL for unverifiable
3476 relay addresses has also been removed.
3478 56. Added hosts_avoid_esmtp to the smtp transport.
3480 57. The exicyclog script was not checking for the esoteric option
3481 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3482 will work only if exicyclog is run under the appropriate euid.
3484 58. Following a discussion on the list, the rules by which Exim recognises line
3485 endings on incoming messages have been changed. The -dropcr and drop_cr
3486 options are now no-ops, retained only for backwards compatibility. The
3487 following line terminators are recognized: LF CRLF CR. However, special
3488 processing applies to CR:
3490 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3491 nor a local message in the state where . is a terminator.
3493 (ii) If a bare CR is encountered in a header line, an extra space is added
3494 after the line terminator so as not to end the header. The reasoning
3495 behind this is that bare CRs in header lines are most likely either
3496 to be mistakes, or people trying to play silly games.
3498 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3499 was being incorrectly given as 18 bytes larger than it should have been.
3500 This is a VOB (very old bug).
3502 60. This may never have affected anything current, but just in case it has:
3503 When the local host is found other than at the start of a list of hosts,
3504 the local host, those with the same MX, and any that follow, are discarded.
3505 When the list in question was part of a longer list of hosts, the following
3506 hosts (not currently being processed) were also being discarded. This no
3507 longer happens. I'm not sure if this situation could ever has previously
3510 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3513 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3514 header. This is something that is recommended in a new Internet Draft, and
3515 is something that is documented as being done by Sendmail. There are two
3516 possible values. For messages generated by the autoreply transport, Exim
3519 Auto-Submitted: auto-replied
3521 whereas for all other generated messages (e.g. bounces) it adds
3523 Auto-Submitted: auto-generated
3525 63. The "personal" condition in filters now includes a test for the
3526 Auto-Submitted: header. If it contains the string "auto-" the message it
3527 not considered personal.
3529 64. Added rcpt_include_affixes as a generic transport option.
3531 65. Added queue_only_override (default true).
3533 66. Added the syslog_duplication option.
3535 67. If what should have been the first header line of a message consisted of
3536 a space followed by a colon, Exim was mis-interpreting it as a header line.
3537 It isn't of course - it is syntactically invalid and should therefore be
3538 treated as the start of the message body. The misbehaviour could have
3539 caused a number of strange effects, including loss of data in subsequent
3540 header lines, and spool format errors.
3542 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3543 client host had authenticated. This control can now be exercised by an ACL
3544 for more flexibility.
3546 69. By default, callouts do not happen when testing with -bh. There is now a
3547 variant, -bhc, which does actually run the callout code, including
3548 consulting and updating the callout cache.
3550 70. Added support for saslauthd authentication, courtesy of Alexander
3553 71. If statvfs() failed on the spool or log directories while checking their
3554 size for availability, Exim confusingly gave the error "space shortage".
3555 Furthermore, in debugging mode it crashed with a floating point exception.
3556 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3557 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3558 serious problem, Exim now writes to the main and panic logs when this
3559 happens, with details of the failure. It then refuses to accept the
3560 incoming message, giving the message "spool directory problem" or "log
3561 directory problem" with a 421 code for SMTP messages.
3563 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3564 0, 1, and 2 exist, because some OS complain for execs without them (see
3565 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3566 descriptors. However, the code omitted to check that the open succeeded,
3567 causing mysterious errors if for some reason the permissions on /dev/null
3568 got screwed. Now Exim writes a message to the main and panic logs, and
3569 bombs out if it can't open /dev/null.
3571 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3572 interact so that it is all more flexible. It is supposed to remain
3573 backwards compatible. Also added extra_local_interfaces.
3575 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3576 to bomb out with an assertion failure - to the client this appears as a
3577 connection drop. This problem occurs in the part of the code that was taken
3578 from the Samba project. Fortunately, the assertion is in a very simple
3579 function, so I have fixed this by reproducing the function inline in the
3580 one place where it is called, and arranging for authentication to fail
3581 instead of killing the process with assert().
3583 75. The SPA client code was not working when the server requested OEM rather
3584 than Unicode encoding.
3586 76. Added code to make require_files with a specific uid setting more usable in
3587 the case where statting the file as root fails - usually a non-root-mounted
3588 NFS file system. When this happens and the failure is EACCES, Exim now
3589 forks a subprocess and does the per-uid checking as the relevant uid.
3591 77. Added process_log_path.
3593 78. If log_file_path was not explicitly set, a setting of check_log_space or
3594 check_log_inodes was ignored.
3596 79. If a space check for the spool or log partitions fails, the incident is now
3597 logged. Of course, in the latter case the data may get lost...
3599 80. Added the %p formatting code to string_format() so that it can be used to
3600 print addresses in debug_print(). Adjusted all the address printing in the
3601 debugging in store.c to use %p rather than %d.
3603 81. There was a concern that a line of code in smtp_in.c could overflow a
3604 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3605 initially expressed, the concern was not well-founded, because trailing
3606 spaces are removed early. However, if the trailing spaces were followed by
3607 a NULL, they did not get removed, so the overflow was possible. Two fixes
3610 (a) I re-wrote the offending code in a cleaner fashion.
3611 (b) If an incoming SMTP command contains a NULL character, it is rejected
3614 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3615 runs initgroups(), so that if the Exim user is in any additional groups,
3616 they will be used during message reception.
3622 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob