1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.291 2006/02/10 16:29:20 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
20 PH/03 Added disable_ipv6.
22 PH/04 Changed $reply_address to use the raw form of the headers instead of the
23 decoded form, because it is most often used to construct To: headers
24 lines in autoreplies, and the decoded form may well be syntactically
25 invalid. However, $reply_address has leading white space removed, and all
26 newlines turned into spaces so that the autoreply transport does not
29 PH/05 If group was specified without a user on a router, and no group or user
30 was specified on a transport, the group from the router was ignored.
32 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
33 for visible compile-time settings that can be used to change these
34 numbers, for those that want even more. Backwards compatibility with old
35 spool files has been maintained. However, going back to a previous Exim
36 release will lost any variables that are in spool files.
38 PH/07 Two small changes when running in the test harness: increase delay when
39 passing a TCP/IP connection to a new process, in case the original
40 process has to generate a bounce, and remove special handling of
41 127.0.0.2 (sic), which is no longer necessary.
43 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
44 be the same on different OS.
46 PH/09 Moved a debug statement in filter processing to avoid a race problem when
49 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
50 whether --show-vars was specified or not
52 JJ/02 exipick: Added support for new ACL variable spool format introduced
55 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
56 syntactically invalid From: or Reply-to: line, and a filter used this to
57 generate an autoreply, and therefore failed to obtain an address for the
58 autoreply, Exim could try to deliver to a non-existent relative file
59 name, causing unrelated and misleading errors. What now happens is that
60 it logs this as a hard delivery error, but does not attempt to create a
63 PH/11 The exinext utility has a -C option for testing purposes, but although
64 the given file was scanned by exinext itself; it wasn't being passed on
67 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
68 an end-of-file indication when reading a command response.
70 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
71 compiled. In many other places in Exim, IPv6 addresses are always
72 recognized, so I have changed this. It also means that IPv4 domain
73 literals of the form [IPV4:n.n.n.n] are now always recognized.
75 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
76 used if the router is not running as root, for example, when verifying at
77 ACL time, or when using -bh. The debugging output from this situation was
78 non-existent - all you got was a failure to exec. I have made two
81 (a) Failures to set uid/gid, the current directory, or a process leader
82 in a subprocess such as that created by queryprogram now generate
83 suitable debugging ouput when -d is set.
85 (b) The queryprogram router detects when it is not running as root,
86 outputs suitable debugging information if -d is set, and then runs
87 the subprocess without attempting to change uid/gid.
89 PH/15 Minor change to Makefile for building test_host (undocumented testing
92 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
93 additional section of a DNS packet that returns MX or SRV records.
94 Instead, it always explicitly searches for A/AAAA records. This avoids
95 major problems that occur when a DNS server includes only records of one
96 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
97 fixed another bug: if SRV records were looked up and the corresponding
98 address records were *not* found in the additional section, the port
99 values from the SRV records were lost.
101 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
102 using the correct key (the original address) when searching the retry
103 rules in order to find which one to use for generating the retry hint.
105 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
106 adding the default one. Similarly, if it contains a Reply-To: header, the
107 errors_reply_to option, if set, is not used.
109 PH/19 When calculating a retry time, Exim used to measure the "time since
110 failure" by looking at the "first failed" field in the retry record. Now
111 it does not use this if it is later than than the arrival time of the
112 message. Instead it uses the arrival time. This makes for better
113 behaviour in cases where some deliveries succeed, thus re-setting the
114 "first failed" field. An example is a quota failure for a huge message
115 when small messages continue to be delivered. Without this change, the
116 "time since failure" will always be short, possible causing more frequent
117 delivery attempts for the huge message than are intended.
119 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
120 $1, $2, $3) because the numerical variables can be reset during some
121 expansion items (e.g. "match"), thereby losing the authentication data.
123 PH/21 Make -bV show the size of off_t variables so that the test suite can
124 decide whether to run tests for quotas > 2G.
126 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
127 mailbox_size, and mailbox_filecount in the appendfile transport. If a
128 filecount value is greater than 2G or if a quota value is greater than 2G
129 on a system where the size of off_t is not greater than 4, a panic error
137 PH/01 Two changes to the default runtime configuration:
139 (1) Move the checks for relay_from_hosts and authenticated clients from
140 after to before the (commented out) DNS black list checks.
142 (2) Add control=submission to the relay_from_hosts and authenticated
143 clients checks, on the grounds that messages accepted by these
144 statements are most likely to be submissions.
146 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
148 (1) Generate an error if the third argument for the ${prvs expansion is
151 (2) Treat a missing third argument of ${prvscheck as if it were an empty
154 (3) Reset the variables that are obtained from the first argument of
155 ${prvscheck and used in the second argument before leaving the code,
156 because their memory is reclaimed, so using them afterwards may do
159 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
160 one (it's much easier than Tom thought :-).
162 (5) Because of (4), we can now allow for the use of $prvscheck_result
163 inside the third argument.
165 PH/03 For some reason, the default setting of PATH when running a command from
166 a pipe transport was just "/usr/bin". I have changed it to
169 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
170 anything to be listed in the output from -bV.
172 PH/05 When a filter generated an autoreply, the entire To: header line was
173 quoted in the delivery log line, like this:
175 => >A.N.Other <ano@some.domain> <original@ddress> ...
177 This has been changed so that it extracts the operative address. There
178 may be more than one such address. If so, they are comma-separated, like
181 => >ano@some.domain,ona@other.domain <original@ddress> ...
183 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
184 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
185 not being looked up in the rDNS to get a host name, Exim was showing the
186 IP address twice in Received: lines, even though the IP addresses were
187 identical. For example:
189 Received: from [1.2.3.4] (helo=[1.2.3.4])
191 However, if the real host name was known, it was omitting the HELO data
192 if it matched the actual IP address. This has been tidied up so that it
193 doesn't show the same IP address twice.
195 PH/07 When both +timestamp and +memory debugging was on, the value given by
196 $tod_xxx expansions could be wrong, because the tod_stamp() function was
197 called by the debug printing, thereby overwriting the timestamp buffer.
198 Debugging no longer uses the tod_stamp() function when +timestamp is set.
200 PH/08 When the original message was included in an autoreply transport, it
201 always said "this is a copy of the message, including all the headers",
202 even if body_only or headers_only was set. It now gives an appropriate
205 PH/09 Applied a patch from the Sieve maintainer which:
207 o fixes some comments
208 o adds the (disabled) notify extension core
209 o adds some debug output for the result of if/elsif tests
210 o points to the current vacation draft in the documentation
211 and documents the missing references header update
215 o fixes a bug in processing the envelope test (when testing
216 multiple envelope elements, the last element determinted the
219 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
220 Electronic Mail") by including:
222 Auto-submitted: auto-generated
224 in the messages that it generates (bounce messages and others, such as
225 warnings). In the case of bounce messages for non-SMTP mesages, there was
226 also a typo: it was using "Auto_submitted" (underscore instead of
227 hyphen). Since every message generated by Exim is necessarily in response
228 to another message, thes have all been changed to:
230 Auto-Submitted: auto-replied
232 in accordance with these statements in the RFC:
234 The auto-replied keyword:
236 - SHOULD be used on messages sent in direct response to another
237 message by an automatic process,
239 - MUST NOT be used on manually-generated messages,
241 - MAY be used on Delivery Status Notifications (DSNs) and Message
242 Disposition Notifications (MDNs),
244 - MUST NOT be used on messages generated by automatic or periodic
245 processes, except for messages which are automatic responses to
248 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
249 to the default Received: header definition.
251 PH/12 Added log selector acl_warn_skipped (default on).
253 PH/13 After a successful wildlsearch lookup, discard the values of numeric
254 variables because (a) they are in the wrong storage pool and (b) even if
255 they were copied, it wouldn't work properly because of the caching.
257 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
258 checking when decoding. Apparently there are clients that generate
259 overlong encoded strings. Why am I not surprised?
261 PH/15 If the first argument of "${if match_address" was not empty, but did not
262 contain an "@" character, Exim crashed. Now it writes a panic log message
263 and treats the condition as false.
265 PH/16 In autoreply, treat an empty string for "once" the same as unset.
267 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
268 extension "envelope-auth". The code is finished and in agreement with
269 other implementations, but there is no documentation so far and in fact,
270 nobody wrote the draft yet. This extension is currently #undef'ed, thus
271 not changing the active code.
273 Print executed "if" and "elsif" statements when debugging is used. This
274 helps a great deal to understand what a filter does.
276 Document more things not specified clearly in RFC3028. I had all this
277 sorted out, when out of a sudden new issues came to my mind. Oops."
279 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
282 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
283 canonical form (as documented). However, after a host name lookup from
284 the IP address, check_host() was doing a simple string comparison with
285 addresses acquired from the DNS when checking that the found name did
286 have the original IP as one of its addresses. Since any found IPv6
287 addresses are likely to be in abbreviated form, the comparison could
288 fail. Luckily, there already exists a function for doing the comparison
289 by converting both addresses to binary, so now that is used instead of
292 PH/20 There was another similar case to PH/19, when a complete host name was
293 given in a host list; looking up its IP address could give an abbreviated
294 form, whereas the current host's name might or might not be abbreviated.
295 The same fix has been applied.
301 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
302 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
305 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
306 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
308 PH/03 Typo: missing ".o" in src/pcre/Makefile.
310 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
311 header line, restrict the check to what is listed in RFCs 2369 and 2929.
312 Also, for "Auto-Submitted", treat anything other than "no" as
313 non-personal, in accordance with RFC 3834. (Previously it treated
314 anything starting "auto-" as non-personal.)
316 TF/01 The control=submission/name=... option had a problem with syntax
317 errors if the name included a slash character. The /name= option
318 now slurps the rest of the string, so it can include any characters
319 but it must come last in the list of options (after /sender_retain
322 PH/05 Some modifications to the interface to the fake nameserver for the new
330 TK/01 Added the "success_on_redirect" address verification option. See
331 NewStuff for rationale and an example.
333 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
335 PH/02 Patch to exigrep to allow it to work on syslog lines.
337 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
338 fread() to skip over the body file's header line, because in Cygwin the
339 header line is locked and is inaccessible.
341 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
342 co-exist for some time) to make it clear that it is the Exim ID that is
343 referenced, not the Message-ID: header line.
345 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
346 string_format() function, because snprintf() does not exist on all
349 PH/06 The use of forbid_filter_existstest now also locks out the use of the
350 ${stat: expansion item.
352 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
353 protocol synchronization error", to keep the pedants happy.
355 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
356 well as for IRIX systems, when gcc is being used. See the host.c source
359 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
361 PH/10 Named domain lists were not working if used in a queue_smtp_domains
364 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
365 transport and to the smtp transport in LMTP mode.
367 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
369 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
370 run a filter in a subprocess. This could lead to confusion in subsequent
371 lookups in the parent process. There should also be a search_tidyup() at
372 the end of the subprocess.
374 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
375 only if the host matched helo_try_verify_hosts, which caused the
376 verification to occur when the EHLO/HELO command was issued. The ACL just
377 tested the remembered result. Now, if a previous verification attempt has
378 not happened, "verify = helo" does it there and then.
380 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
382 TK/03 Fix log output including CR from clamd.
384 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
385 exist provoked a memory error which could cause a segfault.
387 PH/15 Installed PCRE 6.2
389 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
391 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
392 of the problem. Specifically, suggested +O2 rather than +O1 for the
395 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
397 PH/20 If a delivery was routed to a non-standard port by means of an SRV
398 record, the port was not correctly logged when the outgoing_port log
399 selector was set (it logged the transort's default port).
401 PH/21 Added support for host-specific ports to manualroute, queryprogram,
402 fallback_hosts, and "hosts" in the smtp transport.
404 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
405 host errors such as "Connection refused".
407 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
408 authentication with radiusclient 0.4.9:
410 - Error returned from rc_read_config was caught wrongly
411 - Username/password not passed on to radius server due to wrong length.
413 The presumption is that some radiusclient API changes for 4.51/PH/17
414 were not taken care of correctly. The code is still untested by me (my
415 Linux distribution still has 0.3.2 of radiusclient), but it was
416 contributed by a Radius user.
418 PH/24 When doing a callout, the value of $domain wasn't set correctly when
419 expanding the "port" option of the smtp transport.
421 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
422 while reading a MIME header. Thanks to Tom Hughes for a patch.
424 PH/24 Include config.h inside local_scan.h so that configuration settings are
427 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
428 that in an ACL for RCPT (for example), you can examine exactly what was
431 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
432 commands, but it was not correctly comparing the address with the actual
433 client host address. Thus, it would show the EHLO address in Received:
434 header lines when this was not necessary.
436 PH/27 Added the % operator to ${eval:}.
438 PH/28 Exim tries to create and chdir to its spool directory when it starts;
439 it should be ignoring failures (because with -C, for example, it has lost
440 privilege). It wasn't ignoring creation failures other than "already
443 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
446 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
447 because an input line was too long, either on its own, or by virtue of
448 too many continuations, the temporary file was not being removed, and the
449 return code was incorrect.
451 PH/31 Missing "BOOL" in function definition in filtertest.c.
453 PH/32 Applied Sieve patches from the maintainer.
455 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
457 PH/33 Added "verify = not_blind".
459 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
460 Local/Makefile (with some defaults set). These are used in built scripts
461 such as exicyclog, but they have never been used in the exim_install
462 script (though there are many overriding facilities there). I have
463 arranged that the exim_install script now takes note of these two
466 PH/35 Installed configuration files for Dragonfly.
468 PH/36 When a locally submitted message by a trusted user did not contain a
469 From: header, and the sender address was obtained from -f or from an SMTP
470 MAIL command, and the trusted user did not use -F to supply a sender
471 name, $originator_name was incorrectly used when constructing a From:
472 header. Furthermore, $originator_name was used for submission mode
473 messages from external hosts without From: headers in a similar way,
474 which is clearly wrong.
476 PH/37 Added control=suppress_local_fixups.
478 PH/38 When log_selector = +received_sender was set, and the addition of the
479 sender made the log line's construction buffer exactly full, or one byte
480 less than full, an overflow happened when the terminating "\n" was
483 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
484 when the result of a list match is failure because a DNS lookup failed.
486 PH/40 RM_COMMAND is now used in the building process.
488 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
489 the "build-* directories that it finds.
491 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
492 address in a domain literal was a prefix of an interface address.
494 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
495 when verifying a sender address, unless rewrite_headers is false.
497 PH/44 Wrote a long comment about why errors_to addresses are verified as
498 recipients, not senders.
500 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
501 the ratelimit ACL was added.
503 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
505 PH/46 Added extra information about PostgreSQL errors to the error string.
507 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
508 avoiding the need to install special zones in a real server. This is
509 backwards compatible; if it can't find the fake resolver, it drops back.
510 Thus, both old and new test suites can be run.
512 TF/02 Added util/ratelimit.pl
514 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
515 clock is set back in time.
517 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
518 Candler <B.Candler@pobox.com>.
520 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
521 OK for addresses that are the result of redirections.
523 PH/48 A number of further additions for the benefit of the new test suite,
524 including a fake gethostbyname() that interfaces to the fake DNS resolver
527 TF/06 The fix for widen_domains has also been applied to qualify_single and
528 search_parents which are the other dnslookup options that can cause
531 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
534 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
536 TF/07 Exim produced the error message "an SRV record indicated no SMTP
537 service" if it encountered an MX record with an empty target hostname.
538 The message is now "an MX or SRV record indicated no SMTP service".
540 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
541 if the DNS of the sending site is misconfigured. This is quite a
542 common situation. This change restores the behaviour of treating a
543 helo verification defer as a failure.
545 PH/51 If self=fail was set on a router, the bounce message did not include the
546 actual error message.
552 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
554 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
555 command itself ended in error, the underlying message about the transport
556 filter timeout was being overwritten with the pipe command error. Now the
557 underlying error message should be appended to the second error message.
559 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
561 PH/02 Reduce the amount of output that "make" produces by default. Full output
562 can still be requested.
564 PH/03 The warning log line about a condition test deferring for a "warn" verb
565 was being output only once per connection, rather than after each
566 occurrence (because it was using the same function as for successful
567 "warn" verbs). This seems wrong, so I have changed it.
569 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
570 it should not have, which might have caused a crash in the right
571 circumstances, but probably never did.
573 PH/04 Installed a modified version of Tony Finch's patch to make submission
574 mode fix the return path as well as the Sender: header line, and to
575 add a /name= option so that you can make the user's friendly name appear
578 TF/03 Added the control = fakedefer ACL modifier.
580 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
581 Mark Lowes for thorough testing.
583 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
585 TK/03 Merged latest SRS patch from Miles Wilton.
587 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
588 with the definition in sysexits.h (which is #included earlier).
589 Fortunately, Exim does not actually use EX_OK. The code used to try to
590 preserve the sysexits.h value, by assumimg that macro definitions were
591 scanned for macro replacements. I have been disabused of this notion,
592 so now the code just undefines EX_OK before #including unistd.h.
594 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
595 in the smtp transport. When a block could not be written in a single
596 write() function, the timeout was being re-applied to each part-write.
597 This seems wrong - if the receiver was accepting one byte at a time it
598 would take for ever. The timeout is now adjusted when this happens. It
599 doesn't have to be particularly precise.
601 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
602 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
604 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
605 without a domain if the check to <postmaster@domain> fails.
607 SC/01 Eximstats: added -xls and the ability to specify output files
608 (patch written by Frank Heydlauf).
610 SC/02 Eximstats: use FileHandles for outputing results.
612 SC/03 Eximstats: allow any combination of xls, txt, and html output.
614 SC/04 Eximstats: fixed display of large numbers with -nvr option
616 SC/05 Eximstats: fixed merging of reports with empty tables.
618 SC/06 Eximstats: added the -include_original_destination flag
620 SC/07 Eximstats: removed tabs and trailing whitespace.
622 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
624 TK/06 MBOX spool code: Add real "From " MBOX separator line
625 so the .eml file is really in mbox format (even though
626 most programs do not really care). Patch from Alex Miller.
628 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
629 The latter is generated from $received_to and is only set if the
630 message has one envelope recipient. SA can use these headers,
631 obviously out-of-the-box. Patch from Alex Miller.
633 PH/08 The ${def test on a variable was returning false if the variable's
634 value was "0", contrary to what the specification has always said!
635 The result should be true unless the variable is empty.
637 PH/09 The syntax error of a character other than { following "${if
638 def:variable_name" (after optional whitespace) was not being diagnosed.
639 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
640 accidental colon was present, for example, could give incorrect results.
642 PH/10 Tidied the code in a number of places where the st_size field of a stat()
643 result is used (not including appendfile, where other changes are about
646 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
647 This involved changing a lot of size variables from int to off_t. It
648 should work with maildirs and everything.
650 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
651 spamd dying while we are connected to it.
653 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
656 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
657 function for BDB 4.3.
659 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
661 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
662 the sources for PCRE have been re-arranged and more files are now
665 PH/15 The code I had for printing potentially long long variables in PH/11
666 above was not the best (it lost precision). The length of off_t variables
667 is now inspected at build time, and an appropriate printing format (%ld
668 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
669 to be "long long int" or "long int". This is needed for the internal
670 formatting function string_vformat().
672 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
673 the configuration file to be ":syslog", then the script "guesses" where
674 the logs files are, rather than using the compiled in default. In our
675 case the guess is not the same as the compiled default, so the script
676 suddenly stopped working when I started to use syslog. The patch checks
677 to see if log_file_path is "". If so, it attempts to read it from exim
678 with no configuration file to get the compiled in version, before it
679 falls back to the previous guessing code."
681 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
682 implementing BATV in an Exim configuration. See NewStuff for the gory
685 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
686 Makefile that are specific to HP-UX.
688 PH/18 If the "use_postmaster" option was set for a recipient callout together
689 with the "random" option, the postmaster address was used as the MAIL
690 FROM address for the random test, but not for the subsequent recipient
691 test. It is now used for both.
693 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
694 patch removes a few documentation additions to RFC 3028, because the
695 latest draft now contains them. It adds the new en;ascii-case comparator
696 and a new error check for 8bit text in MIME parts. Comparator and
697 require names are now matched exactly. I enabled the subaddress
698 extension, but it is not well tested yet (read: it works for me)."
700 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
701 rework some of the code of TK/09 above to avoid the hardwired use of
702 "%lld" and "long long". Replaced the call to snprintf() with a call to
705 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
706 records point to non-existent hosts", "retry timeout exceeded", and
707 "retry time not reached for any host after a long failure period".
709 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
710 experimental DomainKeys support:
712 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
713 (2) On an error such as an illegally used "control", the wrong name for
714 the control was given.
716 These problems did NOT occur unless DomainKeys support was compiled.
718 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
720 PH/24 Added ${if match_ip condition.
722 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
723 databases so that it will be absolutely obvious if a crash occurs in the
724 DB library. This is a regular occurrence (often caused by mis-matched
727 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
728 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
729 picked up on a user's system that detects such things. There doesn't seem
730 to be a gcc warning option for this - only an attribute that has to be
731 put on the function's prototype. It seems that in Fedora Core 4 they have
732 set this on a number of new functions. No doubt there will be more in due
735 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
736 specify a transport. However, if an address that was verified by such a
737 router was the subject of a callout, Exim crashed because it tried to
738 read the rcpt_include_affixes from the non-existent transport. Now it
739 just assumes that the setting of that option is false. This bug was
740 introduced by 4.51/PH/31.
742 PH/28 Changed -d+all to exclude +memory, because that information is very
743 rarely of interest, but it makes the output a lot bigger. People tend to
744 do -d+all out of habit.
746 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
747 code in os-type was giving problems when libc.so lives in lib64, like on
750 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
751 aren't the modern standard, and it seems that some systems' include files
752 don't always have them. Exim was already checking for some of the newer
753 ones like T_AAAA, and defining it itself. I've added checks for all the
754 record types that Exim uses.
756 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
757 not automatically generating a new one, as it is supposed to. This
758 prevented TLS from working. If the file did exist, but contained invalid
759 data, a new version was generated, as expected. It was only the case of a
760 non-existent file that was broken.
762 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
763 with a change in libdomainkeys > 0.64.
765 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
766 from DNS. If the selector record carries the flag, it now has
767 precedence over the domain-wide flag.
769 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
771 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
772 the use of a port name that isn't defined in /etc/services) occurred, the
773 message was deferred as in a normal delivery, and thus remained on the
774 spool, instead of being failed because of the mua_wrapper setting. This
775 is now fixed, and I tidied up some of the mua_wrapper messages at the
778 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
779 the messages in a hash of arrays rather than using individual hashes.
780 This is a bit cleaner and results in dramatic memory savings, albeit
781 at a slight CPU cost.
783 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
784 as requested by Marc Sherman.
786 SC/10 Eximstats: added histograms for user specified patterns as requested
789 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
791 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
792 fopen() in the content-scanning modules that did not already have it.
798 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
799 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
801 TK/02 Fix ACL "control" statement not being available in MIME ACL.
803 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
805 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
806 to test Sieve filters that use "vacation".
808 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
809 that changes the way the GnuTLS parameters are stored in the cache file.
810 The new format can be generated externally. For backward compatibility,
811 if the data in the cache doesn't make sense, Exim assumes it has read an
812 old-format file, and it generates new data and writes a new file. This
813 means that you can't go back to an older release without removing the
816 PH/03 A redirect router that has both "unseen" and "one_time" set does not
817 work if there are any delivery delays because "one_time" forces the
818 parent to be marked "delivered", so its unseen clone is never tried
819 again. For this reason, Exim now forbids the simultaneous setting of
822 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
823 redirected to themselves ("homonym" addresses). Read the long ChangeLog
824 entry if you want to know the details. The fix, however, neglected to
825 consider the case when local delivery batching is involved. The test for
826 "previously delivered" was not happening when checking to see if an
827 address could be batched with a previous (undelivered) one; under
828 certain circumstances this could lead to multiple deliveries to the same
831 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
832 in its include files, and this causes problems building Exim.
834 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
835 header_syntax) but Exim was just ignoring anything given after a slash.
836 In particular, this caused confusion with an attempt to use "verify =
837 reverse_host_lookup/defer_ok". An error is now given when options are
838 supplied for verify items that do not have them. (Maybe reverse_host_
839 lookup should have a defer_ok option, but that's a different point.)
841 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
842 defined by RFC 821) to 2048, because there were problems with some AUTH
843 commands, and RFC 1869 says the size should be increased for extended
844 SMTP commands that take arguments.
846 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
849 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
850 "unknown" error; now it says that the functionality isn't in the binary.
852 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
853 an address' error message when a string expansion fails (syntax or
854 whatever). Otherwise the password may appear in the log. Following change
855 PH/42 below, there is no longer a chance of it appearing in a bounce
858 PH/11 Installed exipick version 20050225.0 from John Jetmore.
860 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
861 of its addresses was ever tried. (Bugzilla bug #2.)
863 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
864 the result incorrectly in the debug output. (It correctly added a newline
865 to what was transported.)
867 TF/01 Added $received_time.
869 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
870 commented out examples of how to interface to a virus scanner and to
871 SpamAssassin. Also added commented examples of av_scanner and
872 spamd_address settings.
874 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
875 and controls are allowed in which ACLs. There were a couple of minor
876 errors. Some of the entries in the conditions table (which is a table of
877 where they are NOT allowed) were getting very unwieldy; rewrote them as a
878 negation of where the condition IS allowed.
880 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
882 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
883 header file does not have a version number, so I've had to invent a new
884 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
885 API. The code is untested by me (my Linux distribution still has 0.3.2 of
886 radiusclient), but it was contributed by a Radius user.
888 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
889 files or directories, for OpenSSL.
891 PH/19 When an Exim process that is running as root has to create an Exim log
892 file, it does so in a subprocess that runs as exim:exim so as to get the
893 ownership right at creation (otherwise, other Exim processes might see
894 the file with the wrong ownership). There was no test for failure of this
895 fork() call, which would lead to the process getting stuck as it waited
896 for a non-existent subprocess. Forks do occasionally fail when resources
897 run out. I reviewed all the other calls to fork(); they all seem to check
900 PH/20 When checking for unexpected SMTP input at connect time (before writing
901 the banner), Exim was not dealing correctly with a non-positive return
902 from the read() function. If the client had disconnected by this time,
903 the result was a log entry for a synchronization error with an empty
904 string after "input=" when read() returned zero. If read() returned -1
905 (an event I could not check), uninitialized data bytes were printed.
906 There were reports of junk text (parts of files, etc) appearing after
909 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
911 PH/22 Added support for macro redefinition, and (re)definition in between
912 driver and ACL definitions.
914 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
915 forgetting to use the resulting value; it was using the unexpanded value.
917 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
918 hadn't been configured. The fix is from Juergen Kreileder, who
919 understands it better than I do:
921 "Here's what I see happening with three configured cyrus_sasl
922 authenticators configured (plain, login, cram-md5):
924 On startup auth_cyrus_sasl_init() gets called for each of these.
925 This means three calls to sasl_listmech() without a specified mech_list.
926 => SASL tests which mechs of all available mechs actually work
927 => three warnings about OTP not working
928 => the returned list contains: plain, login, cram-md5, digest-md5, ...
930 With the patch, sasl_listmech() also gets called three times. But now
931 SASL's mech_list option is set to the server_mech specified in the the
932 authenticator. Or in other words, the answer from sasl_listmech()
933 gets limited to just the mech you're testing for (which is different
935 => the return list contains just 'plain' or 'login', 'cram-md5' or
936 nothing depending on the value of ob->server_mech.
938 I've just tested the patch: Authentication still works fine,
939 unavailable mechs specified in the exim configuration are still
940 caught, and the auth.log warnings about OTP are gone."
942 PH/25 When debugging is enabled, the contents of the command line are added
943 to the debugging output, even when log_selector=+arguments is not
946 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
947 answer is "GNU", and only if the return is "GNU/something" is the answer
950 PH/27 $acl_verify_message is now set immediately after the failure of a
951 verification in an ACL, and so is available in subsequent modifiers. In
952 particular, the message can be preserved by coding like this:
954 warn !verify = sender
955 set acl_m0 = $acl_verify_message
957 Previously, $acl_verify_message was set only while expanding "message"
958 and "log_message" when a very denied access.
960 PH/28 Modified OS/os.c-Linux with
962 -#ifndef OS_LOAD_AVERAGE
963 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
965 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
966 nomenclature these days.)
968 PH/29 Installed patch from the Sieve maintainer that adds the options
969 sieve_useraddress and sieve_subaddress to the redirect router.
971 PH/30 In these circumstances:
972 . Two addresses routed to the same list of hosts;
973 . First host does not offer TLS;
974 . First host accepts first address;
975 . First host gives temporary error to second address;
976 . Second host offers TLS and a TLS session is established;
977 . Second host accepts second address.
978 Exim incorrectly logged both deliveries with the TLS parameters (cipher
979 and peerdn, if requested) that were in fact used only for the second
982 PH/31 When doing a callout as part of verifying an address, Exim was not paying
983 attention to any local part prefix or suffix that was matched by the
984 router that accepted the address. It now behaves in the same way as it
985 does for delivery: the affixes are removed from the local part unless
986 rcpt_include_affixes is set on the transport.
988 PH/32 Add the sender address, as F=<...>, to the log line when logging a
989 timeout during the DATA phase of an incoming message.
991 PH/33 Sieve envelope tests were broken for match types other than :is. I have
992 applied a patch sanctioned by the Sieve maintainer.
994 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
995 the uid or gid is negative. A case of a negative gid caused this to be
996 noticed. The fix allows for either to be negative.
998 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
999 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1002 PH/36 The change PH/12 above was broken. Fixed it.
1004 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1005 the grounds that routing the same address twice would always produce the
1006 same answer. This might have been true once, but it is certainly no
1007 longer true now. Routing a child address may depend on the previous
1008 routing that produced that child. Some complicated redirection strategies
1009 went wrong when messages had multiple recipients, and made Exim's
1010 behaviour dependent on the order in which the addresses were given.
1012 I have moved the duplicate checking until after the routing is complete.
1013 Exim scans the addresses that are assigned to local and remote
1014 transports, and removes any duplicates. This means that more work will be
1015 done, as duplicates will always all be routed, but duplicates are
1016 presumably rare, so I don't expect this is of any significance.
1018 For deliveries to pipes, files, and autoreplies, the duplicate checking
1019 still happens during the routing process, since they are not going to be
1022 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1023 It corrects a timeout issue with spamd. This is Ian's comment: "The
1024 background is that sometimes spamd either never reads data from a
1025 connection it has accepted, or it never writes response data. The exiscan
1026 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1027 blindly assumes that writes won't block so it may never time out."
1029 PH/39 Allow G after quota size as well as K and M.
1031 PH/40 The value set for $authenticated_id in an authenticator may not contain
1032 binary zeroes or newlines because the value is written to log lines and
1033 to spool files. There was no check on this. Now the value is run through
1034 the string_printing() function so that such characters are converted to
1035 printable escape sequences.
1037 PH/41 $message_linecount is a new variable that contains the total number of
1038 lines in the message. Compare $body_linecount, which is the count for the
1041 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1042 bounce and delay warning messages, except in certain special cases, which
1045 (a) An SMTP error message from a remote host;
1046 (b) A message specified in a :fail: redirection;
1047 (c) A message specified in a "fail" command in a system filter;
1048 (d) A message specified in a FAIL return from the queryprogram router;
1049 (e) A message specified by the cannot_route_message router option.
1051 In these cases only, Exim does include the error details in bounce and
1052 warning messages. There are also a few cases where bland messages such
1053 as "unrouteable address" or "local delivery error" are given.
1055 PH/43 $value is now also set for the "else" part of a ${run expansion.
1057 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1058 being worked on, but at least Exim now implements the latest version to
1061 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1062 process to complete was treated as a delivery failure, a timeout while
1063 writing the message to the pipe was logged, but erroneously treated as a
1064 successful delivery. Such timeouts include transport filter timeouts. For
1065 consistency with the overall process timeout, these timeouts are now
1066 treated as errors, giving rise to delivery failures by default. However,
1067 there is now a new Boolean option for the pipe transport called
1068 timeout_defer, which, if set TRUE, converts the failures into defers for
1069 both kinds of timeout. A transport filter timeout is now identified in
1072 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1073 systems where "make" and "gmake" are different, calling "gmake" at top
1074 level broke things. I've arranged for the value of $(MAKE) to be passed
1075 from the Makefile to this script so that it can call the same version of
1079 A note about Exim versions 4.44 and 4.50
1080 ----------------------------------------
1082 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1083 changes of various kinds. As a consequence, a big documentation update was
1084 needed. This delayed the release for rather longer than seemed good, especially
1085 in the light of a couple of (minor) security issues. Therefore, the changes
1086 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1087 release. So 4.44 and 4.50 are in effect two different branches that both start
1090 I have left the 4.50 change log unchanged; it contains all the changes since
1091 4.43. The change log for 4.44 is below; many of its items are identical to
1092 those for 4.50. This seems to be the most sensible way to preserve the
1093 historical information.
1099 1. Minor wording change to the doc/README.SIEVE file.
1101 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1102 computation of the current number of files was incorrect.
1104 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1105 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1108 4. Give more explanation in the error message when the command for a transport
1109 filter fails to execute.
1111 5. There are several places where Exim runs a non-Exim command in a
1112 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1113 was being done only for the command run by the queryprogram router. It is
1114 now done for all such subprocesses. The other cases are: ${run, transport
1115 filters, and the commands run by the lmtp and pipe transports.
1117 6. Added CONFIGURE_GROUP build-time option.
1119 7. Some older OS have a limit of 256 on the maximum number of file
1120 descriptors. Exim was using setrlimit() to set 1000 as a large value
1121 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1122 systems. I've change it so that if it can't get 1000, it tries for 256.
1124 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1125 was an oversight, and furthermore, ever since the addition of extra
1126 controls (e.g. 4.43/32), the checks on when to allow different forms of
1127 "control" were broken. There should now be diagnostics for all cases when a
1128 control that does not make sense is encountered.
1130 9. Added the /retain_sender option to "control=submission".
1132 10. $recipients is now available in the predata ACL (oversight).
1134 11. Tidy the search cache before the fork to do a delivery from a message
1135 received from the command line. Otherwise the child will trigger a lookup
1136 failure and thereby defer the delivery if it tries to use (for example) a
1137 cached ldap connection that the parent has called unbind on.
1139 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1140 of $address_data from the recipient verification was clobbered by the
1141 sender verification.
1143 13. The value of address_data from a sender verification is now available in
1144 $sender_address_data in subsequent conditions in the ACL statement.
1146 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1148 15. Added a new option "connect=<time>" to callout options, to set a different
1151 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1152 was its contents. (It was OK if the option was not defined at all.)
1154 17. A "Completed" log line is now written for messages that are removed from
1155 the spool by the -Mrm option.
1157 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1158 information about exactly what failed.
1160 19. Added -dd to debug only the daemon process.
1162 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1163 handles timeouts, both on the server side and network timeouts. Renamed the
1164 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1166 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1167 It is now set to "smtps".
1169 22. $host_address is now set to the target address during the checking of
1170 ignore_target_hosts.
1172 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1173 being passed; this would have caused $sender_host_name to have been used if
1174 matching the list had actually called for a host name (not very likely,
1175 since this list is usually IP addresses). A host name is now passed as
1178 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1179 code that specifies a non-restarting handler (typically sigaction() in
1180 modern systems) in an attempt to fix a rare and obscure crash bug.
1182 25. Narrowed the window for a race in the daemon that could cause it to ignore
1183 SIGCHLD signals. This is not a major problem, because they are used only to
1184 wake it up if nothing else does.
1186 26. A malformed maildirsize file could cause Exim to calculate negative values
1187 for the mailbox size or file count. Odd effects could occur as a result.
1188 The maildirsize information is now recalculated if the size or filecount
1191 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1192 support for a long time. Removed HAVE_SYS_VFS_H.
1194 28. Installed the latest version of exipick from John Jetmore.
1196 29. In an address list, if the pattern was not a regular expression, an empty
1197 subject address (from a bounce message) matched only if the pattern was an
1198 empty string. Non-empty patterns were not even tested. This was the wrong
1199 because it is perfectly reasonable to use an empty address as part of a
1200 database query. An empty address is now tested by patterns that are
1201 lookups. However, all the other forms of pattern expect the subject to
1202 contain a local part and a domain, and therefore, for them, an empty
1203 address still always fails if the pattern is not itself empty.
1205 30. Exim went into a mad DNS loop when attempting to do a callout where the
1206 host was specified on an smtp transport, and looking it up yielded more
1207 than one IP address.
1209 31. Re-factored the code for checking spool and log partition space into a
1210 function that finds that data and another that does the check. The former
1211 is then used to implement four new variables: $spool_space, $log_space,
1212 $spool_inodes, and $log_inodes.
1214 32. The RFC2047 encoding function was originally intended for short strings
1215 such as real names; it was not keeping to the 75-character limit for
1216 encoded words that the RFC imposes. It now respects the limit, and
1217 generates multiple encoded words if necessary. To be on the safe side, I
1218 have increased the buffer size for the ${rfc2047: expansion operator from
1221 33. It is now permitted to omit both strings after an "if" condition; if the
1222 condition is true, the result is "true". As before, when the second string
1223 is omitted, a false condition yields an empty string. This makes it less
1224 cumbersome to write custom ACL and router conditions.
1226 34. Failure to deliver a bounce message always caused it to be frozen, even if
1227 there was an errors_to setting on the router. The errors_to setting is now
1230 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1231 canonical form (fully expanded) before being placed in
1232 $sender_host_address.
1234 36. The table in the code that translates DNS record types into text (T_A to
1235 "A" for instance) was missing entries for NS and CNAME. It is just possible
1236 that this could have caused confusion if both these types were looked up
1237 for the same domain, because the text type is used as part of Exim's
1238 per-process caching. But the chance of anyone hitting this buglet seems
1241 37. The dnsdb lookup has been extended in a number of ways.
1243 (1) There is a new type, "zns", which walks up the domain tree until it
1244 finds some nameserver records. It should be used with care.
1246 (2) There is a new type, "mxh", which is like "mx" except that it returns
1247 just the host names, not the priorities.
1249 (3) It is now possible to give a list of domains (or IP addresses) to be
1250 looked up. The behaviour when one of the lookups defers can be
1251 controlled by a keyword.
1253 (4) It is now possible to specify the separator character for use when
1254 multiple records are returned.
1256 38. The dnslists ACL condition has been extended: it is now possible to supply
1257 a list of IP addresses and/or domains to be looked up in a particular DNS
1260 39. Added log_selector=+queue_time_overall.
1262 40. When running the queue in the test harness, wait just a tad after forking a
1263 delivery process, to get repeatability of debugging output.
1265 41. Include certificate and key file names in error message when GnuTLS fails
1266 to set them up, because the GnuTLS error message doesn't include the name
1267 of the failing file when there is a problem reading it.
1269 42. Allow both -bf and -bF in the same test run.
1271 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1273 44. The "Exiscan patch" is now merged into the mainline Exim source.
1275 45. Sometimes the final signoff response after QUIT could fail to get
1276 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1277 < 0 before doing a fflush(). This bug looks as though it goes back to the
1278 introduction of TLS in release 3.20, but "sometimes" must have been rare
1279 because the tests only now provoked it.
1281 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1282 (this can affect the format of dates).
1284 47. exim_tidydb, when checking for the continued existence of a message for
1285 which it has found a message-specific retry record, was not finding
1286 messages that were in split spool directories. Consequently, it was
1287 deleting retry records that should have stayed in existence.
1289 48. Steve fixed some bugs in eximstats.
1291 49. The SPA authentication driver was not abandoning authentication and moving
1292 on to the next authenticator when an expansion was forced to fail,
1293 contradicting the general specification for all authenticators. Instead it
1294 was generating a temporary error. It now behaves as specified.
1296 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1297 (the order specifies the preference for clients). The order is now AES256,
1298 AES128, 3DES, ARCFOUR128.
1300 51. Small patch to Sieve code - explicitly set From: when generating an
1303 52. Exim crashed if a remote delivery caused a very long error message to be
1304 recorded - for instance if somebody sent an entire SpamAssassin report back
1305 as a large number of 550 error lines. This bug was coincidentally fixed by
1306 increasing the size of one of Exim's internal buffers (big_buffer) that
1307 happened as part of the Exiscan merge. However, to be on the safe side, I
1308 have made the code more robust (and fixed the comments that describe what
1311 53. Now that there can be additional text after "Completed" in log lines (if
1312 the queue_time_overall log selector is set), a one-byte patch to exigrep
1313 was needed to allow it to recognize "Completed" as not the last thing in
1316 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1317 patch that reportedly fixes this has been added. I am not expert enough to
1318 create a test for it. This is what the patch creator wrote:
1320 "I found a little strange behaviour of ldap code when working with
1321 Windows 2003 AD Domain, where users was placed in more than one
1322 Organization Units. When I tried to give exim partial DN, the exit code
1323 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1324 But simultaneously result of request was absolutely normal ldap result,
1325 so I produce this patch..."
1327 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1328 REFERENCE, so I have modified the code to exclude the patch when that macro
1331 55. Some experimental protocols are using DNS PTR records for new purposes. The
1332 keys for these records are domain names, not reversed IP addresses. The
1333 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1334 leaves it alone. Component reversal etc. now happens only for IP addresses.
1337 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1339 57. Double the size of the debug message buffer (to 2048) so that more of very
1340 long debug lines gets shown.
1342 58. The exicyclog utility now does better if the number of log files to keep
1343 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1345 59. Two changes related to the smtp_active_hostname option:
1347 (1) $smtp_active_hostname is now available as a variable.
1348 (2) The default for smtp_banner uses $smtp_active_hostname instead
1349 of $primary_hostname.
1351 60. The host_aton() function is supposed to be passed a string that is known
1352 to be a valid IP address. However, in the case of IPv6 addresses, it was
1353 not checking this. This is a hostage to fortune. Exim now panics and dies
1354 if the condition is not met. A case was found where this could be provoked
1355 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1356 components; fortuitously, this particular loophole had already been fixed
1357 by change 4.50/55 above.
1359 If there are any other similar loopholes, the new check in host_aton()
1360 itself should stop them being exploited. The report I received stated that
1361 data on the command line could provoke the exploit when Exim was running as
1362 exim, but did not say which command line option was involved. All I could
1363 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1364 running as the user.
1367 61. There was a buffer overflow vulnerability in the SPA authentication code
1368 (which came originally from the Samba project). I have added a test to the
1369 spa_base64_to_bits() function which I hope fixes it.
1372 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1373 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1375 63. The daemon start-up calls getloadavg() while still root for those OS that
1376 need the first call to be done as root, but it missed one case: when
1377 deliver_queue_load_max is set with deliver_drop_privilege. This is
1378 necessary for the benefit of the queue runner, because there is no re-exec
1379 when deliver_drop_privilege is set.
1381 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1382 This has been fixed.
1384 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1385 was in use, was not putting the data itself into the right store pool;
1386 consequently, it could be overwritten for a subsequent message in the same
1387 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1390 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1392 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1393 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1394 the function were treating the return as a boolean value, which happened to
1395 work because 0=false and not-0=true, but is not correct code.
1397 68. The host_aton() function was not handling scoped IPv6 addresses (those
1398 with, for example, "%eth0" on the end) correctly.
1400 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1401 negated items (that is, ~something) in unsigned ints. Some compilers
1402 apparently mutter when there is no cast.
1404 70. If an address verification called from an ACL failed, and did not produce a
1405 user-specific message (i.e. there was only a "system" message), nothing was
1406 put in $acl_verify_message. In this situation, it now puts the system
1409 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1410 session; change 4.31/43 added the unwanted input to the log line - except
1411 that it did not do this in the start of session case. It now does.
1413 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1414 This is wrong and can cause the other end to generate a synchronization
1415 error if it is another Exim or anything else that does the synchronization
1416 check. A QUIT command is no longer sent after a timeout.
1418 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1419 during host lookups.
1421 74. The defer_ok option of callout verification was not working if it was used
1422 when verifying addresses in header lines, that is, for this case:
1424 verify = header_sender/callout=defer_ok
1426 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1427 those file descriptors could be used for SMTP connections. If anything
1428 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1429 could be sent to the SMTP client, causing chaos. The daemon now opens
1430 stdin, stdout, and stderr to /dev/null when it puts itself into the
1433 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1434 log by default. The user can override this with suitable Perl magic.
1436 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1437 the log message when discard triggers, was not working for the DATA ACL or
1438 for the non-SMTP ACL.
1440 78. Error message wording change in sieve.c.
1442 79. If smtp_accept_max_per_host was set, the number of connections could be
1443 restricted to fewer than expected, because the daemon was trying to set up
1444 a new connection before checking whether the processes handling previous
1445 connections had finished. The check for completed processes is now done
1446 earlier. On busy systems, this bug wouldn't be noticed because something
1447 else would have woken the daemon, and it would have reaped the completed
1450 80. If a message was submitted locally by a user whose login name contained one
1451 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1452 It caused a spool format error. I have fixed the spool reading code. A
1453 related problem was that the "from" clause in the Received: line became
1454 illegal because of the space(s). It is now covered by ${quote_local_part.
1456 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1459 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1461 83. Patch from Sieve maintainer to fix unterminated string problem in
1462 "vacation" handling.
1464 84. Some minor changes to the Linux configuration files to help with other
1465 OS variants using glibc.
1467 85. One more patch for Sieve to update vacation handling to latest spec.
1470 ----------------------------------------------------
1471 See the note above about the 4.44 and 4.50 releases.
1472 ----------------------------------------------------
1478 1. Change 4.43/35 introduced a bug that caused file counts to be
1479 incorrectly computed when quota_filecount was set in an appendfile
1482 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1483 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1486 3. Give more explanation in the error message when the command for a transport
1487 filter fails to execute.
1489 4. There are several places where Exim runs a non-Exim command in a
1490 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1491 was being done only for the command run by the queryprogram router. It is
1492 now done for all such subprocesses. The other cases are: ${run, transport
1493 filters, and the commands run by the lmtp and pipe transports.
1495 5. Some older OS have a limit of 256 on the maximum number of file
1496 descriptors. Exim was using setrlimit() to set 1000 as a large value
1497 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1498 systems. I've change it so that if it can't get 1000, it tries for 256.
1500 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1501 was an oversight, and furthermore, ever since the addition of extra
1502 controls (e.g. 4.43/32), the checks on when to allow different forms of
1503 "control" were broken. There should now be diagnostics for all cases when a
1504 control that does not make sense is encountered.
1506 7. $recipients is now available in the predata ACL (oversight).
1508 8. Tidy the search cache before the fork to do a delivery from a message
1509 received from the command line. Otherwise the child will trigger a lookup
1510 failure and thereby defer the delivery if it tries to use (for example) a
1511 cached ldap connection that the parent has called unbind on.
1513 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1514 of $address_data from the recipient verification was clobbered by the
1515 sender verification.
1517 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1518 was its contents. (It was OK if the option was not defined at all.)
1520 11. A "Completed" log line is now written for messages that are removed from
1521 the spool by the -Mrm option.
1523 12. $host_address is now set to the target address during the checking of
1524 ignore_target_hosts.
1526 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1527 being passed; this would have caused $sender_host_name to have been used if
1528 matching the list had actually called for a host name (not very likely,
1529 since this list is usually IP addresses). A host name is now passed as
1532 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1533 code that specifies a non-restarting handler (typically sigaction() in
1534 modern systems) in an attempt to fix a rare and obscure crash bug.
1536 15. Narrowed the window for a race in the daemon that could cause it to ignore
1537 SIGCHLD signals. This is not a major problem, because they are used only to
1538 wake it up if nothing else does.
1540 16. A malformed maildirsize file could cause Exim to calculate negative values
1541 for the mailbox size or file count. Odd effects could occur as a result.
1542 The maildirsize information is now recalculated if the size or filecount
1545 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1546 support for a long time. Removed HAVE_SYS_VFS_H.
1548 18. Updated exipick to current release from John Jetmore.
1550 19. Allow an empty sender to be matched against a lookup in an address list.
1551 Previously the only cases considered were a regular expression, or an
1554 20. Exim went into a mad DNS lookup loop when doing a callout where the
1555 host was specified on the transport, if the DNS lookup yielded more than
1558 21. The RFC2047 encoding function was originally intended for short strings
1559 such as real names; it was not keeping to the 75-character limit for
1560 encoded words that the RFC imposes. It now respects the limit, and
1561 generates multiple encoded words if necessary. To be on the safe side, I
1562 have increased the buffer size for the ${rfc2047: expansion operator from
1565 22. Failure to deliver a bounce message always caused it to be frozen, even if
1566 there was an errors_to setting on the router. The errors_to setting is now
1569 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1570 canonical form (fully expanded) before being placed in
1571 $sender_host_address.
1573 24. Updated eximstats to version 1.33
1575 25. Include certificate and key file names in error message when GnuTLS fails
1576 to set them up, because the GnuTLS error message doesn't include the name
1577 of the failing file when there is a problem reading it.
1579 26. Expand error message when OpenSSL has problems setting up cert/key files.
1582 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1583 (this can affect the format of dates).
1585 28. exim_tidydb, when checking for the continued existence of a message for
1586 which it has found a message-specific retry record, was not finding
1587 messages that were in split spool directories. Consequently, it was
1588 deleting retry records that should have stayed in existence.
1590 29. eximstats updated to version 1.35
1591 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1592 1.35 - bugfix such that pie charts by volume are generated correctly
1594 30. The SPA authentication driver was not abandoning authentication and moving
1595 on to the next authenticator when an expansion was forced to fail,
1596 contradicting the general specification for all authenticators. Instead it
1597 was generating a temporary error. It now behaves as specified.
1599 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1600 (the order specifies the preference for clients). The order is now AES256,
1601 AES128, 3DES, ARCFOUR128.
1603 31. Small patch to Sieve code - explicitly set From: when generating an
1606 32. Exim crashed if a remote delivery caused a very long error message to be
1607 recorded - for instance if somebody sent an entire SpamAssassin report back
1608 as a large number of 550 error lines. This bug was coincidentally fixed by
1609 increasing the size of one of Exim's internal buffers (big_buffer) that
1610 happened as part of the Exiscan merge. However, to be on the safe side, I
1611 have made the code more robust (and fixed the comments that describe what
1614 33. Some experimental protocols are using DNS PTR records for new purposes. The
1615 keys for these records are domain names, not reversed IP addresses. The
1616 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1617 leaves it alone. Component reversal etc. now happens only for IP addresses.
1620 34. The host_aton() function is supposed to be passed a string that is known
1621 to be a valid IP address. However, in the case of IPv6 addresses, it was
1622 not checking this. This is a hostage to fortune. Exim now panics and dies
1623 if the condition is not met. A case was found where this could be provoked
1624 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1625 components; fortuitously, this particular loophole had already been fixed
1626 by change 4.50/55 or 4.44/33 above.
1628 If there are any other similar loopholes, the new check in host_aton()
1629 itself should stop them being exploited. The report I received stated that
1630 data on the command line could provoke the exploit when Exim was running as
1631 exim, but did not say which command line option was involved. All I could
1632 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1633 running as the user.
1636 35. There was a buffer overflow vulnerability in the SPA authentication code
1637 (which came originally from the Samba project). I have added a test to the
1638 spa_base64_to_bits() function which I hope fixes it.
1641 36. The daemon start-up calls getloadavg() while still root for those OS that
1642 need the first call to be done as root, but it missed one case: when
1643 deliver_queue_load_max is set with deliver_drop_privilege. This is
1644 necessary for the benefit of the queue runner, because there is no re-exec
1645 when deliver_drop_privilege is set.
1647 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1648 was in use, was not putting the data itself into the right store pool;
1649 consequently, it could be overwritten for a subsequent message in the same
1650 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1653 38. Sometimes the final signoff response after QUIT could fail to get
1654 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1655 < 0 before doing a fflush(). This bug looks as though it goes back to the
1656 introduction of TLS in release 3.20, but "sometimes" must have been rare
1657 because the tests only now provoked it.
1663 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1664 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1665 Now it returns an integer. A number of calls were still expecting a T/F
1666 return. Fortuitously, in all cases, the tests worked in OK situations,
1667 which is the norm. However, things would have gone wrong on any write
1668 failures on the smtp file descriptor. This function is used when sending
1669 messages over SMTP and also when doing verify callouts.
1671 2. When Exim is called to do synchronous delivery of a locally submitted
1672 message (the -odf or -odi options), it no longer closes stderr before doing
1675 3. Implemented the mua_wrapper option.
1677 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1679 5. Implemented the functions header_remove(), header_testname(),
1680 header_add_at_position(), and receive_remove_recipient(), and exported them
1683 6. If an ACL "warn" statement specified the addition of headers, Exim already
1684 inserted X-ACL-Warn: at the start if there was no header name. However, it
1685 was not making this test for the second and subsequent header lines if
1686 there were newlines in the string. This meant that an invalid header could
1687 be inserted if Exim was badly configured.
1689 7. Allow an ACL "warn" statement to add header lines at the start or after all
1690 the Received: headers, as well as at the end.
1692 8. Added the rcpt_4xx retry error code.
1694 9. Added postmaster_mailfrom=xxx to callout verification option.
1696 10. Added mailfrom=xxxx to the callout verification option, for verify=
1699 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1700 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1702 12. Inserted some casts to stop certain compilers warning when using pointer
1703 differences as field lengths or precisions in printf-type calls (mostly
1704 affecting debugging statements).
1706 13. Added optional readline() support for -be (dynamically loaded).
1708 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1709 same clock tick as a message's arrival, so that its received time was the
1710 same as the "first fail" time on the retry record, and that message
1711 remained on the queue past the ultimate address timeout, every queue runner
1712 would try a delivery (because it was past the ultimate address timeout) but
1713 after another failure, the ultimate address timeout, which should have then
1714 bounced the address, did not kick in. This was a "< instead of <=" error;
1715 in most cases the first failure would have been in the next clock tick
1716 after the received time, and all would be well.
1718 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1719 being recognized when the domain list was tested by the match_domain
1720 condition in an expansion string.
1722 16. Added the ${str2b64: operator.
1724 17. Exim was always calling setrlimit() to set a large limit for the number of
1725 processes, without checking whether the existing limit was already
1726 adequate. (It did check for the limit on file descriptors.) Furthermore,
1727 errors from getrlimit() and setrlimit() were being ignored. Now they are
1728 logged to the main and panic logs, but Exim does carry on, to try to do its
1729 job under whatever limits there are.
1731 18. Imported PCRE 5.0.
1733 19. Trivial typo in log message " temporarily refused connection" (the leading
1736 20. If the log selector return_path_on_delivery was set and an address was
1737 redirected to /dev/null, the delivery process crashed because it assumed
1738 that a return path would always be set for a "successful" delivery. In this
1739 case, the whole delivery is bypassed as an optimization, and therefore no
1742 21. Internal re-arrangement: the function for sending a challenge and reading
1743 a response while authentication was assuming a zero-terminated challenge
1744 string. It's now changed to take a pointer and a length, to allow for
1745 binary data in such strings.
1747 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1749 23. Exim was not respecting finduser_retries when seeking the login of the
1750 uid under which it was called; it was always trying 10 times. (The default
1751 setting of finduser_retries is zero.) Also, it was sleeping after the final
1752 failure, which is pointless.
1754 24. Implemented tls_on_connect_ports.
1756 25. Implemented acl_smtp_predata.
1758 26. If the domain in control=submission is set empty, Exim assumes that the
1759 authenticated id is a complete email address when it generates From: or
1760 Sender: header lines.
1762 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1763 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1764 chown and chgrp in /bin and hostname in /usr/bin.
1766 28. Exim was keeping the "process log" file open after each use, just as it
1767 does for the main log. This opens the possibility of it remaining open for
1768 long periods when the USR1 signal hits a daemon. Occasional processlog
1769 errors were reported, that could have been caused by this. Anyway, it seems
1770 much more sensible not to leave this file open at all, so that is what now
1773 29. The long-running daemon process does not normally write to the log once it
1774 has entered its main loop, and it closes the log before doing so. This is
1775 so that log files can straightforwardly be renamed and moved. However,
1776 there are a couple of unusual error situations where the daemon does write
1777 log entries, and I had neglected to close the log afterwards.
1779 30. The text of an SMTP error response that was received during a remote
1780 delivery was being truncated at 512 bytes. This is too short for some of
1781 the long messages that one sometimes sees. I've increased the limit to
1784 31. It is now possible to make retry rules that apply only when a message has a
1785 specific sender, in particular, an empty sender.
1787 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1788 it possible to be selective about when SMTP synchronization is enforced.
1790 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1792 32. Implemented hosts_connection_nolog.
1794 33. Added an ACL for QUIT.
1796 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1799 35. Added mailbox_size and mailbox_filecount to appendfile.
1801 36. Added control = no_multiline_responses to ACLs.
1803 37. There was a bug in the logic of the code that waits for the clock to tick
1804 in the case where the clock went backwards by a substantial amount such
1805 that the microsecond fraction of "now" was more than the microsecond
1806 fraction of "then" (but the whole seconds number was less).
1808 38. Added support for the libradius Radius client library this is found on
1809 FreeBSD (previously only the radiusclient library was supported).
1815 1. When certain lookups returned multiple values in the form name=value, the
1816 quoting of the values was not always being done properly. Specifically:
1817 (a) If the value started with a double quote, but contained no whitespace,
1819 (b) If the value contained whitespace other than a space character (i.e.
1820 tabs or newlines or carriage returns) it was not quoted.
1821 This fix has been applied to the mysql and pgsql lookups by writing a
1822 separate quoting function and calling it from the lookup code. The fix
1823 should probably also be applied to nisplus, ibase and oracle lookups, but
1824 since I cannot test any of those, I have not disturbed their existing code.
1826 2. A hit in the callout cache for a specific address caused a log line with no
1827 reason for rejecting RCPT. Now it says "Previous (cached) callout
1828 verification failure".
1830 3. There was an off-by-one bug in the queryprogram router. An over-long
1831 return line was truncated at 256 instead of 255 characters, thereby
1832 overflowing its buffer with the terminating zero. As well as fixing this, I
1833 have increased the buffer size to 1024 (and made a note to document this).
1835 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1836 when Exim is waiting for an SMTP response from a remote server, Exim
1837 restarts its select() call on the socket, thereby resetting its timeout.
1838 This is not a problem when such interrupts are rare. Somebody set up a cron
1839 job to run exiwhat every 2 minutes, which is less than the normal select()
1840 timeout (5 or 10 minutes). This meant that the select() timeout never
1841 kicked in because it was always reset. I have fixed this by comparing the
1842 time when an interrupt arrives with the time at the start of the first call
1843 to select(). If more time than the timeout has elapsed, the interrupt is
1844 treated as a timeout.
1846 5. Some internal re-factoring in preparation for the addition of Sieve
1847 extensions (by MH). In particular, the "personal" test is moved to a
1848 separate function, and given an option for scanning Cc: and Bcc: (which is
1849 not set for Exim filters).
1851 6. When Exim created an email address using the login of the caller as the
1852 local part (e.g. when creating a From: or Sender: header line), it was not
1853 quoting the local part when it contained special characters such as @.
1855 7. Installed new OpenBSD configuration files.
1857 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1858 try to make them clearer.
1860 9. Callout options, other than the timeout value, were being ignored when
1861 verifying sender addresses in header lines. For example, when using
1863 verify = header_sender/callout=no_cache
1865 the cache was (incorrectly) being used.
1867 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1868 only the Cygwin environment.
1870 11. When return_path_on_delivery was set as a log selector, if different remote
1871 addresses in the same message used different return paths and parallel
1872 remote delivery occurred, the wrong values would sometimes be logged.
1873 (Whenever a remote delivery process finished, the return path value from
1874 the most recently started remote delivery process was logged.)
1876 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1877 lines when AUTH and/or TLS are in use. This is the "received protocol"
1878 field. Exim used to use "asmtp" for authenticated SMTP, without any
1879 indication (in the protocol name) for TLS use. Now it follows the RFC and
1880 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1881 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1882 names appear in log lines as well as in Received: header lines.
1884 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1885 extensions, and comparison tests, and to fix some bugs.
1887 14. Changes to the "personal" filter test:
1889 (1) The test was buggy in that it was just doing the equivalent of
1890 "contains" tests on header lines. For example, if a user's address was
1891 anne@some.where, the "personal" test would incorrectly be true for
1893 To: susanne@some.where
1895 This test is now done by extracting each address from the header in turn,
1896 and checking the entire address. Other tests that are part of "personal"
1897 are now done using regular expressions (for example, to check local parts
1898 of addresses in From: header lines).
1900 (2) The list of non-personal local parts in From: addresses has been
1901 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1902 taken from the Sieve specification recommendations.
1904 (3) If the message contains any header line starting with "List-" it is
1905 treated as non-personal.
1907 (4) The test for "circular" in the Subject: header line has been removed
1908 because it now seems ill-conceived.
1910 15. Minor typos in src/EDITME comments corrected.
1912 16. Installed latest exipick from John Jetmore.
1914 17. If headers_add on a router specified a text string that was too long for
1915 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1916 of string_sprintf() is now avoided.
1918 18. $message_body_size was not set (it was always zero) when running the DATA
1919 ACL and the local_scan() function.
1921 19. For the "mail" command in an Exim filter, no default was being set for
1922 the once_repeat time, causing a random time value to be used if "once" was
1923 specified. (If the value happened to be <= 0, no repeat happened.) The
1924 default is now 0s, meaning "never repeat". The "vacation" command was OK
1925 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1926 (I found it when inspecting the code).
1928 20. There is now an overall timeout for performing a callout verification. It
1929 defaults to 4 times the callout timeout, which applies to individual SMTP
1930 commands during the callout. The overall timeout applies when there is more
1931 than one host that can be tried. The timeout is checked before trying the
1932 next host. This prevents very long delays if there are a large number of
1933 hosts and all are timing out (e.g. when the network connections are timing
1934 out). The value of the overall timeout can be changed by specifying an
1935 additional sub-option for "callout", called "maxwait". For example:
1937 verify = sender/callout=5s,maxwait=20s
1939 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1940 to the end before writing, but this should make it even safer).
1942 22. Exim was forgetting that it had advertised PIPELINING for the second and
1943 subsequent messages on an SMTP connection. It was also not resetting its
1944 memory on STARTTLS and an internal HELO.
1946 23. When Exim logs an SMTP synchronization error within a session, it now
1947 records whether PIPELINING has been advertised or not.
1949 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1950 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1951 rather than long int.
1953 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1955 26. Added the never_mail option to autoreply.
1961 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1962 crash if the getsockname() call failed; this can happen if a connection is
1963 closed very soon after it is established. The problem was simply in the
1964 order in which certain operations were done, causing Exim to try to write
1965 to the SMTP stream before it had set up the file descriptor. The bug has
1966 been fixed by making things happen in the correct order.
1972 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1973 before the connection was closed, thus losing the rejection response.
1975 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1976 some early Solaris releases, but causes trouble in current releases where
1977 socklen_t is defined.
1979 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1982 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1985 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1986 that releasing the top of it at the end releases what was used for sub-
1987 expansions (unless the block got too big). However, discard this block if
1988 the first thing is a variable or header, so that we can use its block when
1989 it is dynamic (useful for very large $message_headers, for example).
1991 6. Lookups now cache *every* query, not just the most recent. A new, separate
1992 store pool is used for this. It can be recovered when all lookup caches are
1993 flushed. Lookups now release memory at the end of their result strings.
1994 This has involved some general refactoring of the lookup sources.
1996 7. Some code has been added to the store_xxx() functions to reduce the amount
1997 of flapping under certain conditions.
1999 8. log_incoming_interface used to affect only the <= reception log lines. Now
2000 it causes the local interface and port to be added to several more SMTP log
2001 lines, for example "SMTP connection from", and rejection lines.
2003 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2005 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2007 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2008 could be overwritten at the end of the current message (or the start of a
2009 new message if it was set in a HELO ACL). The value is now preserved for
2010 the duration of the SMTP connection.
2012 12. If a transport had a headers_rewrite setting, and a matching header line
2013 contained an unqualified address, that address was qualified, even if it
2014 did not match any rewriting rules. The underlying bug was that the values
2015 of the flags that permit the existence of unqualified sender and recipient
2016 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2017 non-local messages, and by -bnq for local messages) were not being
2018 preserved with the message after it was received.
2020 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2021 "next input=" as part of the text comprising the host identity instead of
2022 the correct text. The code was using the same buffer for two different
2023 strings. However, depending on which order the printing function evaluated
2024 its arguments, the bug did not always show up. Under Linux, for example, my
2025 test suite worked just fine.
2027 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2028 doesn't work with some older versions of Perl. It has been changed to "my",
2029 which in any case is probably the better facility to use.
2031 15. A really picky compiler found some instances of statements for creating
2032 error messages that either had too many or two few arguments for the format
2035 16. The size of the buffer for calls to the DNS resolver has been increased
2036 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2037 for addresses that have a lot of PTR records. This alleviates a problem; it
2038 does not fully solve it.
2040 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2041 buffer now truncates the list and logs the incident, which is the same
2042 action as happens when Exim is looking up a host name and its aliases.
2043 Previously in this situation something unpredictable would happen;
2044 sometimes it was "internal error: store_reset failed".
2046 18. If a server dropped the connection unexpectedly when an Exim client was
2047 using GnuTLS and trying to read a response, the client delivery process
2048 crashed while trying to generate an error log message.
2050 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2051 string, for example:
2053 warn message = H1: something\nH2: something
2055 the text was added as a single header line from Exim's point of view
2056 though it ended up OK in the delivered message. However, searching for the
2057 second and subsequent header lines using $h_h2: did not work. This has been
2058 fixed. Similarly, if a system filter added multiple headers in this way,
2059 the routers could not see them.
2061 20. Expanded the error message when iplsearch is called with an invalid key to
2062 suggest using net-iplsearch in a host list.
2064 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2065 ACLs are no longer actually imposed (and a message to that effect is
2068 22. If a "gecos" field in a passwd entry contained escaped characters, in
2069 particular, if it contained a \" sequence, Exim got it wrong when building
2070 a From: or a Sender: header from that name. A second bug also caused
2071 incorrect handling when an unquoted " was present following a character
2072 that needed quoting.
2074 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2075 was not being matched caselessly.
2077 24. Arranged for all hyphens in the exim.8 source to be escaped with
2080 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2081 properly. Recipient callout cache records were still being keyed to include
2082 the sender, even when use_sender was set false. This led to far more
2083 callouts that were necessary. The sender is no longer included in the key
2084 when use_sender is false.
2086 26. Added "control = submission" modifier to ACLs.
2088 27. Added the ${base62d: operator to decode base 62 numbers.
2090 28. dnsdb lookups can now access SRV records.
2092 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2093 the configuration file.
2095 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2096 (-v) mode. This makes the output for a verbose queue run more intelligible.
2098 31. Added a use_postmaster feature to recipient callouts.
2100 32. Added the $body_zerocount variable, containing the number of binary zero
2101 bytes in the message body.
2103 33. The time of last modification of the "new" subdirectory is now used as the
2104 "mailbox time last read" when there is a quota error for a maildir
2107 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2109 35. Added +ignore_unknown as a special item in host lists.
2111 36. Code for decoding IPv6 addresses in host lists is now included, even if
2112 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2113 address was recognized as an IP address, but was then not correctly decoded
2114 into binary, causing unexpected and incorrect effects when compared with
2121 1. Very minor rewording of debugging text in manualroute to say "list of
2122 hosts" instead of "hostlist".
2124 2. If verify=header_syntax was set, and a header line with an unqualified
2125 address (no domain) and a large number of spaces between the end of the
2126 name and the colon was received, the reception process suffered a buffer
2127 overflow, and (when I tested it) crashed. This was caused by some obsolete
2128 code that should have been removed. The fix is to remove it!
2130 3. When running in the test harness, delay a bit after writing a bounce
2131 message to get a bit more predictability in the log output.
2133 4. Added a call to search_tidyup() just before forking a reception process. In
2134 theory, someone could use a lookup in the expansion of smtp_accept_max_
2135 per_host which, without the tidyup, could leave open a database connection.
2137 5. Added the variables $recipient_data and $sender_data which get set from a
2138 lookup success in an ACL "recipients" or "senders" condition, or a router
2139 "senders" option, similar to $domain_data and $local_part_data.
2141 6. Moved the writing of debug_print from before to after the "senders" test
2144 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2145 problems for message scanning, either using a data ACL, or using
2146 local_scan() because the Received: header was not generated till after they
2147 were called (in order to set the time as the time of reception completion).
2148 I have revised the way this works. The header is now generated after the
2149 body is received, but before the ACL or local_scan() are called. After they
2150 are run, the timestamp in the header is updated.
2156 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2157 before starting a queue runner without re-exec. This happened only when
2158 deliver_drop_privilege was set or when the Exim user was set to root. The
2159 effect of the bug was that timeouts during subsequent deliveries caused
2160 crashes instead of being properly handled. The handler is now left at its
2161 default (and expected) setting.
2163 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2164 message, again when deliver_drop_privilege is set or Exim is run as root.
2165 The bug described in (1) was not present in this case, but the tidying up
2166 of the other signals was missing. I have made the two cases consistent.
2168 3. The ignore_target_hosts setting on a manualroute router was being ignored
2169 for hosts that were looked up using the /MX notation.
2171 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2174 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2175 operated on the sender address. After changing the $sender_address to <>
2176 for the sender address verify, Exim was re-instated it as the original
2177 (before rewriting) address, but remembering that it had rewritten it, so it
2178 wasn't rewriting it again. This bug also had the effect of breaking the
2179 sender address verification caching when the sender address was rewritten.
2181 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2182 This has been changed so that if the ip literal address matches
2183 ignore_target_hosts, the router declines.
2185 7. Added expansion conditions match_domain, match_address, and match_local_
2186 part (NOT match_host).
2188 8. The placeholder for the Received: header didn't have a length field set.
2190 9. Added code to Exim itself and to exim_lock to test for a specific race
2191 condition that could lead to file corruption when using MBX delivery. The
2192 issue is with the lockfile that is created in /tmp. If this file is removed
2193 after a process has opened it but before that process has acquired a lock,
2194 there is the potential for a second process to recreate the file and also
2195 acquire a lock. This could lead to two Exim processes writing to the file
2196 at the same time. The added code performs the same test as UW imapd; it
2197 checks after acquiring the lock that its file descriptor still refers to
2198 the same named file.
2200 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2201 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2204 11. Added the smtp_active_hostname option. If used, this will typically be made
2205 to depend on the incoming interface address. Because $interface_address is
2206 not set up until the daemon has forked a reception process, error responses
2207 that can happen earlier (such as "too many connections") no longer contain
2210 12. If an expansion in a condition on a "warn" statement fails because a lookup
2211 defers, the "warn" statement is abandoned, and the next ACL statement is
2212 processed. Previously this caused the whole ACL to be aborted.
2214 13. Added the iplsearch lookup type.
2216 14. Added ident_timeout as a log selector.
2218 15. Added tls_certificate_verified as a log selector.
2220 16. Added a global option tls_require_ciphers (compare the smtp transport
2221 option of the same name). This controls incoming TLS connections.
2223 17. I finally figured out how to make tls_require_ciphers do a similar thing
2224 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2225 before starting the TLS session.
2227 18. Tabs are now shown as \t in -bP output.
2229 19. If the log selector return_path_on_delivery was set, Exim crashed when
2230 bouncing a message because it had too many Received: header lines.
2232 20. If two routers both had headers_remove settings, and the first one included
2233 a superfluous trailing colon, the final name in the first list and the
2234 first name in the second list were incorrectly joined into one item (with a
2235 colon in the middle).
2241 1. Added -C and -D options to the exinext utility, mainly to make it easier
2242 to include in the automated testing, but these could be helpful when
2243 multiple configurations are in use.
2245 2. The exinext utility was not formatting the output nicely when there was
2246 an alternate port involved in the retry record key, nor when there was a
2247 message id as well (for retries that were specific to a specific message
2248 and a specific host). It was also confused by IPv6 addresses, because of
2249 the additional colons they contain. I have fixed the IPv4 problem, and
2250 patched it up to do a reasonable job for IPv6.
2252 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2253 delivery, the log line now contains "pipelined" if PIPELINING was used.
2255 4. An SMTP transport process used to panic and die if the bind() call to set
2256 an explicit outgoing interface failed. This has been changed; it is now
2257 treated in the same way as a connect() failure.
2259 5. A reference to $sender_host_name in the part of a conditional expansion
2260 that was being skipped was still causing a DNS lookup. This no longer
2263 6. The def: expansion condition was not recognizing references to header lines
2264 that used bh_ and bheader_.
2266 7. Added the _cache feature to named lists.
2268 8. The code for checking quota_filecount in the appendfile transport was
2269 allowing one more file than it should have been.
2271 9. For compatibility with Sendmail, the command line option
2279 and sets the incoming protocol and host name (for trusted callers). The
2280 host name and its colon can be omitted when only the protocol is to be set.
2281 Note the Exim already has two private options, -pd and -ps, that refer to
2282 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2283 "s", but I don't think that's a major issue.
2285 10. A number of refactoring changes to the code, none of which should affect
2288 (a) The number of logging options was getting close to filling up the
2289 32-bit word that was used as a bit map. I have split them into two classes:
2290 those that are passed in the argument to log_write(), and those that are
2291 only ever tested independently outside of that function. These are now in
2292 separate 32-bit words, so there is plenty of room for expansion again.
2293 There is no change in the user interface or the logging behaviour.
2295 (b) When building, for example, log lines, the code previously used a
2296 macro that called string_cat() twice, in order to add two strings. This is
2297 not really sufficiently general. Furthermore, there was one instance where
2298 it was actually wrong because one of the argument was used twice, and in
2299 one call a function was used. (As it happened, calling the function twice
2300 did not affect the overall behaviour.) The macro has been replaced by a
2301 function that can join an arbitrary number of extra strings onto a growing
2304 (c) The code for expansion conditions now uses a table and a binary chop
2305 instead of a serial search (which was left over from when there were very
2306 few conditions). Also, it now recognizes conditions like "pam" even when
2307 the relevant support is not compiled in: a suitably worded error message is
2308 given if an attempt is made to use such a condition.
2310 11. Added ${time_interval:xxxxx}.
2312 12. A bug was causing one of the ddress fields not to be passed back correctly
2313 from remote delivery subprocesses. The field in question was not being
2314 subsequently used, so this caused to problems in practice.
2316 13. Added new log selectors queue_time and deliver_time.
2318 14. Might have fixed a bug in maildirsizefile handling that threw up
2319 "unexpected character" debug warnings, and recalculated the data
2320 unnecessarily. In any case, I expanded the warning message to give more
2323 15. Added the message "Restricted characters in address" to the statements in
2324 the default ACL that block characters like @ and % in local parts.
2326 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2327 Three changes have been made:
2329 (a) There was a serious bug; a negative response to MAIL caused the whole
2330 recipient domain to be cached as invalid, thereby blocking all messages
2331 to all local parts at the same domain, from all senders. This bug has
2332 been fixed. The domain is no longer cached after a negative response to
2333 MAIL if the sender used is not empty.
2335 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2338 (c) A new callout option, "use_sender" has been added for people who want
2339 the modified behaviour.
2345 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2348 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2349 indeed breaks things for older releases.
2351 3. Added additional logging to the case where there is a problem reading data
2352 from a filter that is running in a subprocess using a pipe, in order to
2353 try to track down a specific problem.
2355 4. Testing facility fudge: when running in the test harness and attempting
2356 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2357 getting "No route to host". Convert this to a timeout.
2359 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2362 6. Some OS don't have socklen_t but use size_t instead. This affects the
2363 fifth argument of getsockopt() amongst other things. This is now
2364 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2365 can be set for individual OS. I have set it for SunOS5, OSF1, and
2366 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2367 some earlier ones do not.
2369 7. Change 4.30/15 was not doing the test caselessly.
2371 8. The standard form for an IPv6 address literal was being rejected by address
2372 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2373 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2374 this, as well as the form without the "IPv6" on the front (but only when
2375 address literals are enabled, of course).
2377 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2379 10. Exim crashed if a message with an empty sender address specified by -f
2380 encountered a router with an errors_to setting. This could be provoked only
2381 by a command such as
2385 where an empty string was supplied; "<>" did not hit this bug.
2387 11. Installed PCRE release 4.5.
2389 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2390 remained set. It is now erased.
2392 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2393 times from message ids (which are base 36 rather than the normal 62).
2395 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2396 were being counted as actual protocol errors, and logged if the log
2397 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2398 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2399 and DATA following a set of rejected RCPTs do not count as protocol errors.
2400 In other words, Exim assumes they were pipelined, though this may not
2401 actually be the case. Of course, in all cases the client gets an
2402 appropriate error code.
2404 15. If a lookup fails in an ACL condition, a message about the failure may
2405 be available; it is used if testing the ACL cannot continue, because most
2406 such messages specify what the cause of the deferral is. However, some
2407 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2408 that caused an old message to be retained and used if a later statement
2409 caused a defer, replacing the real cause of the deferral.
2411 16. If an IP address had so many PTR records that the DNS lookup buffer
2412 was not large enough to hold them, Exim could crash while trying to process
2413 the truncated data. It now detects and logs this case.
2415 17. Further to 4.21/58, another change has been made: if (and only if) the
2416 first line of a message (the first header line) ends with CRLF, a bare LF
2417 in a subsequent header line has a space inserted after it, so as not to
2418 terminate the header.
2420 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2421 unnecessarily, used atoi() instead of strtol(), and didn't check the
2422 termination when getting file sizes from file names by regex.
2424 19. Completely re-implemented the support for maildirsize files, in the light
2425 of a number of problems with the previous contributed implementation
2426 (4.30/29). In particular:
2428 . If the quota is zero, the maildirsize file is maintained, but no quota is
2431 . If the maildir directory does not exist, it is created before any attempt
2432 to write a maildirsize file.
2434 . The quota value in the file is just a cache; if the quota is changed in
2435 the transport, the new value overrides.
2437 . A regular expression is available for excluding directories from the
2440 20. The autoreply transport checks the characters in options that define the
2441 message's headers; it allows continued headers, but it was checking with
2442 isspace() after an embedded newline instead of explicitly looking for a
2445 21. If all the "regular" hosts to which an address was routed had passed their
2446 expiry times, and had not reached their retry times, the address was
2447 bounced, even if fallback hosts were defined. Now Exim should go on to try
2450 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2451 equivalent code in the SMTP transport. Some hosts send humungous responses
2452 to HELO/EHLO, more than 1024 it seems.
2454 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2455 gives compiler warnings in some environments. I've now done it "properly",
2458 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2459 (because of problems with the built-in one) was declared to return uschar *
2460 instead of char *, causing compiler failure.
2462 25. Fixed a file descriptor leak when processing alias/forward files.
2464 26. Fixed a minor format string issue in dbfn.c.
2466 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2468 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2469 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2472 29. When a sender address is verified, it is cached, to save repeating the test
2473 when there is more than one recipient in a message. However, when the
2474 verification involves a callout, it is possible for different callout
2475 options to be set for different recipients. It is too complicated to keep
2476 track of this in the cache, so now Exim always runs a verification when a
2477 callout is required, relying on the callout cache for the optimization.
2478 The overhead is duplication of the address routing, but this should not be
2481 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2482 to be verified with callout=postmaster, and the main callout worked but the
2483 postmaster check failed, the verification correctly failed. However, if a
2484 subsequent RCPT command asked for sender verification *without* the
2485 postmaster check, incorrect caching caused this verification also to fail,
2488 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2489 it was not caching the DNS options (qualify_single, search_parents) that
2490 were used when the lookup failed. A subsequent lookup with different
2491 options therefore always gave the same answer, though there were cases
2492 where it should not have. (Example: a "domains = !$mx_any" option on a
2493 dnslookup router: the "domains" option is always processed without any
2494 widening, but the router might have qualify_single set.) Now Exim uses the
2495 cached value only when the same options are set.
2497 32. Added John Jetmore's "exipick" utility to the distribution.
2499 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2500 than a timeout (e.g. a certificate is required, and is not provided), an
2501 Exim server now closes the connection immediately. Previously it waited for
2502 the client to close - but if the client is SSL, it seems that they each
2503 wait for each other, leading to a delay before one of them times out.
2505 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2506 maintained 0.8.x compatibility because I don't think many are using it, and
2507 it is clearly obsolete.
2509 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2512 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2513 client certificate was expired. A simple patch fixes this, though I don't
2514 understand the full logic of why the verify callback is called multiple
2517 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2518 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2519 which causes problems with some clients (such as the Certicom SSL Plus
2520 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2521 disables the coutermeasure allowing Eudora to connect."
2523 38. Exim was not checking that a write() to a log file succeeded. This could
2524 lead to Bad Things if a log got too big, in particular if it hit a file
2525 size limit. Exim now panics and dies if it cannot write to a log file, just
2526 as it does if it cannot open a log file.
2528 39. Modified OS/Makefile-Linux so that it now contains
2530 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2532 The two -D definitions ensure that Exim is compiled with large file
2533 support, which makes it possible to handle log files that are bigger than
2536 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2537 instance) a domain was checked against a named list that involved a lookup,
2538 causing $domain_data to be set, then another domain was checked against the
2539 same list, then the first domain was re-checked, the value of $domain_data
2540 after the final check could be wrong. In particular, if the second check
2541 failed, it could be set empty. This bug probably also applied to
2544 41. The strip_trailing_dot option was not being applied to the address given
2545 with the -f command-line option.
2547 42. The code for reading a message's header from the spool was incrementing
2548 $received_count, but never initializing it. This meant that the value was
2549 incorrect (doubled) while delivering a message in the same process in which
2550 it was received. In the most common configuration of Exim, this never
2551 happens - a fresh exec is done - but it can happen when
2552 deliver_drop_privilege is set.
2554 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2555 it now includes up to 150 characters of the unexpected data in the log
2558 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2559 and building data strings. The size of both of these buffers was 10 000
2560 bytes - far larger than anybody would *ever* want, thought I. Needless to
2561 say, somebody hit the limit. I have increased the maximum line length to
2562 20 000 and the maximum data length of concatenated lines to 100 000. I have
2563 also fixed two bugs, because there was no checking on these buffers. Tsk,
2564 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2565 buffer is too small.
2567 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2568 lsearch lookups. Now it does.
2570 46. When parsing a route_list item in a manualroute router, a fixed-length
2571 buffer was used for the list of hosts. I made this 1024 bytes long,
2572 thinking that nobody would ever have a list of hosts that long. Wrong.
2573 Somebody had a whole pile of complicated expansion conditions, and the
2574 string was silently truncated, leading to an expansion error. It turns out
2575 that it is easier to change to an unlimited length (owing to other changes
2576 that have happened since this code was originally written) than to build
2577 structure for giving a limitation error. The length of the item that
2578 expands into the list of hosts is now unlimited.
2580 47. The lsearch lookup could not handle data where the length of text line was
2581 more than 4095 characters. Such lines were truncated, leading to shortened
2582 data being returned. It should now handle lines of any length.
2584 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2585 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2588 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2589 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2590 no longer happens in comment lines. A list of replacements is now placed
2591 at the head of all of the source files, except those whose only change is
2592 to replace PERL_COMMAND in the very first #! line.
2594 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2595 messages on the queue, with a bottom-up merge sort, using code contributed
2596 by Michael Haardt. This should make operations like -bp somewhat faster on
2597 large queues. It won't affect queue runners, except when queue_run_in_order
2600 51. Installed eximstats 1.31 in the distribution.
2602 52. Added support for SRV lookups to the dnslookup router.
2604 53. If an ACL referred to $message_body or $message_body_end, the value was not
2605 reset for any messages that followed in the same SMTP session.
2607 54. The store-handling optimization for building very long strings was not
2608 differentiating between the different store pools. I don't think this
2609 actually made any difference in practice, but I've tidied it.
2611 55. While running the routers to verify a sender address, $sender_address
2612 was still set to the sender address. This is wrong, because when routing to
2613 send a bounce to the sender, it would be empty. Therefore, I have changed
2614 it so that, while verifying a sender address, $sender_address is set to <>.
2615 (There is no change to what happens when verifying a recipient address.)
2617 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2618 target A or AAAA records (if not already returned) without resetting the
2619 qualify_single or search_parents options of the DNS resolver. These are
2620 inappropriate in this case because the targets of MX and SRV records must
2621 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2622 target that, when qualified, matched something in the local domain. These
2623 two options are now turned off when doing these lookups.
2625 57. It seems that at least some releases of Reiserfs (which does not have the
2626 concept of a fixed number of inodes) returns zero and not -1 for the
2627 number of available inodes. This interacted badly with check_spool_inodes,
2628 which assumed that -1 was the "no such thing" setting. What I have done is
2629 to check that the total number of inodes is greater than zero before doing
2630 the test of how many are available.
2632 58. When a "warn" ACL statement has a log_message modifier, the message is
2633 remembered, and not repeated. This is to avoid a lot of repetition when a
2634 message has many recipients that cause the same warning to be written.
2635 Howewer, Exim was preserving the list of already written lines for an
2636 entire SMTP session, which doesn't seem right. The memory is now reset if a
2637 new message is started.
2639 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2640 debugging output unless log_rewrite was also set.
2642 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2643 of (int)(handle) when we know that handle contains (void *)(-1).
2645 61. The Exim daemon panic-logs an error return when it closes the incoming
2646 connection. However "connection reset by peer" seems to be common, and
2647 isn't really an error worthy of noting specially, so that particular error
2650 62. When Exim is trying to find all the local interfaces, it used to panic and
2651 die if the ioctl to get the interface flags failed. However, it seems that
2652 on at least one OS (Solaris 9) it is possible to have an interface that is
2653 included in the list of interfaces, but for which you get a failure error
2654 for this call. This happens when the interface is not "plumbed" into a
2655 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2656 failure of the "get flags" call assumes that the interface is down.
2658 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2659 makes life easier for people who are doing arithmetic on fields extracted
2660 from dates, where you often get leading zeros that should not be
2661 interpreted as octal.
2663 64. Added qualify_domain to the redirect router, to override the global
2666 65. If a pathologically long header line contained very many addresses (the
2667 report of this problem mentioned 10 000) and each of them was rewritten,
2668 Exim could use up a very large amount of memory. (It kept on making new
2669 copies of the header line as it rewrote, and never released the old ones.)
2670 At the expense of a bit more processing, the header rewriting function has
2671 been changed so that it no longer eats memory in this way.
2673 66. The generation of the Received: header has been moved from the time that a
2674 message starts to be received, to the time that it finishes. The timestamp
2675 in the Received: header should now be very close to that of the <= log
2676 line. There are two side-effects of this change:
2678 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2679 logged header lines no longer include the local Received: line, because
2680 it has not yet been created. The same applies to a copy of the message
2681 that is returned to a non-SMTP sender when a message is rejected.
2683 (b) When a filter file is tested using -bf, no additional Received: header
2684 is added to the test message. After some thought, I decided that this
2687 This change does not affect the value of $received_for. It is still set
2688 after address rewriting, but before local_scan() is called.
2690 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2692 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2693 gave an unhelpful panic error message, and a defer error. I have managed to
2694 change this behaviour so that it now rejects any supplied certificate,
2695 which seems right, as the list of acceptable certificates is empty.
2697 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2698 gave an unhelpful defer error. I have not managed to make this reject any
2699 supplied certificates, but the error message it gives is "no certificate
2700 supplied", which is not helpful.
2702 70. exigrep's output now also includes lines that are not associated with any
2703 message, but which match the given pattern. Implemented by a patch from
2704 Martin Sluka, which also tidied up the Perl a bit.
2706 71. Recipient callout verification, like sender verification, was using <> in
2707 the MAIL FROM command. This isn't really the right thing, since the actual
2708 sender may affect whether the remote host accepts the recipient or not. I
2709 have changed it to use the actual sender in the callout; this means that
2710 the cache record is now keyed on a recipient/sender pair, not just the
2711 recipient address. There doesn't seem to be a real danger of callout loops,
2712 since a callout by the remote host to check the sender would use <>.
2713 [SEE ABOVE: changed after hitting problems.]
2715 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2716 temporary errors. However, in the case of such a code being given after
2717 the end of a data transmission (i.e. after ".") Exim was failing to write
2718 a retry record for the message. (Yes, there was some broken host that was
2719 actually sending 8xx at this point.)
2721 73. An unknown lookup type in a host list could cause Exim to panic-die when
2722 the list was checked. (An example that provoked this was putting <; in the
2723 middle of a list instead of at the start.) If this happened during a DATA
2724 ACL check, a -D file could be left lying around. This kind of configuration
2725 error no longer causes Exim to die; instead it causes a defer errror. The
2726 incident is still logged to the main and panic logs.
2728 74. Buglet left over from Exim 3 conversion. The message "too many messages
2729 in one connection" was written to the rejectlog but not the mainlog, except
2730 when address rewriting (yes!) was being logged.
2732 75. Added write_rejectlog option.
2734 76. When a system filter was run not as root (that is, when system_filter_user
2735 was set), the values of the $n variables were not being returned to the
2736 main process; thus, they were not subsequently available in the $sn
2739 77. Added +return_path_on_delivery log selector.
2741 78. A connection timeout was being treated differently from recipients deferred
2742 when testing hosts_max_try with a message that was older than the host's
2743 retry timeout. (The host should not be counted, thus allowing all hosts to
2744 be tried at least once before bouncing.) This may have been the cause of an
2745 occasionally reported bug whereby a message would remain on the queue
2746 longer than the retry timeout, but would be bounced if a delivery was
2747 forced. I say "may" because I never totally pinned down the problem;
2748 setting up timeout/retry tests is difficult. See also the next item.
2750 79. The ultimate address timeout was not being applied to errors that involved
2751 a combination of host plus message (for example, a timeout on a MAIL
2752 command). When an address resolved to a number of possible hosts, and they
2753 were not all tried for each delivery (e.g. because of hosts_max_try), a
2754 message could remain on the queue longer than the retry timeout.
2756 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2759 81. Fixed an obscure SMTP outgoing bug which required at least the following
2760 conditions: (a) there was another message waiting for the same server;
2761 (b) the server returned 5xx to all RCPT commands in the first message so
2762 that the message was not completed; (c) the server dropped the connection
2763 or gave a negative response to the RSET that Exim sends to abort the
2764 transaction. The observed case was a dropped connection after DATA that had
2765 been sent in pipelining mode. That is, the server had advertised PIPELINING
2766 but was not implementing it correctly. The effect of the bug was incorrect
2767 behaviour, such as trying another host, and this could lead to a crash.
2773 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2774 and daemon.c were passed as pointers to ints; they should have been
2775 pointers to socklen_t variables (which are typically unsigned ints).
2777 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2780 3. Fixed a really odd bug that affected only the testing scheme; patching a
2781 certain fixed string in the binary changed the value of another string that
2782 happened to be identical to the end of the original first string.
2784 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2785 name", it returns that address as the IP address. On some operating
2786 systems (e.g. Solaris), it also passes back the IP address string as the
2787 "host name". However, on others (e.g. Linux), it passes back an empty
2788 string. Exim wasn't checking for this, and was changing the host name to an
2789 empty string, assuming it had been canonicized.
2791 5. Although rare, it is permitted to have more than one PTR record for a given
2792 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2793 all the names associated with an address, because they do in Solaris.
2794 However, it seems that they do not in Linux for data that comes from the
2795 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2796 I found this out when I moved to a new Linux workstation and tried to run
2797 the Exim test suite.
2799 To get round this problem I have changed the code so that it now does its
2800 own call to the DNS to look up PTR records when searching for a host name.
2801 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2802 addresses that are only in /etc/hosts are still found.
2804 This behaviour is, however, controlled by an option called host_lookup_
2805 order, which defaults to "bydns:byaddr". If people want to use the other
2806 order, or indeed, just use one or the other means of lookup, they can
2807 specify it in this variable.
2809 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2810 some operating systems, this comes back from gethostbyaddr() as an empty
2811 string, and this is what Exim used to test for. However, it seems that in
2812 other systems, "." is yielded. Exim now tests for this case too.
2814 7. The values of check_spool_space and check_log_space are now held internally
2815 as a number of kilobytes instead of an absolute number of bytes. If a
2816 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2817 kilobyte. This means that much larger values can be stored.
2819 8. Exim monitor: an attempt to get the action menu when not actually pointing
2820 at a message produces an empty menu entitled "No message selected". This
2821 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2822 no entries in it ("Shell widget menu has zero width and/or height"). So I
2823 have added a single, blank menu entry in this case.
2825 9. Added ${quote_local_part.
2827 10. MIME decoding is now applied to the contents of Subject: header lines when
2830 11. Now that a reference to $sender_host_address automatically causes a reverse
2831 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2832 host lookup before query-style lookups in lists that might use this
2833 variable. This has therefore been abolished, and the "net-" prefix is no
2834 longer necessary for query-style lookups.
2836 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2837 have been a typo for LFLAGS, so it has been changed.
2839 13. The install script calls Exim with "-C /dev/null" in order to find the
2840 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2841 to be output. Howeve, since Exim outputs its version number before the
2842 error, it didn't break the script. It just looked ugly. I fixed this by
2843 always allowing "-C /dev/null" if the caller is root.
2845 14. Ignore overlarge ACL variable number when reading spool file - insurance
2846 against a later release with more variables having written the file.
2848 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2849 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2850 this, as well as the form without the "IPv6" on the front.
2852 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2853 OS/Makefile-Darwin file.
2855 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2856 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2858 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2859 variable. However, some OS use ints for inodes. I've added cast to long int
2860 to get rid of the compiler warning.
2862 19. I had forgotten to lock out "/../" in configuration file names when
2863 ALT_CONFIG_PREFIX was set.
2865 20. Routers used for verification do not need to specify transports. However,
2866 if such a router generated a host list, and callout was configured, Exim
2867 crashed, because it could not find a port number from the (non-existent)
2868 transport. It now assumes port 25 in this circumstance.
2870 21. Added the -t option to exigrep.
2872 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2873 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2874 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2875 them is set, LOOKUP_LSEARCH is forced.
2877 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2879 24. Added sender and host information to the "rejected by local_scan()" log
2880 line; previously there was no indication of these.
2882 25. Added .include_if_exists.
2884 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2885 Linux. It turns out that not all file systems support this. Apparently some
2886 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2887 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2888 not supported on the file descriptor, is now ignored when Exim is trying to
2889 sync a directory. This applies only to Linux.
2891 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2893 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2894 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2897 29. Added support for maildirsize files from supplied patch (modified a bit).
2899 30. The use of :fail: followed by an empty string could lead Exim to respond to
2900 sender verification failures with (e.g.):
2902 550 Verification failed for <xxx>
2903 550 Sender verify failed
2905 where the first response line was missing the '-' that indicates it is not
2906 the final line of the response.
2908 31. The loop for finding the name of the user that called Exim had a hardwired
2909 limit of 10; it now uses the value of finduser_retries, which is used for
2910 all other user lookups.
2912 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2915 33. Exim was neglecting to zero errno before one call of strtol() when
2916 expanding a string and expecting an integer value. On some systems this
2917 resulted in spurious "integer overflow" errors. Also, it was casting the
2918 result into an int without checking.
2920 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2921 did not work. The code looks as if it has *never* worked, though it appears
2922 to have been documented since at least releast 1.62. I have made it work.
2924 35. The "timeout_DNS" error in retry rules, also documented since at least
2925 1.62, also never worked. As it isn't clear exactly what this means, and
2926 clearly it isn't a major issue, I have abolished the feature by treating it
2927 as "timeout", and writing a warning to the main and panic logs.
2929 36. The display of retry rules for -brt wasn't always showing the error code
2932 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2933 timeout_connect_A, timeout_connect_MX.
2935 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2936 to the empty sender.
2938 39. The daemon was not analysing the content of -oX till after it had closed
2939 stderr and disconnected from the controlling terminal. This meant that any
2940 syntax errors were only noted on the panic log, and the return code from
2941 the command was 0. By re-arranging the code a little, I've made the
2942 decoding happen first, so such errors now appear on stderr, and the return
2943 code is 1. However, the actual setting up of the sockets still happens in
2944 the disconnected process, so errors there are still only recorded on the
2947 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2948 connections (as happens on some IP stacks) was logged at start up time as
2949 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2950 it from "IPv6 and IPv4", which means that two separate sockets are being
2953 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2954 says whether AF_INET or AF_INET6 was passed as an argument.
2956 42. Exiwhat output was messed up when time zones were included in log
2959 43. Exiwhat now gives more information about the daemon's listening ports,
2960 and whether -tls-on-connect was used.
2962 44. The "port" option of the smtp transport is now expanded.
2964 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2965 silently ignored. Now an error message is written to the main and panic
2968 46. There's a new ACL modifier called "logwrite" which writes to a log file
2969 as soon as it is encountered.
2971 47. Added $local_user_uid and $local_user_gid at routing time.
2973 48. Exim crashed when trying to verify a sender address that was being
2976 49. Exim was recognizing only a space character after ".include". It now also
2977 recognizes a tab character.
2979 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2980 extracting the relevant information from the specification. The man page no
2981 longer contains scrambled data for the -d option, and I've added a section
2982 at the front about calling Exim under different names.
2984 51. Added "extra_headers" argument to the "mail" command in filter files.
2986 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2989 53. Installed eximstats 1.29.
2991 54. Added transport_filter_timeout as a generic transport option.
2993 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2994 Cc: header lines. This was required by RFC 822, but it not required by RFC
2997 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2998 incoming messages that did not have them. Now it does so only if the
2999 message originates locally, that is, if there is no associated remote host
3000 address. When Resent- header lines are present, this applies to the Resent-
3001 lines rather than the non-Resent- lines.
3003 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3004 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3006 58. Messages for configuration errors now include the name of the main
3007 configuration file - useful now that there may be more than one file in a
3008 list (.included file names were always shown).
3010 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3011 for those rare installations that do not start the daemon as root or run it
3012 setuid root. I've cut out the call to initgroups() if the daemon is not
3015 60. The Exim user and group can now be bound into the binary as text strings
3016 that are looked up at the start of Exim's processing.
3018 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3020 62. Added $mailstore_basename variable.
3022 63. Installed patch to sieve.c from Michael Haardt.
3024 64. When Exim failed to open the panic log after failing to open the main log,
3025 the original message it was trying to log was written to stderr and debug
3026 output, but if they were not available (the usual case in production), it
3027 was lost. Now it is written to syslog before the two lines that record the
3028 failures to open the logs.
3030 65. Users' Exim filters run in subprocesses under the user's uid. It is
3031 possible for a "deliver" command or an alias in a "personal" command to
3032 provoke an address rewrite. If logging of address rewriting is configured,
3033 this fails because the process is not running as root or exim. There may be
3034 a better way of dealing with this, but for the moment (because 4.30 needs
3035 to be released), I have disabled address rewrite logging when running a
3036 filter in a non-root, non-exim process.
3042 1. The buildconfig auxiliary program wasn't quoting the value set for
3043 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3044 not defined. This bug was masked in 4.22 by the effect that was fixed in
3047 2. Some messages that were rejected after a message id was allocated were
3048 shown as "incomplete" by exigrep. It no longer does this for messages that
3049 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3051 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3052 have allow_domain_literals set, the ID did not get logged in the <= line.
3053 Domain literals are now always recognized in Message-ID: header lines.
3055 4. The first argument for a ${extract expansion item is the key name or field
3056 number. Leading and trailing spaces in this item were not being ignored,
3057 causing some misleading effects.
3059 5. When deliver_drop_privilege was set, single queue runner processes started
3060 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3061 same command in the process it spins off) were not dropping privilege.
3063 6. When the daemon running as "exim" started a queue runner, it always
3064 re-executed Exim in the spun-off process. This is a waste of effort when
3065 deliver_drop_privilege is set. The new process now just calls the
3066 queue-runner function directly.
3072 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3075 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3076 ignored. Though the use of -oP was forcing the writing of a pid file, it
3077 was always written to the default place.
3079 3. If the message "no IP address found for host xxxx" is generated during
3080 incoming verification, it is now followed by identification of the incoming
3081 connection (so you can more easily find what provoked it).
3083 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3085 5. Added some features to "harden" Exim a bit more against certain attacks:
3087 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3088 be put in Local/Makefile. This is like the never_users runtime option,
3089 but it cannot be overridden. The default setting is "root".
3091 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3092 prefix string with which any file named in a -C command line option
3095 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3096 is retained for -C and -D only if the caller of Exim is root. Without
3097 it, the exim user may also use -C and -D and retain privilege.
3099 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3100 command line option is disabled.
3102 6. Macro names set by the -D option must start with an upper case letter, just
3103 like macro names defined in the configuration file.
3105 7. Added "dereference=" facility to LDAP.
3107 8. Two instances of the typo "uknown" in the source files are fixed.
3109 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3110 the Configure-Makefile script screwed up while processing it.
3112 10. Incorporated PCRE 4.4.
3114 11. The SMTP synchronization check was not operating right at the start of an
3115 SMTP session. For example, it could not catch a HELO sent before the client
3116 waited for the greeting. There is now a check for outstanding input at the
3117 point when the greeting is written. Because of the duplex, asynchronous
3118 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3119 way, but not yet received, when the check is performed.
3121 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3122 on TCP/IP sockets, because this apparently causes some broken clients to
3125 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3126 unchanged) from the Cygwin maintainer.
3128 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3129 was supported instead of testing for mbx. Effectively a typo.
3131 15. The spa authenticator server code was not checking that the input it
3132 received was valid base64.
3134 16. The debug output line for the "set" modifier in ACLs was not showing the
3135 name of the variable that was being set.
3137 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3139 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3140 lookup on its own. Something else was needed to trigger the lookup. For
3141 example, a match in host_lookup or the need for a host name in a host list.
3142 Now, if $sender_host_name is referenced and the host name has not yet been
3143 looked up, a lookup is performed. If the lookup fails, the variable remains
3144 empty, and $host_lookup_failed is set to "1".
3146 19. Added "eqi" as a case-independent comparison operator.
3148 20. The saslauthd authentication condition could segfault if neither service
3149 nor realm was specified.
3151 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3152 error message that was logged was misleading, and incoming SMTP
3153 connections were dropped. The message is now more accurate, and temporary
3154 errors are given to SMTP connections.
3156 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3157 (except RSET) with a 421 error, until QUIT is received. However, it was
3158 failing to send a response to QUIT.
3160 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3161 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3162 ACL is now run after the helo_try_verify_hosts code.
3164 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3165 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3166 case-independent, so other case variants are also recognized). Apparently
3167 some systems use these upper case variants.
3169 25. If more than two messages were waiting for the same host, and a transport
3170 filter was specified for the transport, Exim sent two messages over the
3171 same TCP/IP connection, and then failed with "socket operation on non-
3172 socket" when it tried to send the third.
3174 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3176 27. The extern definition of crypt16() in expand.c was not being excluded when
3177 the OS had its own crypt16() function.
3179 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3180 as a preferred synonym for return_size_limit, both as an option and as an
3183 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3185 30. Changed the default configuration ACL to relax the local part checking rule
3186 for addresses that are not in any local domains. For these addresses,
3187 slashes and pipe symbols are allowed within local parts, but the sequence
3188 /../ is explicitly forbidden.
3190 31. SPA server authentication was not clearing the challenge buffer before
3193 32. log_message in a "warn" ACL statement was writing to the reject log as
3194 well as to the main log, which contradicts the documentation and doesn't
3195 seem right (because no rejection is happening). So I have stopped it.
3197 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3198 However, I am unable to do any testing of this.
3200 34. Fixed an infelicity in the appendfile transport. When checking directories
3201 for a mailbox, to see if any needed to be created, it was accidentally
3202 using path names with one or more superfluous leading slashes; tracing
3203 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3205 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3206 contents are added to the log line that is written for every discarded
3207 recipient. (Previously a log_message setting was ignored.)
3209 36. The ${quote: operator now quotes the string if it is empty.
3211 37. The install script runs exim in order to find its version number. If for
3212 some reason other than non-existence or emptiness, which it checks, it
3213 could not run './exim', it was installing it with an empty version number,
3214 i.e. as "exim-". This error state is now caught, and the installation is
3217 38. An argument was missing from the function that creates an error message
3218 when Exim fails to connect to the socket for saslauthd authentication.
3219 This could cause Exim to crash, or give a corrupted message.
3221 39. Added isip, isip4, and isip6 to ${if conditions.
3223 40. The ACL variables $acl_xx are now saved with the message, and can be
3224 accessed later in routers, transports, and filters.
3226 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3227 strings in the file are not string-expanded.
3229 42. If a MAIL command specified a SIZE value that was too large to fit into an
3230 int variable, the check against message_size_limit failed. Such values are
3231 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3232 day this will have to be increased, but I don't think I want to be around
3233 when emails are that large.
3240 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3241 iconv() is not standard in FreeBSD.
3243 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3244 IPv6 enabled. The observed symptom was a segmentation fault on return from
3245 the function os_common_find_running_interfaces() in src/os.c.
3247 3. In the check_special_case() function in daemon.c I had used "errno" as an
3248 argument name, which causes warnings on some systems. This was basically a
3249 typo, since it was named "eno" in the comments!
3251 4. The code that waits for the clock to tick (at a resolution of some fraction
3252 of a second) so as to ensure message-id uniqueness was always waiting for
3253 at least one whole tick, when it could have waited for less. [This is
3254 almost certainly not relevant at current processor speeds, where it is
3255 unlikely to ever wait at all. But we try to future-proof.]
3257 5. The function that sleeps for a time interval that includes fractions of a
3258 second contained a race. It did not block SIGALRM between setting the
3259 timer, and suspending (a couple of lines later). If the interval was short
3260 and the sigsuspend() was delayed until after it had expired, the suspension
3261 never ended. On busy systems this could lead to processes getting stuck for
3264 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3265 process, before it forks any delivery processes. The open lookup caching
3266 mechanism meant that the open file or database connection was passed into
3267 the delivery process. The problem was that delivery processes always tidy
3268 up cached lookup data. This could cause a problem for the next delivery
3269 process started by the queue runner, because the external queue runner
3270 process does not know about the closure. So the next delivery process
3271 still has data in the lookup cache. In the case of a file lookup, there was
3272 no problem because closing a file descriptor in a subprocess doesn't affect
3273 the parent. However, if the lookup was caching a connection to a database,
3274 the connection was closed, and the second delivery process was likely to
3275 see errors such as "PGSQL: query failed: server closed the connection
3276 unexpectedly". The problem has been fixed by closing all cached lookups
3277 in a queue runner before running a delivery process.
3279 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3280 seem to have the "const" qualifier which it has on other OS. I've
3283 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3284 *of the same type* (client or server) with the same public name that an
3285 error should be diagnosed.
3287 9. When Exim looked up a host name for an IP address, but failed to find the
3288 original IP address when looking up the host name (a safety check), it
3289 output the message "<ip address> does not match any IP for NULL", which was
3290 confusing, to say the least. The bug was that the host name should have
3291 appeared instead of "NULL".
3293 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3294 user that is built into the binary, and the -C or -D options is used, root
3295 privilege is dropped before the configuration file is read. In addition,
3296 logging is switched to stderr instead of the normal log files. If the
3297 configuration then re-defines the Exim user, the unprivileged environment
3298 is probably not what is expected, so Exim logs a panic warning message (but
3301 However, if deliver_drop_privilege is set, the unprivileged state may well
3302 be exactly what is intended, so the warning has been cut out in that case,
3303 and Exim is allowed to try to write to its normal log files.
3309 1. smtp_return_error_details was not giving details for temporary sender
3310 or receiver verification errors.
3312 2. Diagnose a configuration error if two authenticators have the same public
3315 3. Exim used not to create the message log file for a message until the first
3316 delivery attempt. This could be confusing when incoming messages were held
3317 for policy or load reasons. The message log file is now created at the time
3318 the message is received, and an initial "Received" line is written to it.
3320 4. The automatically generated man page for command line options had a minor
3321 bug that caused no ill effects; however, a more serious problem was that
3322 the procedure for building the man page automatically didn't always
3323 operate. Consequently, release 4.20 contains an out-of-date version. This
3324 shouldn't happen again.
3326 5. When building Exim with embedded Perl support, the script that builds the
3327 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3328 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3330 6. The freeze_tell option was not being used for messages that were frozen on
3331 arrival, either by an ACL or by local_scan().
3333 7. Added the smtp_incomplete_transaction log selector.
3335 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3336 was accepting AUTH without a new EHLO.
3338 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3339 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3342 10. Logging of TCP/IP connections (when configured) now happens in the main
3343 daemon process instead of the child process, so that the TCP/IP connection
3344 count is more accurate (but it can never be perfect).
3346 11. The use of "drop" in a nested ACL was not being handled correctly in the
3347 outer ACL. Now, if condition failure induced by the nested "drop" causes
3348 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3349 or "require"), the connection is dropped.
3351 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3352 that yield "discard" can now be used with an "accept" or a "discard" verb,
3353 but an error is generated for any others (because I can't see a useful way
3354 to define what should happen).
3356 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3357 are now processed in the same way as lines in the Exim configuration file.
3358 In particular, continuation lines are supported.
3360 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3362 15. Added -ti meaning -t -i.
3364 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3365 domains, and warn by logging if others are found.
3367 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3368 returned from GIFCONF when Exim is trying to find the list of interfaces on
3369 a host. The code in os.c has been modified to copy each ifreq to an aligned
3370 structure in all cases.
3372 Also, in some cases, the returned ifreq's were being copied to a 'struct
3373 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3374 means the last couple of bytes of an IPv6 address could be chopped if the
3375 ifreq contained only a normal sockaddr (14 bytes storage).
3377 18. Named domain lists were not supported in the hosts_treat_as_local option.
3378 An entry such as +xxxx was not recognized, and was treated as a literal
3381 19. Ensure that header lines added by a DATA ACL are included in the reject log
3382 if the ACL subsequently rejects the message.
3384 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3385 MD5 (which is deprecated).
3387 21. When testing a filter file using -bf, Exim was writing a message when it
3388 took the sender from a "From " line in the message, but it was not doing so
3389 when it took $return_path from a Return-Path: header line. It now does.
3391 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3392 with a valid header line field name (a series of printing characters
3393 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3395 23. Changed "disc" in the source to "disk" to conform to the documentation and
3396 the book and for uniformity.
3398 24. Ignore Sendmail's -Ooption=value command line item.
3400 25. When execve() failed while trying to run a command in a pipe transport,
3401 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3402 could be confused with a return value of 69 from the command itself. This
3403 has been changed to 127, the value the shell returns if it is asked to run
3404 a non-existent command. The wording for the related log line suggests a
3405 non-existent command as the problem.
3407 26. If received_header_text expands to an empty string, do not add a Received:
3408 header line to the message. (Well, it adds a token one on the spool, but
3409 marks it "old" so that it doesn't get used or transmitted.)
3411 27. Installed eximstats 1.28 (addition of -nt option).
3413 28. There was no check for failure on the call to getsockname() in the daemon
3414 code. This can fail if there is a shortage of resources on the system, with
3415 ENOMEM, for example. A temporary error is now given on failure.
3417 29. Contrary to the C standard, it seems that in some environments, the
3418 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3419 program. Exim now does this explicitly; it affects the formatting of
3420 timestamps using strftime().
3422 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3423 complaints. I've now initialized all the variables, to avoid this.
3425 32. Header lines added by a system filter were not being "seen" during
3426 transport-time rewrites.
3428 33. The info_callback() function passed to OpenSSL is set up with type void
3429 (*)(SSL *, int, int), as described somewhere. However, when calling the
3430 function (actually a macro) that sets it up, the type void(*)() is
3431 expected. I've put in a cast to prevent warnings from picky compilers.
3433 34. If a DNS black list lookup found a CNAME record, but there were no A
3434 records associated with the domain it pointed at, Exim crashed.
3436 35. If a DNS black list lookup returned more than one A record, Exim ignored
3437 all but the first. It now scans all returned addresses if a particular IP
3438 value is being sought. In this situation, the contents of the
3439 $dnslist_value variable are a list of all the addresses, separated by a
3442 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3443 to accept a host name and all its aliases if the forward lookup for any of
3444 them yielded the IP address of the incoming connection. Now it accepts only
3445 those names whose forward lookup yields the correct IP address. Any other
3446 names are discarded. This closes a loophole whereby a rogue DNS
3447 administrator could create reverse DNS records to break through a
3448 wildcarded host restriction in an ACL.
3450 37. If a user filter or a system filter that ran in a subprocess used any of
3451 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3452 the wrong values were passed to the pipe command ($thisaddress had the
3453 value of $0, $0 had the value of $1, etc). This bug was introduced by
3454 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3456 38. Improved the line breaking for long SMTP error messages from ACLs.
3457 Previously, if there was no break point between 40 and 75 characters, Exim
3458 left the rest of the message alone. Two changes have been made: (a) I've
3459 reduced the minimum length to 35 characters; (b) if it can't find a break
3460 point between 35 and 75 characters, it looks ahead and uses the first one
3461 that it finds. This may give the occasional overlong line, but at least the
3462 remaining text gets split now.
3464 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3465 file descriptors might be low, and that setting 1000 would always raise it.
3466 It turns out that in some environments, the limit is already over 1000 and
3467 that lowering it causes trouble. So now Exim takes care not to decrease it.
3469 40. When delivering a message, the value of $return_path is set to $sender_
3470 address at the start of routing (routers may change the value). By an
3471 oversight, this default was not being set up when an address was tested by
3472 -bt or -bv, which affected the outcome if any router or filter referred to
3475 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3476 log only when "message" or "log_message" are set. However, if one of the
3477 conditions was an address verification, or a call to a nested ACL, the
3478 messages generated by the underlying test were being passed through. This
3479 no longer happens. The underlying message is available in $acl_verify_
3480 message for both "message" and "log_message" expansions, so it can be
3481 passed through if needed.
3483 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3484 new expansion $bh_ to give the encoded byte string without charset
3485 translation. Translation happens only if iconv() is available; HAVE_ICONV
3486 indicates this at build time. HEADERS_CHARSET gives the charset to
3487 translate to; headers_charset can change it in the configuration, and
3488 "headers charset" can change it in an individual filter file.
3490 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3491 that creates RFC 2047 encoded "words" labels them as that charset instead
3492 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3493 expansion operator; (ii) when Exim creates a From: line for a local
3494 message; (iii) when a header line is rewritten to include a "phrase" part.
3496 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3497 buggy, causing it to skip the first lines of messages whose message ID
3498 ended in 'D'. This would not have bitten before Exim release 4.14, because
3499 message IDs were unlikely to end in 'D' before then. The effect was to have
3500 incorrect size information for certain domains.
3502 45. #include "config.h" was missing at the start of the crypt16.c module. This
3503 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3506 46. If there was a timeout during a "random" callout check, Exim treated it as
3507 a failure of the random address, and carried on sending RSET and the real
3508 address. If the delay was just some slowness somewhere, the response to the
3509 original RCPT would be taken as a response to RSET and so on, causing
3510 mayhem of various kinds.
3512 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3513 when I implemented it. It didn't allow for the fact that some option values
3514 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3515 the right test for positive values.
3517 48. Domain names in DNS records are case-independent. Exim always looks them up
3518 in lower case. Some resolvers return domain names in exactly the case they
3519 appear in the zone file, that is, they may contain uppercase letters. Not
3520 all resolvers do this - some return always lower case. Exim was treating a
3521 change of case by a resolver as a change of domain, similar to a widening
3522 of a domain abbreviation. This triggered its re-routing code and so it was
3523 trying to route what was effectively the same domain again. This normally
3524 caused routing to fail (because the router wouldn't handle the domain
3525 twice). Now Exim checks for this case specially, and just changes the
3526 casing of the domain that it ultimately uses when it transmits the message
3529 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3532 50. If a filter generated a file delivery with a non-absolute name (possible if
3533 no home directory exists for the router), the forbid_file option was not
3536 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3537 the existing equality matching.
3539 52. Exim was using ints instead of ino_t variables in some places where it was
3540 dealing with inode numbers.
3542 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3543 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3544 called TMPDIR, and if it finds it is different, it changes its value.
3546 54. The smtp_printf() function is now made available to local_scan() so
3547 additional output lines can be written before returning. There is also an
3548 smtp_fflush() function to enable the detection of a dropped connection.
3549 The variables smtp_input and smtp_batched_input are exported to
3552 55. Changed the default runtime configuration: the message "Unknown user"
3553 has been removed from the ACL, and instead placed on the localuser router,
3554 using the cannot_route_message feature. This means that any verification
3555 failures that generate their own messages won't get overridden. Similarly,
3556 the "Unrouteable address" message that was in the ACL for unverifiable
3557 relay addresses has also been removed.
3559 56. Added hosts_avoid_esmtp to the smtp transport.
3561 57. The exicyclog script was not checking for the esoteric option
3562 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3563 will work only if exicyclog is run under the appropriate euid.
3565 58. Following a discussion on the list, the rules by which Exim recognises line
3566 endings on incoming messages have been changed. The -dropcr and drop_cr
3567 options are now no-ops, retained only for backwards compatibility. The
3568 following line terminators are recognized: LF CRLF CR. However, special
3569 processing applies to CR:
3571 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3572 nor a local message in the state where . is a terminator.
3574 (ii) If a bare CR is encountered in a header line, an extra space is added
3575 after the line terminator so as not to end the header. The reasoning
3576 behind this is that bare CRs in header lines are most likely either
3577 to be mistakes, or people trying to play silly games.
3579 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3580 was being incorrectly given as 18 bytes larger than it should have been.
3581 This is a VOB (very old bug).
3583 60. This may never have affected anything current, but just in case it has:
3584 When the local host is found other than at the start of a list of hosts,
3585 the local host, those with the same MX, and any that follow, are discarded.
3586 When the list in question was part of a longer list of hosts, the following
3587 hosts (not currently being processed) were also being discarded. This no
3588 longer happens. I'm not sure if this situation could ever has previously
3591 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3594 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3595 header. This is something that is recommended in a new Internet Draft, and
3596 is something that is documented as being done by Sendmail. There are two
3597 possible values. For messages generated by the autoreply transport, Exim
3600 Auto-Submitted: auto-replied
3602 whereas for all other generated messages (e.g. bounces) it adds
3604 Auto-Submitted: auto-generated
3606 63. The "personal" condition in filters now includes a test for the
3607 Auto-Submitted: header. If it contains the string "auto-" the message it
3608 not considered personal.
3610 64. Added rcpt_include_affixes as a generic transport option.
3612 65. Added queue_only_override (default true).
3614 66. Added the syslog_duplication option.
3616 67. If what should have been the first header line of a message consisted of
3617 a space followed by a colon, Exim was mis-interpreting it as a header line.
3618 It isn't of course - it is syntactically invalid and should therefore be
3619 treated as the start of the message body. The misbehaviour could have
3620 caused a number of strange effects, including loss of data in subsequent
3621 header lines, and spool format errors.
3623 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3624 client host had authenticated. This control can now be exercised by an ACL
3625 for more flexibility.
3627 69. By default, callouts do not happen when testing with -bh. There is now a
3628 variant, -bhc, which does actually run the callout code, including
3629 consulting and updating the callout cache.
3631 70. Added support for saslauthd authentication, courtesy of Alexander
3634 71. If statvfs() failed on the spool or log directories while checking their
3635 size for availability, Exim confusingly gave the error "space shortage".
3636 Furthermore, in debugging mode it crashed with a floating point exception.
3637 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3638 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3639 serious problem, Exim now writes to the main and panic logs when this
3640 happens, with details of the failure. It then refuses to accept the
3641 incoming message, giving the message "spool directory problem" or "log
3642 directory problem" with a 421 code for SMTP messages.
3644 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3645 0, 1, and 2 exist, because some OS complain for execs without them (see
3646 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3647 descriptors. However, the code omitted to check that the open succeeded,
3648 causing mysterious errors if for some reason the permissions on /dev/null
3649 got screwed. Now Exim writes a message to the main and panic logs, and
3650 bombs out if it can't open /dev/null.
3652 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3653 interact so that it is all more flexible. It is supposed to remain
3654 backwards compatible. Also added extra_local_interfaces.
3656 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3657 to bomb out with an assertion failure - to the client this appears as a
3658 connection drop. This problem occurs in the part of the code that was taken
3659 from the Samba project. Fortunately, the assertion is in a very simple
3660 function, so I have fixed this by reproducing the function inline in the
3661 one place where it is called, and arranging for authentication to fail
3662 instead of killing the process with assert().
3664 75. The SPA client code was not working when the server requested OEM rather
3665 than Unicode encoding.
3667 76. Added code to make require_files with a specific uid setting more usable in
3668 the case where statting the file as root fails - usually a non-root-mounted
3669 NFS file system. When this happens and the failure is EACCES, Exim now
3670 forks a subprocess and does the per-uid checking as the relevant uid.
3672 77. Added process_log_path.
3674 78. If log_file_path was not explicitly set, a setting of check_log_space or
3675 check_log_inodes was ignored.
3677 79. If a space check for the spool or log partitions fails, the incident is now
3678 logged. Of course, in the latter case the data may get lost...
3680 80. Added the %p formatting code to string_format() so that it can be used to
3681 print addresses in debug_print(). Adjusted all the address printing in the
3682 debugging in store.c to use %p rather than %d.
3684 81. There was a concern that a line of code in smtp_in.c could overflow a
3685 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3686 initially expressed, the concern was not well-founded, because trailing
3687 spaces are removed early. However, if the trailing spaces were followed by
3688 a NULL, they did not get removed, so the overflow was possible. Two fixes
3691 (a) I re-wrote the offending code in a cleaner fashion.
3692 (b) If an incoming SMTP command contains a NULL character, it is rejected
3695 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3696 runs initgroups(), so that if the Exim user is in any additional groups,
3697 they will be used during message reception.
3703 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob