1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.286 2006/02/07 16:36:25 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.
106 PH/01 Two changes to the default runtime configuration:
108 (1) Move the checks for relay_from_hosts and authenticated clients from
109 after to before the (commented out) DNS black list checks.
111 (2) Add control=submission to the relay_from_hosts and authenticated
112 clients checks, on the grounds that messages accepted by these
113 statements are most likely to be submissions.
115 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
117 (1) Generate an error if the third argument for the ${prvs expansion is
120 (2) Treat a missing third argument of ${prvscheck as if it were an empty
123 (3) Reset the variables that are obtained from the first argument of
124 ${prvscheck and used in the second argument before leaving the code,
125 because their memory is reclaimed, so using them afterwards may do
128 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
129 one (it's much easier than Tom thought :-).
131 (5) Because of (4), we can now allow for the use of $prvscheck_result
132 inside the third argument.
134 PH/03 For some reason, the default setting of PATH when running a command from
135 a pipe transport was just "/usr/bin". I have changed it to
138 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
139 anything to be listed in the output from -bV.
141 PH/05 When a filter generated an autoreply, the entire To: header line was
142 quoted in the delivery log line, like this:
144 => >A.N.Other <ano@some.domain> <original@ddress> ...
146 This has been changed so that it extracts the operative address. There
147 may be more than one such address. If so, they are comma-separated, like
150 => >ano@some.domain,ona@other.domain <original@ddress> ...
152 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
153 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
154 not being looked up in the rDNS to get a host name, Exim was showing the
155 IP address twice in Received: lines, even though the IP addresses were
156 identical. For example:
158 Received: from [1.2.3.4] (helo=[1.2.3.4])
160 However, if the real host name was known, it was omitting the HELO data
161 if it matched the actual IP address. This has been tidied up so that it
162 doesn't show the same IP address twice.
164 PH/07 When both +timestamp and +memory debugging was on, the value given by
165 $tod_xxx expansions could be wrong, because the tod_stamp() function was
166 called by the debug printing, thereby overwriting the timestamp buffer.
167 Debugging no longer uses the tod_stamp() function when +timestamp is set.
169 PH/08 When the original message was included in an autoreply transport, it
170 always said "this is a copy of the message, including all the headers",
171 even if body_only or headers_only was set. It now gives an appropriate
174 PH/09 Applied a patch from the Sieve maintainer which:
176 o fixes some comments
177 o adds the (disabled) notify extension core
178 o adds some debug output for the result of if/elsif tests
179 o points to the current vacation draft in the documentation
180 and documents the missing references header update
184 o fixes a bug in processing the envelope test (when testing
185 multiple envelope elements, the last element determinted the
188 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
189 Electronic Mail") by including:
191 Auto-submitted: auto-generated
193 in the messages that it generates (bounce messages and others, such as
194 warnings). In the case of bounce messages for non-SMTP mesages, there was
195 also a typo: it was using "Auto_submitted" (underscore instead of
196 hyphen). Since every message generated by Exim is necessarily in response
197 to another message, thes have all been changed to:
199 Auto-Submitted: auto-replied
201 in accordance with these statements in the RFC:
203 The auto-replied keyword:
205 - SHOULD be used on messages sent in direct response to another
206 message by an automatic process,
208 - MUST NOT be used on manually-generated messages,
210 - MAY be used on Delivery Status Notifications (DSNs) and Message
211 Disposition Notifications (MDNs),
213 - MUST NOT be used on messages generated by automatic or periodic
214 processes, except for messages which are automatic responses to
217 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
218 to the default Received: header definition.
220 PH/12 Added log selector acl_warn_skipped (default on).
222 PH/13 After a successful wildlsearch lookup, discard the values of numeric
223 variables because (a) they are in the wrong storage pool and (b) even if
224 they were copied, it wouldn't work properly because of the caching.
226 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
227 checking when decoding. Apparently there are clients that generate
228 overlong encoded strings. Why am I not surprised?
230 PH/15 If the first argument of "${if match_address" was not empty, but did not
231 contain an "@" character, Exim crashed. Now it writes a panic log message
232 and treats the condition as false.
234 PH/16 In autoreply, treat an empty string for "once" the same as unset.
236 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
237 extension "envelope-auth". The code is finished and in agreement with
238 other implementations, but there is no documentation so far and in fact,
239 nobody wrote the draft yet. This extension is currently #undef'ed, thus
240 not changing the active code.
242 Print executed "if" and "elsif" statements when debugging is used. This
243 helps a great deal to understand what a filter does.
245 Document more things not specified clearly in RFC3028. I had all this
246 sorted out, when out of a sudden new issues came to my mind. Oops."
248 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
251 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
252 canonical form (as documented). However, after a host name lookup from
253 the IP address, check_host() was doing a simple string comparison with
254 addresses acquired from the DNS when checking that the found name did
255 have the original IP as one of its addresses. Since any found IPv6
256 addresses are likely to be in abbreviated form, the comparison could
257 fail. Luckily, there already exists a function for doing the comparison
258 by converting both addresses to binary, so now that is used instead of
261 PH/20 There was another similar case to PH/19, when a complete host name was
262 given in a host list; looking up its IP address could give an abbreviated
263 form, whereas the current host's name might or might not be abbreviated.
264 The same fix has been applied.
270 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
271 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
274 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
275 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
277 PH/03 Typo: missing ".o" in src/pcre/Makefile.
279 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
280 header line, restrict the check to what is listed in RFCs 2369 and 2929.
281 Also, for "Auto-Submitted", treat anything other than "no" as
282 non-personal, in accordance with RFC 3834. (Previously it treated
283 anything starting "auto-" as non-personal.)
285 TF/01 The control=submission/name=... option had a problem with syntax
286 errors if the name included a slash character. The /name= option
287 now slurps the rest of the string, so it can include any characters
288 but it must come last in the list of options (after /sender_retain
291 PH/05 Some modifications to the interface to the fake nameserver for the new
299 TK/01 Added the "success_on_redirect" address verification option. See
300 NewStuff for rationale and an example.
302 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
304 PH/02 Patch to exigrep to allow it to work on syslog lines.
306 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
307 fread() to skip over the body file's header line, because in Cygwin the
308 header line is locked and is inaccessible.
310 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
311 co-exist for some time) to make it clear that it is the Exim ID that is
312 referenced, not the Message-ID: header line.
314 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
315 string_format() function, because snprintf() does not exist on all
318 PH/06 The use of forbid_filter_existstest now also locks out the use of the
319 ${stat: expansion item.
321 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
322 protocol synchronization error", to keep the pedants happy.
324 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
325 well as for IRIX systems, when gcc is being used. See the host.c source
328 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
330 PH/10 Named domain lists were not working if used in a queue_smtp_domains
333 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
334 transport and to the smtp transport in LMTP mode.
336 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
338 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
339 run a filter in a subprocess. This could lead to confusion in subsequent
340 lookups in the parent process. There should also be a search_tidyup() at
341 the end of the subprocess.
343 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
344 only if the host matched helo_try_verify_hosts, which caused the
345 verification to occur when the EHLO/HELO command was issued. The ACL just
346 tested the remembered result. Now, if a previous verification attempt has
347 not happened, "verify = helo" does it there and then.
349 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
351 TK/03 Fix log output including CR from clamd.
353 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
354 exist provoked a memory error which could cause a segfault.
356 PH/15 Installed PCRE 6.2
358 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
360 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
361 of the problem. Specifically, suggested +O2 rather than +O1 for the
364 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
366 PH/20 If a delivery was routed to a non-standard port by means of an SRV
367 record, the port was not correctly logged when the outgoing_port log
368 selector was set (it logged the transort's default port).
370 PH/21 Added support for host-specific ports to manualroute, queryprogram,
371 fallback_hosts, and "hosts" in the smtp transport.
373 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
374 host errors such as "Connection refused".
376 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
377 authentication with radiusclient 0.4.9:
379 - Error returned from rc_read_config was caught wrongly
380 - Username/password not passed on to radius server due to wrong length.
382 The presumption is that some radiusclient API changes for 4.51/PH/17
383 were not taken care of correctly. The code is still untested by me (my
384 Linux distribution still has 0.3.2 of radiusclient), but it was
385 contributed by a Radius user.
387 PH/24 When doing a callout, the value of $domain wasn't set correctly when
388 expanding the "port" option of the smtp transport.
390 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
391 while reading a MIME header. Thanks to Tom Hughes for a patch.
393 PH/24 Include config.h inside local_scan.h so that configuration settings are
396 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
397 that in an ACL for RCPT (for example), you can examine exactly what was
400 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
401 commands, but it was not correctly comparing the address with the actual
402 client host address. Thus, it would show the EHLO address in Received:
403 header lines when this was not necessary.
405 PH/27 Added the % operator to ${eval:}.
407 PH/28 Exim tries to create and chdir to its spool directory when it starts;
408 it should be ignoring failures (because with -C, for example, it has lost
409 privilege). It wasn't ignoring creation failures other than "already
412 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
415 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
416 because an input line was too long, either on its own, or by virtue of
417 too many continuations, the temporary file was not being removed, and the
418 return code was incorrect.
420 PH/31 Missing "BOOL" in function definition in filtertest.c.
422 PH/32 Applied Sieve patches from the maintainer.
424 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
426 PH/33 Added "verify = not_blind".
428 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
429 Local/Makefile (with some defaults set). These are used in built scripts
430 such as exicyclog, but they have never been used in the exim_install
431 script (though there are many overriding facilities there). I have
432 arranged that the exim_install script now takes note of these two
435 PH/35 Installed configuration files for Dragonfly.
437 PH/36 When a locally submitted message by a trusted user did not contain a
438 From: header, and the sender address was obtained from -f or from an SMTP
439 MAIL command, and the trusted user did not use -F to supply a sender
440 name, $originator_name was incorrectly used when constructing a From:
441 header. Furthermore, $originator_name was used for submission mode
442 messages from external hosts without From: headers in a similar way,
443 which is clearly wrong.
445 PH/37 Added control=suppress_local_fixups.
447 PH/38 When log_selector = +received_sender was set, and the addition of the
448 sender made the log line's construction buffer exactly full, or one byte
449 less than full, an overflow happened when the terminating "\n" was
452 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
453 when the result of a list match is failure because a DNS lookup failed.
455 PH/40 RM_COMMAND is now used in the building process.
457 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
458 the "build-* directories that it finds.
460 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
461 address in a domain literal was a prefix of an interface address.
463 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
464 when verifying a sender address, unless rewrite_headers is false.
466 PH/44 Wrote a long comment about why errors_to addresses are verified as
467 recipients, not senders.
469 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
470 the ratelimit ACL was added.
472 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
474 PH/46 Added extra information about PostgreSQL errors to the error string.
476 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
477 avoiding the need to install special zones in a real server. This is
478 backwards compatible; if it can't find the fake resolver, it drops back.
479 Thus, both old and new test suites can be run.
481 TF/02 Added util/ratelimit.pl
483 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
484 clock is set back in time.
486 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
487 Candler <B.Candler@pobox.com>.
489 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
490 OK for addresses that are the result of redirections.
492 PH/48 A number of further additions for the benefit of the new test suite,
493 including a fake gethostbyname() that interfaces to the fake DNS resolver
496 TF/06 The fix for widen_domains has also been applied to qualify_single and
497 search_parents which are the other dnslookup options that can cause
500 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
503 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
505 TF/07 Exim produced the error message "an SRV record indicated no SMTP
506 service" if it encountered an MX record with an empty target hostname.
507 The message is now "an MX or SRV record indicated no SMTP service".
509 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
510 if the DNS of the sending site is misconfigured. This is quite a
511 common situation. This change restores the behaviour of treating a
512 helo verification defer as a failure.
514 PH/51 If self=fail was set on a router, the bounce message did not include the
515 actual error message.
521 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
523 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
524 command itself ended in error, the underlying message about the transport
525 filter timeout was being overwritten with the pipe command error. Now the
526 underlying error message should be appended to the second error message.
528 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
530 PH/02 Reduce the amount of output that "make" produces by default. Full output
531 can still be requested.
533 PH/03 The warning log line about a condition test deferring for a "warn" verb
534 was being output only once per connection, rather than after each
535 occurrence (because it was using the same function as for successful
536 "warn" verbs). This seems wrong, so I have changed it.
538 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
539 it should not have, which might have caused a crash in the right
540 circumstances, but probably never did.
542 PH/04 Installed a modified version of Tony Finch's patch to make submission
543 mode fix the return path as well as the Sender: header line, and to
544 add a /name= option so that you can make the user's friendly name appear
547 TF/03 Added the control = fakedefer ACL modifier.
549 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
550 Mark Lowes for thorough testing.
552 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
554 TK/03 Merged latest SRS patch from Miles Wilton.
556 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
557 with the definition in sysexits.h (which is #included earlier).
558 Fortunately, Exim does not actually use EX_OK. The code used to try to
559 preserve the sysexits.h value, by assumimg that macro definitions were
560 scanned for macro replacements. I have been disabused of this notion,
561 so now the code just undefines EX_OK before #including unistd.h.
563 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
564 in the smtp transport. When a block could not be written in a single
565 write() function, the timeout was being re-applied to each part-write.
566 This seems wrong - if the receiver was accepting one byte at a time it
567 would take for ever. The timeout is now adjusted when this happens. It
568 doesn't have to be particularly precise.
570 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
571 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
573 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
574 without a domain if the check to <postmaster@domain> fails.
576 SC/01 Eximstats: added -xls and the ability to specify output files
577 (patch written by Frank Heydlauf).
579 SC/02 Eximstats: use FileHandles for outputing results.
581 SC/03 Eximstats: allow any combination of xls, txt, and html output.
583 SC/04 Eximstats: fixed display of large numbers with -nvr option
585 SC/05 Eximstats: fixed merging of reports with empty tables.
587 SC/06 Eximstats: added the -include_original_destination flag
589 SC/07 Eximstats: removed tabs and trailing whitespace.
591 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
593 TK/06 MBOX spool code: Add real "From " MBOX separator line
594 so the .eml file is really in mbox format (even though
595 most programs do not really care). Patch from Alex Miller.
597 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
598 The latter is generated from $received_to and is only set if the
599 message has one envelope recipient. SA can use these headers,
600 obviously out-of-the-box. Patch from Alex Miller.
602 PH/08 The ${def test on a variable was returning false if the variable's
603 value was "0", contrary to what the specification has always said!
604 The result should be true unless the variable is empty.
606 PH/09 The syntax error of a character other than { following "${if
607 def:variable_name" (after optional whitespace) was not being diagnosed.
608 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
609 accidental colon was present, for example, could give incorrect results.
611 PH/10 Tidied the code in a number of places where the st_size field of a stat()
612 result is used (not including appendfile, where other changes are about
615 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
616 This involved changing a lot of size variables from int to off_t. It
617 should work with maildirs and everything.
619 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
620 spamd dying while we are connected to it.
622 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
625 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
626 function for BDB 4.3.
628 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
630 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
631 the sources for PCRE have been re-arranged and more files are now
634 PH/15 The code I had for printing potentially long long variables in PH/11
635 above was not the best (it lost precision). The length of off_t variables
636 is now inspected at build time, and an appropriate printing format (%ld
637 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
638 to be "long long int" or "long int". This is needed for the internal
639 formatting function string_vformat().
641 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
642 the configuration file to be ":syslog", then the script "guesses" where
643 the logs files are, rather than using the compiled in default. In our
644 case the guess is not the same as the compiled default, so the script
645 suddenly stopped working when I started to use syslog. The patch checks
646 to see if log_file_path is "". If so, it attempts to read it from exim
647 with no configuration file to get the compiled in version, before it
648 falls back to the previous guessing code."
650 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
651 implementing BATV in an Exim configuration. See NewStuff for the gory
654 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
655 Makefile that are specific to HP-UX.
657 PH/18 If the "use_postmaster" option was set for a recipient callout together
658 with the "random" option, the postmaster address was used as the MAIL
659 FROM address for the random test, but not for the subsequent recipient
660 test. It is now used for both.
662 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
663 patch removes a few documentation additions to RFC 3028, because the
664 latest draft now contains them. It adds the new en;ascii-case comparator
665 and a new error check for 8bit text in MIME parts. Comparator and
666 require names are now matched exactly. I enabled the subaddress
667 extension, but it is not well tested yet (read: it works for me)."
669 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
670 rework some of the code of TK/09 above to avoid the hardwired use of
671 "%lld" and "long long". Replaced the call to snprintf() with a call to
674 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
675 records point to non-existent hosts", "retry timeout exceeded", and
676 "retry time not reached for any host after a long failure period".
678 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
679 experimental DomainKeys support:
681 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
682 (2) On an error such as an illegally used "control", the wrong name for
683 the control was given.
685 These problems did NOT occur unless DomainKeys support was compiled.
687 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
689 PH/24 Added ${if match_ip condition.
691 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
692 databases so that it will be absolutely obvious if a crash occurs in the
693 DB library. This is a regular occurrence (often caused by mis-matched
696 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
697 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
698 picked up on a user's system that detects such things. There doesn't seem
699 to be a gcc warning option for this - only an attribute that has to be
700 put on the function's prototype. It seems that in Fedora Core 4 they have
701 set this on a number of new functions. No doubt there will be more in due
704 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
705 specify a transport. However, if an address that was verified by such a
706 router was the subject of a callout, Exim crashed because it tried to
707 read the rcpt_include_affixes from the non-existent transport. Now it
708 just assumes that the setting of that option is false. This bug was
709 introduced by 4.51/PH/31.
711 PH/28 Changed -d+all to exclude +memory, because that information is very
712 rarely of interest, but it makes the output a lot bigger. People tend to
713 do -d+all out of habit.
715 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
716 code in os-type was giving problems when libc.so lives in lib64, like on
719 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
720 aren't the modern standard, and it seems that some systems' include files
721 don't always have them. Exim was already checking for some of the newer
722 ones like T_AAAA, and defining it itself. I've added checks for all the
723 record types that Exim uses.
725 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
726 not automatically generating a new one, as it is supposed to. This
727 prevented TLS from working. If the file did exist, but contained invalid
728 data, a new version was generated, as expected. It was only the case of a
729 non-existent file that was broken.
731 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
732 with a change in libdomainkeys > 0.64.
734 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
735 from DNS. If the selector record carries the flag, it now has
736 precedence over the domain-wide flag.
738 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
740 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
741 the use of a port name that isn't defined in /etc/services) occurred, the
742 message was deferred as in a normal delivery, and thus remained on the
743 spool, instead of being failed because of the mua_wrapper setting. This
744 is now fixed, and I tidied up some of the mua_wrapper messages at the
747 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
748 the messages in a hash of arrays rather than using individual hashes.
749 This is a bit cleaner and results in dramatic memory savings, albeit
750 at a slight CPU cost.
752 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
753 as requested by Marc Sherman.
755 SC/10 Eximstats: added histograms for user specified patterns as requested
758 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
760 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
761 fopen() in the content-scanning modules that did not already have it.
767 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
768 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
770 TK/02 Fix ACL "control" statement not being available in MIME ACL.
772 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
774 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
775 to test Sieve filters that use "vacation".
777 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
778 that changes the way the GnuTLS parameters are stored in the cache file.
779 The new format can be generated externally. For backward compatibility,
780 if the data in the cache doesn't make sense, Exim assumes it has read an
781 old-format file, and it generates new data and writes a new file. This
782 means that you can't go back to an older release without removing the
785 PH/03 A redirect router that has both "unseen" and "one_time" set does not
786 work if there are any delivery delays because "one_time" forces the
787 parent to be marked "delivered", so its unseen clone is never tried
788 again. For this reason, Exim now forbids the simultaneous setting of
791 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
792 redirected to themselves ("homonym" addresses). Read the long ChangeLog
793 entry if you want to know the details. The fix, however, neglected to
794 consider the case when local delivery batching is involved. The test for
795 "previously delivered" was not happening when checking to see if an
796 address could be batched with a previous (undelivered) one; under
797 certain circumstances this could lead to multiple deliveries to the same
800 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
801 in its include files, and this causes problems building Exim.
803 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
804 header_syntax) but Exim was just ignoring anything given after a slash.
805 In particular, this caused confusion with an attempt to use "verify =
806 reverse_host_lookup/defer_ok". An error is now given when options are
807 supplied for verify items that do not have them. (Maybe reverse_host_
808 lookup should have a defer_ok option, but that's a different point.)
810 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
811 defined by RFC 821) to 2048, because there were problems with some AUTH
812 commands, and RFC 1869 says the size should be increased for extended
813 SMTP commands that take arguments.
815 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
818 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
819 "unknown" error; now it says that the functionality isn't in the binary.
821 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
822 an address' error message when a string expansion fails (syntax or
823 whatever). Otherwise the password may appear in the log. Following change
824 PH/42 below, there is no longer a chance of it appearing in a bounce
827 PH/11 Installed exipick version 20050225.0 from John Jetmore.
829 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
830 of its addresses was ever tried. (Bugzilla bug #2.)
832 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
833 the result incorrectly in the debug output. (It correctly added a newline
834 to what was transported.)
836 TF/01 Added $received_time.
838 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
839 commented out examples of how to interface to a virus scanner and to
840 SpamAssassin. Also added commented examples of av_scanner and
841 spamd_address settings.
843 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
844 and controls are allowed in which ACLs. There were a couple of minor
845 errors. Some of the entries in the conditions table (which is a table of
846 where they are NOT allowed) were getting very unwieldy; rewrote them as a
847 negation of where the condition IS allowed.
849 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
851 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
852 header file does not have a version number, so I've had to invent a new
853 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
854 API. The code is untested by me (my Linux distribution still has 0.3.2 of
855 radiusclient), but it was contributed by a Radius user.
857 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
858 files or directories, for OpenSSL.
860 PH/19 When an Exim process that is running as root has to create an Exim log
861 file, it does so in a subprocess that runs as exim:exim so as to get the
862 ownership right at creation (otherwise, other Exim processes might see
863 the file with the wrong ownership). There was no test for failure of this
864 fork() call, which would lead to the process getting stuck as it waited
865 for a non-existent subprocess. Forks do occasionally fail when resources
866 run out. I reviewed all the other calls to fork(); they all seem to check
869 PH/20 When checking for unexpected SMTP input at connect time (before writing
870 the banner), Exim was not dealing correctly with a non-positive return
871 from the read() function. If the client had disconnected by this time,
872 the result was a log entry for a synchronization error with an empty
873 string after "input=" when read() returned zero. If read() returned -1
874 (an event I could not check), uninitialized data bytes were printed.
875 There were reports of junk text (parts of files, etc) appearing after
878 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
880 PH/22 Added support for macro redefinition, and (re)definition in between
881 driver and ACL definitions.
883 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
884 forgetting to use the resulting value; it was using the unexpanded value.
886 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
887 hadn't been configured. The fix is from Juergen Kreileder, who
888 understands it better than I do:
890 "Here's what I see happening with three configured cyrus_sasl
891 authenticators configured (plain, login, cram-md5):
893 On startup auth_cyrus_sasl_init() gets called for each of these.
894 This means three calls to sasl_listmech() without a specified mech_list.
895 => SASL tests which mechs of all available mechs actually work
896 => three warnings about OTP not working
897 => the returned list contains: plain, login, cram-md5, digest-md5, ...
899 With the patch, sasl_listmech() also gets called three times. But now
900 SASL's mech_list option is set to the server_mech specified in the the
901 authenticator. Or in other words, the answer from sasl_listmech()
902 gets limited to just the mech you're testing for (which is different
904 => the return list contains just 'plain' or 'login', 'cram-md5' or
905 nothing depending on the value of ob->server_mech.
907 I've just tested the patch: Authentication still works fine,
908 unavailable mechs specified in the exim configuration are still
909 caught, and the auth.log warnings about OTP are gone."
911 PH/25 When debugging is enabled, the contents of the command line are added
912 to the debugging output, even when log_selector=+arguments is not
915 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
916 answer is "GNU", and only if the return is "GNU/something" is the answer
919 PH/27 $acl_verify_message is now set immediately after the failure of a
920 verification in an ACL, and so is available in subsequent modifiers. In
921 particular, the message can be preserved by coding like this:
923 warn !verify = sender
924 set acl_m0 = $acl_verify_message
926 Previously, $acl_verify_message was set only while expanding "message"
927 and "log_message" when a very denied access.
929 PH/28 Modified OS/os.c-Linux with
931 -#ifndef OS_LOAD_AVERAGE
932 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
934 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
935 nomenclature these days.)
937 PH/29 Installed patch from the Sieve maintainer that adds the options
938 sieve_useraddress and sieve_subaddress to the redirect router.
940 PH/30 In these circumstances:
941 . Two addresses routed to the same list of hosts;
942 . First host does not offer TLS;
943 . First host accepts first address;
944 . First host gives temporary error to second address;
945 . Second host offers TLS and a TLS session is established;
946 . Second host accepts second address.
947 Exim incorrectly logged both deliveries with the TLS parameters (cipher
948 and peerdn, if requested) that were in fact used only for the second
951 PH/31 When doing a callout as part of verifying an address, Exim was not paying
952 attention to any local part prefix or suffix that was matched by the
953 router that accepted the address. It now behaves in the same way as it
954 does for delivery: the affixes are removed from the local part unless
955 rcpt_include_affixes is set on the transport.
957 PH/32 Add the sender address, as F=<...>, to the log line when logging a
958 timeout during the DATA phase of an incoming message.
960 PH/33 Sieve envelope tests were broken for match types other than :is. I have
961 applied a patch sanctioned by the Sieve maintainer.
963 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
964 the uid or gid is negative. A case of a negative gid caused this to be
965 noticed. The fix allows for either to be negative.
967 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
968 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
971 PH/36 The change PH/12 above was broken. Fixed it.
973 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
974 the grounds that routing the same address twice would always produce the
975 same answer. This might have been true once, but it is certainly no
976 longer true now. Routing a child address may depend on the previous
977 routing that produced that child. Some complicated redirection strategies
978 went wrong when messages had multiple recipients, and made Exim's
979 behaviour dependent on the order in which the addresses were given.
981 I have moved the duplicate checking until after the routing is complete.
982 Exim scans the addresses that are assigned to local and remote
983 transports, and removes any duplicates. This means that more work will be
984 done, as duplicates will always all be routed, but duplicates are
985 presumably rare, so I don't expect this is of any significance.
987 For deliveries to pipes, files, and autoreplies, the duplicate checking
988 still happens during the routing process, since they are not going to be
991 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
992 It corrects a timeout issue with spamd. This is Ian's comment: "The
993 background is that sometimes spamd either never reads data from a
994 connection it has accepted, or it never writes response data. The exiscan
995 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
996 blindly assumes that writes won't block so it may never time out."
998 PH/39 Allow G after quota size as well as K and M.
1000 PH/40 The value set for $authenticated_id in an authenticator may not contain
1001 binary zeroes or newlines because the value is written to log lines and
1002 to spool files. There was no check on this. Now the value is run through
1003 the string_printing() function so that such characters are converted to
1004 printable escape sequences.
1006 PH/41 $message_linecount is a new variable that contains the total number of
1007 lines in the message. Compare $body_linecount, which is the count for the
1010 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1011 bounce and delay warning messages, except in certain special cases, which
1014 (a) An SMTP error message from a remote host;
1015 (b) A message specified in a :fail: redirection;
1016 (c) A message specified in a "fail" command in a system filter;
1017 (d) A message specified in a FAIL return from the queryprogram router;
1018 (e) A message specified by the cannot_route_message router option.
1020 In these cases only, Exim does include the error details in bounce and
1021 warning messages. There are also a few cases where bland messages such
1022 as "unrouteable address" or "local delivery error" are given.
1024 PH/43 $value is now also set for the "else" part of a ${run expansion.
1026 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1027 being worked on, but at least Exim now implements the latest version to
1030 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1031 process to complete was treated as a delivery failure, a timeout while
1032 writing the message to the pipe was logged, but erroneously treated as a
1033 successful delivery. Such timeouts include transport filter timeouts. For
1034 consistency with the overall process timeout, these timeouts are now
1035 treated as errors, giving rise to delivery failures by default. However,
1036 there is now a new Boolean option for the pipe transport called
1037 timeout_defer, which, if set TRUE, converts the failures into defers for
1038 both kinds of timeout. A transport filter timeout is now identified in
1041 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1042 systems where "make" and "gmake" are different, calling "gmake" at top
1043 level broke things. I've arranged for the value of $(MAKE) to be passed
1044 from the Makefile to this script so that it can call the same version of
1048 A note about Exim versions 4.44 and 4.50
1049 ----------------------------------------
1051 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1052 changes of various kinds. As a consequence, a big documentation update was
1053 needed. This delayed the release for rather longer than seemed good, especially
1054 in the light of a couple of (minor) security issues. Therefore, the changes
1055 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1056 release. So 4.44 and 4.50 are in effect two different branches that both start
1059 I have left the 4.50 change log unchanged; it contains all the changes since
1060 4.43. The change log for 4.44 is below; many of its items are identical to
1061 those for 4.50. This seems to be the most sensible way to preserve the
1062 historical information.
1068 1. Minor wording change to the doc/README.SIEVE file.
1070 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1071 computation of the current number of files was incorrect.
1073 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1074 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1077 4. Give more explanation in the error message when the command for a transport
1078 filter fails to execute.
1080 5. There are several places where Exim runs a non-Exim command in a
1081 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1082 was being done only for the command run by the queryprogram router. It is
1083 now done for all such subprocesses. The other cases are: ${run, transport
1084 filters, and the commands run by the lmtp and pipe transports.
1086 6. Added CONFIGURE_GROUP build-time option.
1088 7. Some older OS have a limit of 256 on the maximum number of file
1089 descriptors. Exim was using setrlimit() to set 1000 as a large value
1090 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1091 systems. I've change it so that if it can't get 1000, it tries for 256.
1093 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1094 was an oversight, and furthermore, ever since the addition of extra
1095 controls (e.g. 4.43/32), the checks on when to allow different forms of
1096 "control" were broken. There should now be diagnostics for all cases when a
1097 control that does not make sense is encountered.
1099 9. Added the /retain_sender option to "control=submission".
1101 10. $recipients is now available in the predata ACL (oversight).
1103 11. Tidy the search cache before the fork to do a delivery from a message
1104 received from the command line. Otherwise the child will trigger a lookup
1105 failure and thereby defer the delivery if it tries to use (for example) a
1106 cached ldap connection that the parent has called unbind on.
1108 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1109 of $address_data from the recipient verification was clobbered by the
1110 sender verification.
1112 13. The value of address_data from a sender verification is now available in
1113 $sender_address_data in subsequent conditions in the ACL statement.
1115 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1117 15. Added a new option "connect=<time>" to callout options, to set a different
1120 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1121 was its contents. (It was OK if the option was not defined at all.)
1123 17. A "Completed" log line is now written for messages that are removed from
1124 the spool by the -Mrm option.
1126 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1127 information about exactly what failed.
1129 19. Added -dd to debug only the daemon process.
1131 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1132 handles timeouts, both on the server side and network timeouts. Renamed the
1133 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1135 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1136 It is now set to "smtps".
1138 22. $host_address is now set to the target address during the checking of
1139 ignore_target_hosts.
1141 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1142 being passed; this would have caused $sender_host_name to have been used if
1143 matching the list had actually called for a host name (not very likely,
1144 since this list is usually IP addresses). A host name is now passed as
1147 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1148 code that specifies a non-restarting handler (typically sigaction() in
1149 modern systems) in an attempt to fix a rare and obscure crash bug.
1151 25. Narrowed the window for a race in the daemon that could cause it to ignore
1152 SIGCHLD signals. This is not a major problem, because they are used only to
1153 wake it up if nothing else does.
1155 26. A malformed maildirsize file could cause Exim to calculate negative values
1156 for the mailbox size or file count. Odd effects could occur as a result.
1157 The maildirsize information is now recalculated if the size or filecount
1160 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1161 support for a long time. Removed HAVE_SYS_VFS_H.
1163 28. Installed the latest version of exipick from John Jetmore.
1165 29. In an address list, if the pattern was not a regular expression, an empty
1166 subject address (from a bounce message) matched only if the pattern was an
1167 empty string. Non-empty patterns were not even tested. This was the wrong
1168 because it is perfectly reasonable to use an empty address as part of a
1169 database query. An empty address is now tested by patterns that are
1170 lookups. However, all the other forms of pattern expect the subject to
1171 contain a local part and a domain, and therefore, for them, an empty
1172 address still always fails if the pattern is not itself empty.
1174 30. Exim went into a mad DNS loop when attempting to do a callout where the
1175 host was specified on an smtp transport, and looking it up yielded more
1176 than one IP address.
1178 31. Re-factored the code for checking spool and log partition space into a
1179 function that finds that data and another that does the check. The former
1180 is then used to implement four new variables: $spool_space, $log_space,
1181 $spool_inodes, and $log_inodes.
1183 32. The RFC2047 encoding function was originally intended for short strings
1184 such as real names; it was not keeping to the 75-character limit for
1185 encoded words that the RFC imposes. It now respects the limit, and
1186 generates multiple encoded words if necessary. To be on the safe side, I
1187 have increased the buffer size for the ${rfc2047: expansion operator from
1190 33. It is now permitted to omit both strings after an "if" condition; if the
1191 condition is true, the result is "true". As before, when the second string
1192 is omitted, a false condition yields an empty string. This makes it less
1193 cumbersome to write custom ACL and router conditions.
1195 34. Failure to deliver a bounce message always caused it to be frozen, even if
1196 there was an errors_to setting on the router. The errors_to setting is now
1199 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1200 canonical form (fully expanded) before being placed in
1201 $sender_host_address.
1203 36. The table in the code that translates DNS record types into text (T_A to
1204 "A" for instance) was missing entries for NS and CNAME. It is just possible
1205 that this could have caused confusion if both these types were looked up
1206 for the same domain, because the text type is used as part of Exim's
1207 per-process caching. But the chance of anyone hitting this buglet seems
1210 37. The dnsdb lookup has been extended in a number of ways.
1212 (1) There is a new type, "zns", which walks up the domain tree until it
1213 finds some nameserver records. It should be used with care.
1215 (2) There is a new type, "mxh", which is like "mx" except that it returns
1216 just the host names, not the priorities.
1218 (3) It is now possible to give a list of domains (or IP addresses) to be
1219 looked up. The behaviour when one of the lookups defers can be
1220 controlled by a keyword.
1222 (4) It is now possible to specify the separator character for use when
1223 multiple records are returned.
1225 38. The dnslists ACL condition has been extended: it is now possible to supply
1226 a list of IP addresses and/or domains to be looked up in a particular DNS
1229 39. Added log_selector=+queue_time_overall.
1231 40. When running the queue in the test harness, wait just a tad after forking a
1232 delivery process, to get repeatability of debugging output.
1234 41. Include certificate and key file names in error message when GnuTLS fails
1235 to set them up, because the GnuTLS error message doesn't include the name
1236 of the failing file when there is a problem reading it.
1238 42. Allow both -bf and -bF in the same test run.
1240 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1242 44. The "Exiscan patch" is now merged into the mainline Exim source.
1244 45. Sometimes the final signoff response after QUIT could fail to get
1245 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1246 < 0 before doing a fflush(). This bug looks as though it goes back to the
1247 introduction of TLS in release 3.20, but "sometimes" must have been rare
1248 because the tests only now provoked it.
1250 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1251 (this can affect the format of dates).
1253 47. exim_tidydb, when checking for the continued existence of a message for
1254 which it has found a message-specific retry record, was not finding
1255 messages that were in split spool directories. Consequently, it was
1256 deleting retry records that should have stayed in existence.
1258 48. Steve fixed some bugs in eximstats.
1260 49. The SPA authentication driver was not abandoning authentication and moving
1261 on to the next authenticator when an expansion was forced to fail,
1262 contradicting the general specification for all authenticators. Instead it
1263 was generating a temporary error. It now behaves as specified.
1265 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1266 (the order specifies the preference for clients). The order is now AES256,
1267 AES128, 3DES, ARCFOUR128.
1269 51. Small patch to Sieve code - explicitly set From: when generating an
1272 52. Exim crashed if a remote delivery caused a very long error message to be
1273 recorded - for instance if somebody sent an entire SpamAssassin report back
1274 as a large number of 550 error lines. This bug was coincidentally fixed by
1275 increasing the size of one of Exim's internal buffers (big_buffer) that
1276 happened as part of the Exiscan merge. However, to be on the safe side, I
1277 have made the code more robust (and fixed the comments that describe what
1280 53. Now that there can be additional text after "Completed" in log lines (if
1281 the queue_time_overall log selector is set), a one-byte patch to exigrep
1282 was needed to allow it to recognize "Completed" as not the last thing in
1285 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1286 patch that reportedly fixes this has been added. I am not expert enough to
1287 create a test for it. This is what the patch creator wrote:
1289 "I found a little strange behaviour of ldap code when working with
1290 Windows 2003 AD Domain, where users was placed in more than one
1291 Organization Units. When I tried to give exim partial DN, the exit code
1292 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1293 But simultaneously result of request was absolutely normal ldap result,
1294 so I produce this patch..."
1296 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1297 REFERENCE, so I have modified the code to exclude the patch when that macro
1300 55. Some experimental protocols are using DNS PTR records for new purposes. The
1301 keys for these records are domain names, not reversed IP addresses. The
1302 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1303 leaves it alone. Component reversal etc. now happens only for IP addresses.
1306 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1308 57. Double the size of the debug message buffer (to 2048) so that more of very
1309 long debug lines gets shown.
1311 58. The exicyclog utility now does better if the number of log files to keep
1312 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1314 59. Two changes related to the smtp_active_hostname option:
1316 (1) $smtp_active_hostname is now available as a variable.
1317 (2) The default for smtp_banner uses $smtp_active_hostname instead
1318 of $primary_hostname.
1320 60. The host_aton() function is supposed to be passed a string that is known
1321 to be a valid IP address. However, in the case of IPv6 addresses, it was
1322 not checking this. This is a hostage to fortune. Exim now panics and dies
1323 if the condition is not met. A case was found where this could be provoked
1324 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1325 components; fortuitously, this particular loophole had already been fixed
1326 by change 4.50/55 above.
1328 If there are any other similar loopholes, the new check in host_aton()
1329 itself should stop them being exploited. The report I received stated that
1330 data on the command line could provoke the exploit when Exim was running as
1331 exim, but did not say which command line option was involved. All I could
1332 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1333 running as the user.
1336 61. There was a buffer overflow vulnerability in the SPA authentication code
1337 (which came originally from the Samba project). I have added a test to the
1338 spa_base64_to_bits() function which I hope fixes it.
1341 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1342 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1344 63. The daemon start-up calls getloadavg() while still root for those OS that
1345 need the first call to be done as root, but it missed one case: when
1346 deliver_queue_load_max is set with deliver_drop_privilege. This is
1347 necessary for the benefit of the queue runner, because there is no re-exec
1348 when deliver_drop_privilege is set.
1350 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1351 This has been fixed.
1353 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1354 was in use, was not putting the data itself into the right store pool;
1355 consequently, it could be overwritten for a subsequent message in the same
1356 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1359 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1361 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1362 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1363 the function were treating the return as a boolean value, which happened to
1364 work because 0=false and not-0=true, but is not correct code.
1366 68. The host_aton() function was not handling scoped IPv6 addresses (those
1367 with, for example, "%eth0" on the end) correctly.
1369 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1370 negated items (that is, ~something) in unsigned ints. Some compilers
1371 apparently mutter when there is no cast.
1373 70. If an address verification called from an ACL failed, and did not produce a
1374 user-specific message (i.e. there was only a "system" message), nothing was
1375 put in $acl_verify_message. In this situation, it now puts the system
1378 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1379 session; change 4.31/43 added the unwanted input to the log line - except
1380 that it did not do this in the start of session case. It now does.
1382 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1383 This is wrong and can cause the other end to generate a synchronization
1384 error if it is another Exim or anything else that does the synchronization
1385 check. A QUIT command is no longer sent after a timeout.
1387 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1388 during host lookups.
1390 74. The defer_ok option of callout verification was not working if it was used
1391 when verifying addresses in header lines, that is, for this case:
1393 verify = header_sender/callout=defer_ok
1395 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1396 those file descriptors could be used for SMTP connections. If anything
1397 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1398 could be sent to the SMTP client, causing chaos. The daemon now opens
1399 stdin, stdout, and stderr to /dev/null when it puts itself into the
1402 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1403 log by default. The user can override this with suitable Perl magic.
1405 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1406 the log message when discard triggers, was not working for the DATA ACL or
1407 for the non-SMTP ACL.
1409 78. Error message wording change in sieve.c.
1411 79. If smtp_accept_max_per_host was set, the number of connections could be
1412 restricted to fewer than expected, because the daemon was trying to set up
1413 a new connection before checking whether the processes handling previous
1414 connections had finished. The check for completed processes is now done
1415 earlier. On busy systems, this bug wouldn't be noticed because something
1416 else would have woken the daemon, and it would have reaped the completed
1419 80. If a message was submitted locally by a user whose login name contained one
1420 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1421 It caused a spool format error. I have fixed the spool reading code. A
1422 related problem was that the "from" clause in the Received: line became
1423 illegal because of the space(s). It is now covered by ${quote_local_part.
1425 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1428 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1430 83. Patch from Sieve maintainer to fix unterminated string problem in
1431 "vacation" handling.
1433 84. Some minor changes to the Linux configuration files to help with other
1434 OS variants using glibc.
1436 85. One more patch for Sieve to update vacation handling to latest spec.
1439 ----------------------------------------------------
1440 See the note above about the 4.44 and 4.50 releases.
1441 ----------------------------------------------------
1447 1. Change 4.43/35 introduced a bug that caused file counts to be
1448 incorrectly computed when quota_filecount was set in an appendfile
1451 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1452 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1455 3. Give more explanation in the error message when the command for a transport
1456 filter fails to execute.
1458 4. There are several places where Exim runs a non-Exim command in a
1459 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1460 was being done only for the command run by the queryprogram router. It is
1461 now done for all such subprocesses. The other cases are: ${run, transport
1462 filters, and the commands run by the lmtp and pipe transports.
1464 5. Some older OS have a limit of 256 on the maximum number of file
1465 descriptors. Exim was using setrlimit() to set 1000 as a large value
1466 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1467 systems. I've change it so that if it can't get 1000, it tries for 256.
1469 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1470 was an oversight, and furthermore, ever since the addition of extra
1471 controls (e.g. 4.43/32), the checks on when to allow different forms of
1472 "control" were broken. There should now be diagnostics for all cases when a
1473 control that does not make sense is encountered.
1475 7. $recipients is now available in the predata ACL (oversight).
1477 8. Tidy the search cache before the fork to do a delivery from a message
1478 received from the command line. Otherwise the child will trigger a lookup
1479 failure and thereby defer the delivery if it tries to use (for example) a
1480 cached ldap connection that the parent has called unbind on.
1482 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1483 of $address_data from the recipient verification was clobbered by the
1484 sender verification.
1486 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1487 was its contents. (It was OK if the option was not defined at all.)
1489 11. A "Completed" log line is now written for messages that are removed from
1490 the spool by the -Mrm option.
1492 12. $host_address is now set to the target address during the checking of
1493 ignore_target_hosts.
1495 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1496 being passed; this would have caused $sender_host_name to have been used if
1497 matching the list had actually called for a host name (not very likely,
1498 since this list is usually IP addresses). A host name is now passed as
1501 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1502 code that specifies a non-restarting handler (typically sigaction() in
1503 modern systems) in an attempt to fix a rare and obscure crash bug.
1505 15. Narrowed the window for a race in the daemon that could cause it to ignore
1506 SIGCHLD signals. This is not a major problem, because they are used only to
1507 wake it up if nothing else does.
1509 16. A malformed maildirsize file could cause Exim to calculate negative values
1510 for the mailbox size or file count. Odd effects could occur as a result.
1511 The maildirsize information is now recalculated if the size or filecount
1514 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1515 support for a long time. Removed HAVE_SYS_VFS_H.
1517 18. Updated exipick to current release from John Jetmore.
1519 19. Allow an empty sender to be matched against a lookup in an address list.
1520 Previously the only cases considered were a regular expression, or an
1523 20. Exim went into a mad DNS lookup loop when doing a callout where the
1524 host was specified on the transport, if the DNS lookup yielded more than
1527 21. The RFC2047 encoding function was originally intended for short strings
1528 such as real names; it was not keeping to the 75-character limit for
1529 encoded words that the RFC imposes. It now respects the limit, and
1530 generates multiple encoded words if necessary. To be on the safe side, I
1531 have increased the buffer size for the ${rfc2047: expansion operator from
1534 22. Failure to deliver a bounce message always caused it to be frozen, even if
1535 there was an errors_to setting on the router. The errors_to setting is now
1538 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1539 canonical form (fully expanded) before being placed in
1540 $sender_host_address.
1542 24. Updated eximstats to version 1.33
1544 25. Include certificate and key file names in error message when GnuTLS fails
1545 to set them up, because the GnuTLS error message doesn't include the name
1546 of the failing file when there is a problem reading it.
1548 26. Expand error message when OpenSSL has problems setting up cert/key files.
1551 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1552 (this can affect the format of dates).
1554 28. exim_tidydb, when checking for the continued existence of a message for
1555 which it has found a message-specific retry record, was not finding
1556 messages that were in split spool directories. Consequently, it was
1557 deleting retry records that should have stayed in existence.
1559 29. eximstats updated to version 1.35
1560 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1561 1.35 - bugfix such that pie charts by volume are generated correctly
1563 30. The SPA authentication driver was not abandoning authentication and moving
1564 on to the next authenticator when an expansion was forced to fail,
1565 contradicting the general specification for all authenticators. Instead it
1566 was generating a temporary error. It now behaves as specified.
1568 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1569 (the order specifies the preference for clients). The order is now AES256,
1570 AES128, 3DES, ARCFOUR128.
1572 31. Small patch to Sieve code - explicitly set From: when generating an
1575 32. Exim crashed if a remote delivery caused a very long error message to be
1576 recorded - for instance if somebody sent an entire SpamAssassin report back
1577 as a large number of 550 error lines. This bug was coincidentally fixed by
1578 increasing the size of one of Exim's internal buffers (big_buffer) that
1579 happened as part of the Exiscan merge. However, to be on the safe side, I
1580 have made the code more robust (and fixed the comments that describe what
1583 33. Some experimental protocols are using DNS PTR records for new purposes. The
1584 keys for these records are domain names, not reversed IP addresses. The
1585 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1586 leaves it alone. Component reversal etc. now happens only for IP addresses.
1589 34. The host_aton() function is supposed to be passed a string that is known
1590 to be a valid IP address. However, in the case of IPv6 addresses, it was
1591 not checking this. This is a hostage to fortune. Exim now panics and dies
1592 if the condition is not met. A case was found where this could be provoked
1593 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1594 components; fortuitously, this particular loophole had already been fixed
1595 by change 4.50/55 or 4.44/33 above.
1597 If there are any other similar loopholes, the new check in host_aton()
1598 itself should stop them being exploited. The report I received stated that
1599 data on the command line could provoke the exploit when Exim was running as
1600 exim, but did not say which command line option was involved. All I could
1601 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1602 running as the user.
1605 35. There was a buffer overflow vulnerability in the SPA authentication code
1606 (which came originally from the Samba project). I have added a test to the
1607 spa_base64_to_bits() function which I hope fixes it.
1610 36. The daemon start-up calls getloadavg() while still root for those OS that
1611 need the first call to be done as root, but it missed one case: when
1612 deliver_queue_load_max is set with deliver_drop_privilege. This is
1613 necessary for the benefit of the queue runner, because there is no re-exec
1614 when deliver_drop_privilege is set.
1616 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1617 was in use, was not putting the data itself into the right store pool;
1618 consequently, it could be overwritten for a subsequent message in the same
1619 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1622 38. Sometimes the final signoff response after QUIT could fail to get
1623 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1624 < 0 before doing a fflush(). This bug looks as though it goes back to the
1625 introduction of TLS in release 3.20, but "sometimes" must have been rare
1626 because the tests only now provoked it.
1632 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1633 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1634 Now it returns an integer. A number of calls were still expecting a T/F
1635 return. Fortuitously, in all cases, the tests worked in OK situations,
1636 which is the norm. However, things would have gone wrong on any write
1637 failures on the smtp file descriptor. This function is used when sending
1638 messages over SMTP and also when doing verify callouts.
1640 2. When Exim is called to do synchronous delivery of a locally submitted
1641 message (the -odf or -odi options), it no longer closes stderr before doing
1644 3. Implemented the mua_wrapper option.
1646 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1648 5. Implemented the functions header_remove(), header_testname(),
1649 header_add_at_position(), and receive_remove_recipient(), and exported them
1652 6. If an ACL "warn" statement specified the addition of headers, Exim already
1653 inserted X-ACL-Warn: at the start if there was no header name. However, it
1654 was not making this test for the second and subsequent header lines if
1655 there were newlines in the string. This meant that an invalid header could
1656 be inserted if Exim was badly configured.
1658 7. Allow an ACL "warn" statement to add header lines at the start or after all
1659 the Received: headers, as well as at the end.
1661 8. Added the rcpt_4xx retry error code.
1663 9. Added postmaster_mailfrom=xxx to callout verification option.
1665 10. Added mailfrom=xxxx to the callout verification option, for verify=
1668 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1669 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1671 12. Inserted some casts to stop certain compilers warning when using pointer
1672 differences as field lengths or precisions in printf-type calls (mostly
1673 affecting debugging statements).
1675 13. Added optional readline() support for -be (dynamically loaded).
1677 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1678 same clock tick as a message's arrival, so that its received time was the
1679 same as the "first fail" time on the retry record, and that message
1680 remained on the queue past the ultimate address timeout, every queue runner
1681 would try a delivery (because it was past the ultimate address timeout) but
1682 after another failure, the ultimate address timeout, which should have then
1683 bounced the address, did not kick in. This was a "< instead of <=" error;
1684 in most cases the first failure would have been in the next clock tick
1685 after the received time, and all would be well.
1687 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1688 being recognized when the domain list was tested by the match_domain
1689 condition in an expansion string.
1691 16. Added the ${str2b64: operator.
1693 17. Exim was always calling setrlimit() to set a large limit for the number of
1694 processes, without checking whether the existing limit was already
1695 adequate. (It did check for the limit on file descriptors.) Furthermore,
1696 errors from getrlimit() and setrlimit() were being ignored. Now they are
1697 logged to the main and panic logs, but Exim does carry on, to try to do its
1698 job under whatever limits there are.
1700 18. Imported PCRE 5.0.
1702 19. Trivial typo in log message " temporarily refused connection" (the leading
1705 20. If the log selector return_path_on_delivery was set and an address was
1706 redirected to /dev/null, the delivery process crashed because it assumed
1707 that a return path would always be set for a "successful" delivery. In this
1708 case, the whole delivery is bypassed as an optimization, and therefore no
1711 21. Internal re-arrangement: the function for sending a challenge and reading
1712 a response while authentication was assuming a zero-terminated challenge
1713 string. It's now changed to take a pointer and a length, to allow for
1714 binary data in such strings.
1716 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1718 23. Exim was not respecting finduser_retries when seeking the login of the
1719 uid under which it was called; it was always trying 10 times. (The default
1720 setting of finduser_retries is zero.) Also, it was sleeping after the final
1721 failure, which is pointless.
1723 24. Implemented tls_on_connect_ports.
1725 25. Implemented acl_smtp_predata.
1727 26. If the domain in control=submission is set empty, Exim assumes that the
1728 authenticated id is a complete email address when it generates From: or
1729 Sender: header lines.
1731 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1732 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1733 chown and chgrp in /bin and hostname in /usr/bin.
1735 28. Exim was keeping the "process log" file open after each use, just as it
1736 does for the main log. This opens the possibility of it remaining open for
1737 long periods when the USR1 signal hits a daemon. Occasional processlog
1738 errors were reported, that could have been caused by this. Anyway, it seems
1739 much more sensible not to leave this file open at all, so that is what now
1742 29. The long-running daemon process does not normally write to the log once it
1743 has entered its main loop, and it closes the log before doing so. This is
1744 so that log files can straightforwardly be renamed and moved. However,
1745 there are a couple of unusual error situations where the daemon does write
1746 log entries, and I had neglected to close the log afterwards.
1748 30. The text of an SMTP error response that was received during a remote
1749 delivery was being truncated at 512 bytes. This is too short for some of
1750 the long messages that one sometimes sees. I've increased the limit to
1753 31. It is now possible to make retry rules that apply only when a message has a
1754 specific sender, in particular, an empty sender.
1756 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1757 it possible to be selective about when SMTP synchronization is enforced.
1759 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1761 32. Implemented hosts_connection_nolog.
1763 33. Added an ACL for QUIT.
1765 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1768 35. Added mailbox_size and mailbox_filecount to appendfile.
1770 36. Added control = no_multiline_responses to ACLs.
1772 37. There was a bug in the logic of the code that waits for the clock to tick
1773 in the case where the clock went backwards by a substantial amount such
1774 that the microsecond fraction of "now" was more than the microsecond
1775 fraction of "then" (but the whole seconds number was less).
1777 38. Added support for the libradius Radius client library this is found on
1778 FreeBSD (previously only the radiusclient library was supported).
1784 1. When certain lookups returned multiple values in the form name=value, the
1785 quoting of the values was not always being done properly. Specifically:
1786 (a) If the value started with a double quote, but contained no whitespace,
1788 (b) If the value contained whitespace other than a space character (i.e.
1789 tabs or newlines or carriage returns) it was not quoted.
1790 This fix has been applied to the mysql and pgsql lookups by writing a
1791 separate quoting function and calling it from the lookup code. The fix
1792 should probably also be applied to nisplus, ibase and oracle lookups, but
1793 since I cannot test any of those, I have not disturbed their existing code.
1795 2. A hit in the callout cache for a specific address caused a log line with no
1796 reason for rejecting RCPT. Now it says "Previous (cached) callout
1797 verification failure".
1799 3. There was an off-by-one bug in the queryprogram router. An over-long
1800 return line was truncated at 256 instead of 255 characters, thereby
1801 overflowing its buffer with the terminating zero. As well as fixing this, I
1802 have increased the buffer size to 1024 (and made a note to document this).
1804 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1805 when Exim is waiting for an SMTP response from a remote server, Exim
1806 restarts its select() call on the socket, thereby resetting its timeout.
1807 This is not a problem when such interrupts are rare. Somebody set up a cron
1808 job to run exiwhat every 2 minutes, which is less than the normal select()
1809 timeout (5 or 10 minutes). This meant that the select() timeout never
1810 kicked in because it was always reset. I have fixed this by comparing the
1811 time when an interrupt arrives with the time at the start of the first call
1812 to select(). If more time than the timeout has elapsed, the interrupt is
1813 treated as a timeout.
1815 5. Some internal re-factoring in preparation for the addition of Sieve
1816 extensions (by MH). In particular, the "personal" test is moved to a
1817 separate function, and given an option for scanning Cc: and Bcc: (which is
1818 not set for Exim filters).
1820 6. When Exim created an email address using the login of the caller as the
1821 local part (e.g. when creating a From: or Sender: header line), it was not
1822 quoting the local part when it contained special characters such as @.
1824 7. Installed new OpenBSD configuration files.
1826 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1827 try to make them clearer.
1829 9. Callout options, other than the timeout value, were being ignored when
1830 verifying sender addresses in header lines. For example, when using
1832 verify = header_sender/callout=no_cache
1834 the cache was (incorrectly) being used.
1836 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1837 only the Cygwin environment.
1839 11. When return_path_on_delivery was set as a log selector, if different remote
1840 addresses in the same message used different return paths and parallel
1841 remote delivery occurred, the wrong values would sometimes be logged.
1842 (Whenever a remote delivery process finished, the return path value from
1843 the most recently started remote delivery process was logged.)
1845 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1846 lines when AUTH and/or TLS are in use. This is the "received protocol"
1847 field. Exim used to use "asmtp" for authenticated SMTP, without any
1848 indication (in the protocol name) for TLS use. Now it follows the RFC and
1849 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1850 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1851 names appear in log lines as well as in Received: header lines.
1853 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1854 extensions, and comparison tests, and to fix some bugs.
1856 14. Changes to the "personal" filter test:
1858 (1) The test was buggy in that it was just doing the equivalent of
1859 "contains" tests on header lines. For example, if a user's address was
1860 anne@some.where, the "personal" test would incorrectly be true for
1862 To: susanne@some.where
1864 This test is now done by extracting each address from the header in turn,
1865 and checking the entire address. Other tests that are part of "personal"
1866 are now done using regular expressions (for example, to check local parts
1867 of addresses in From: header lines).
1869 (2) The list of non-personal local parts in From: addresses has been
1870 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1871 taken from the Sieve specification recommendations.
1873 (3) If the message contains any header line starting with "List-" it is
1874 treated as non-personal.
1876 (4) The test for "circular" in the Subject: header line has been removed
1877 because it now seems ill-conceived.
1879 15. Minor typos in src/EDITME comments corrected.
1881 16. Installed latest exipick from John Jetmore.
1883 17. If headers_add on a router specified a text string that was too long for
1884 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1885 of string_sprintf() is now avoided.
1887 18. $message_body_size was not set (it was always zero) when running the DATA
1888 ACL and the local_scan() function.
1890 19. For the "mail" command in an Exim filter, no default was being set for
1891 the once_repeat time, causing a random time value to be used if "once" was
1892 specified. (If the value happened to be <= 0, no repeat happened.) The
1893 default is now 0s, meaning "never repeat". The "vacation" command was OK
1894 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1895 (I found it when inspecting the code).
1897 20. There is now an overall timeout for performing a callout verification. It
1898 defaults to 4 times the callout timeout, which applies to individual SMTP
1899 commands during the callout. The overall timeout applies when there is more
1900 than one host that can be tried. The timeout is checked before trying the
1901 next host. This prevents very long delays if there are a large number of
1902 hosts and all are timing out (e.g. when the network connections are timing
1903 out). The value of the overall timeout can be changed by specifying an
1904 additional sub-option for "callout", called "maxwait". For example:
1906 verify = sender/callout=5s,maxwait=20s
1908 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1909 to the end before writing, but this should make it even safer).
1911 22. Exim was forgetting that it had advertised PIPELINING for the second and
1912 subsequent messages on an SMTP connection. It was also not resetting its
1913 memory on STARTTLS and an internal HELO.
1915 23. When Exim logs an SMTP synchronization error within a session, it now
1916 records whether PIPELINING has been advertised or not.
1918 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1919 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1920 rather than long int.
1922 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1924 26. Added the never_mail option to autoreply.
1930 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1931 crash if the getsockname() call failed; this can happen if a connection is
1932 closed very soon after it is established. The problem was simply in the
1933 order in which certain operations were done, causing Exim to try to write
1934 to the SMTP stream before it had set up the file descriptor. The bug has
1935 been fixed by making things happen in the correct order.
1941 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1942 before the connection was closed, thus losing the rejection response.
1944 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1945 some early Solaris releases, but causes trouble in current releases where
1946 socklen_t is defined.
1948 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1951 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1954 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1955 that releasing the top of it at the end releases what was used for sub-
1956 expansions (unless the block got too big). However, discard this block if
1957 the first thing is a variable or header, so that we can use its block when
1958 it is dynamic (useful for very large $message_headers, for example).
1960 6. Lookups now cache *every* query, not just the most recent. A new, separate
1961 store pool is used for this. It can be recovered when all lookup caches are
1962 flushed. Lookups now release memory at the end of their result strings.
1963 This has involved some general refactoring of the lookup sources.
1965 7. Some code has been added to the store_xxx() functions to reduce the amount
1966 of flapping under certain conditions.
1968 8. log_incoming_interface used to affect only the <= reception log lines. Now
1969 it causes the local interface and port to be added to several more SMTP log
1970 lines, for example "SMTP connection from", and rejection lines.
1972 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1974 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1976 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1977 could be overwritten at the end of the current message (or the start of a
1978 new message if it was set in a HELO ACL). The value is now preserved for
1979 the duration of the SMTP connection.
1981 12. If a transport had a headers_rewrite setting, and a matching header line
1982 contained an unqualified address, that address was qualified, even if it
1983 did not match any rewriting rules. The underlying bug was that the values
1984 of the flags that permit the existence of unqualified sender and recipient
1985 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1986 non-local messages, and by -bnq for local messages) were not being
1987 preserved with the message after it was received.
1989 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1990 "next input=" as part of the text comprising the host identity instead of
1991 the correct text. The code was using the same buffer for two different
1992 strings. However, depending on which order the printing function evaluated
1993 its arguments, the bug did not always show up. Under Linux, for example, my
1994 test suite worked just fine.
1996 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1997 doesn't work with some older versions of Perl. It has been changed to "my",
1998 which in any case is probably the better facility to use.
2000 15. A really picky compiler found some instances of statements for creating
2001 error messages that either had too many or two few arguments for the format
2004 16. The size of the buffer for calls to the DNS resolver has been increased
2005 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2006 for addresses that have a lot of PTR records. This alleviates a problem; it
2007 does not fully solve it.
2009 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2010 buffer now truncates the list and logs the incident, which is the same
2011 action as happens when Exim is looking up a host name and its aliases.
2012 Previously in this situation something unpredictable would happen;
2013 sometimes it was "internal error: store_reset failed".
2015 18. If a server dropped the connection unexpectedly when an Exim client was
2016 using GnuTLS and trying to read a response, the client delivery process
2017 crashed while trying to generate an error log message.
2019 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2020 string, for example:
2022 warn message = H1: something\nH2: something
2024 the text was added as a single header line from Exim's point of view
2025 though it ended up OK in the delivered message. However, searching for the
2026 second and subsequent header lines using $h_h2: did not work. This has been
2027 fixed. Similarly, if a system filter added multiple headers in this way,
2028 the routers could not see them.
2030 20. Expanded the error message when iplsearch is called with an invalid key to
2031 suggest using net-iplsearch in a host list.
2033 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2034 ACLs are no longer actually imposed (and a message to that effect is
2037 22. If a "gecos" field in a passwd entry contained escaped characters, in
2038 particular, if it contained a \" sequence, Exim got it wrong when building
2039 a From: or a Sender: header from that name. A second bug also caused
2040 incorrect handling when an unquoted " was present following a character
2041 that needed quoting.
2043 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2044 was not being matched caselessly.
2046 24. Arranged for all hyphens in the exim.8 source to be escaped with
2049 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2050 properly. Recipient callout cache records were still being keyed to include
2051 the sender, even when use_sender was set false. This led to far more
2052 callouts that were necessary. The sender is no longer included in the key
2053 when use_sender is false.
2055 26. Added "control = submission" modifier to ACLs.
2057 27. Added the ${base62d: operator to decode base 62 numbers.
2059 28. dnsdb lookups can now access SRV records.
2061 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2062 the configuration file.
2064 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2065 (-v) mode. This makes the output for a verbose queue run more intelligible.
2067 31. Added a use_postmaster feature to recipient callouts.
2069 32. Added the $body_zerocount variable, containing the number of binary zero
2070 bytes in the message body.
2072 33. The time of last modification of the "new" subdirectory is now used as the
2073 "mailbox time last read" when there is a quota error for a maildir
2076 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2078 35. Added +ignore_unknown as a special item in host lists.
2080 36. Code for decoding IPv6 addresses in host lists is now included, even if
2081 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2082 address was recognized as an IP address, but was then not correctly decoded
2083 into binary, causing unexpected and incorrect effects when compared with
2090 1. Very minor rewording of debugging text in manualroute to say "list of
2091 hosts" instead of "hostlist".
2093 2. If verify=header_syntax was set, and a header line with an unqualified
2094 address (no domain) and a large number of spaces between the end of the
2095 name and the colon was received, the reception process suffered a buffer
2096 overflow, and (when I tested it) crashed. This was caused by some obsolete
2097 code that should have been removed. The fix is to remove it!
2099 3. When running in the test harness, delay a bit after writing a bounce
2100 message to get a bit more predictability in the log output.
2102 4. Added a call to search_tidyup() just before forking a reception process. In
2103 theory, someone could use a lookup in the expansion of smtp_accept_max_
2104 per_host which, without the tidyup, could leave open a database connection.
2106 5. Added the variables $recipient_data and $sender_data which get set from a
2107 lookup success in an ACL "recipients" or "senders" condition, or a router
2108 "senders" option, similar to $domain_data and $local_part_data.
2110 6. Moved the writing of debug_print from before to after the "senders" test
2113 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2114 problems for message scanning, either using a data ACL, or using
2115 local_scan() because the Received: header was not generated till after they
2116 were called (in order to set the time as the time of reception completion).
2117 I have revised the way this works. The header is now generated after the
2118 body is received, but before the ACL or local_scan() are called. After they
2119 are run, the timestamp in the header is updated.
2125 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2126 before starting a queue runner without re-exec. This happened only when
2127 deliver_drop_privilege was set or when the Exim user was set to root. The
2128 effect of the bug was that timeouts during subsequent deliveries caused
2129 crashes instead of being properly handled. The handler is now left at its
2130 default (and expected) setting.
2132 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2133 message, again when deliver_drop_privilege is set or Exim is run as root.
2134 The bug described in (1) was not present in this case, but the tidying up
2135 of the other signals was missing. I have made the two cases consistent.
2137 3. The ignore_target_hosts setting on a manualroute router was being ignored
2138 for hosts that were looked up using the /MX notation.
2140 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2143 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2144 operated on the sender address. After changing the $sender_address to <>
2145 for the sender address verify, Exim was re-instated it as the original
2146 (before rewriting) address, but remembering that it had rewritten it, so it
2147 wasn't rewriting it again. This bug also had the effect of breaking the
2148 sender address verification caching when the sender address was rewritten.
2150 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2151 This has been changed so that if the ip literal address matches
2152 ignore_target_hosts, the router declines.
2154 7. Added expansion conditions match_domain, match_address, and match_local_
2155 part (NOT match_host).
2157 8. The placeholder for the Received: header didn't have a length field set.
2159 9. Added code to Exim itself and to exim_lock to test for a specific race
2160 condition that could lead to file corruption when using MBX delivery. The
2161 issue is with the lockfile that is created in /tmp. If this file is removed
2162 after a process has opened it but before that process has acquired a lock,
2163 there is the potential for a second process to recreate the file and also
2164 acquire a lock. This could lead to two Exim processes writing to the file
2165 at the same time. The added code performs the same test as UW imapd; it
2166 checks after acquiring the lock that its file descriptor still refers to
2167 the same named file.
2169 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2170 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2173 11. Added the smtp_active_hostname option. If used, this will typically be made
2174 to depend on the incoming interface address. Because $interface_address is
2175 not set up until the daemon has forked a reception process, error responses
2176 that can happen earlier (such as "too many connections") no longer contain
2179 12. If an expansion in a condition on a "warn" statement fails because a lookup
2180 defers, the "warn" statement is abandoned, and the next ACL statement is
2181 processed. Previously this caused the whole ACL to be aborted.
2183 13. Added the iplsearch lookup type.
2185 14. Added ident_timeout as a log selector.
2187 15. Added tls_certificate_verified as a log selector.
2189 16. Added a global option tls_require_ciphers (compare the smtp transport
2190 option of the same name). This controls incoming TLS connections.
2192 17. I finally figured out how to make tls_require_ciphers do a similar thing
2193 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2194 before starting the TLS session.
2196 18. Tabs are now shown as \t in -bP output.
2198 19. If the log selector return_path_on_delivery was set, Exim crashed when
2199 bouncing a message because it had too many Received: header lines.
2201 20. If two routers both had headers_remove settings, and the first one included
2202 a superfluous trailing colon, the final name in the first list and the
2203 first name in the second list were incorrectly joined into one item (with a
2204 colon in the middle).
2210 1. Added -C and -D options to the exinext utility, mainly to make it easier
2211 to include in the automated testing, but these could be helpful when
2212 multiple configurations are in use.
2214 2. The exinext utility was not formatting the output nicely when there was
2215 an alternate port involved in the retry record key, nor when there was a
2216 message id as well (for retries that were specific to a specific message
2217 and a specific host). It was also confused by IPv6 addresses, because of
2218 the additional colons they contain. I have fixed the IPv4 problem, and
2219 patched it up to do a reasonable job for IPv6.
2221 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2222 delivery, the log line now contains "pipelined" if PIPELINING was used.
2224 4. An SMTP transport process used to panic and die if the bind() call to set
2225 an explicit outgoing interface failed. This has been changed; it is now
2226 treated in the same way as a connect() failure.
2228 5. A reference to $sender_host_name in the part of a conditional expansion
2229 that was being skipped was still causing a DNS lookup. This no longer
2232 6. The def: expansion condition was not recognizing references to header lines
2233 that used bh_ and bheader_.
2235 7. Added the _cache feature to named lists.
2237 8. The code for checking quota_filecount in the appendfile transport was
2238 allowing one more file than it should have been.
2240 9. For compatibility with Sendmail, the command line option
2248 and sets the incoming protocol and host name (for trusted callers). The
2249 host name and its colon can be omitted when only the protocol is to be set.
2250 Note the Exim already has two private options, -pd and -ps, that refer to
2251 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2252 "s", but I don't think that's a major issue.
2254 10. A number of refactoring changes to the code, none of which should affect
2257 (a) The number of logging options was getting close to filling up the
2258 32-bit word that was used as a bit map. I have split them into two classes:
2259 those that are passed in the argument to log_write(), and those that are
2260 only ever tested independently outside of that function. These are now in
2261 separate 32-bit words, so there is plenty of room for expansion again.
2262 There is no change in the user interface or the logging behaviour.
2264 (b) When building, for example, log lines, the code previously used a
2265 macro that called string_cat() twice, in order to add two strings. This is
2266 not really sufficiently general. Furthermore, there was one instance where
2267 it was actually wrong because one of the argument was used twice, and in
2268 one call a function was used. (As it happened, calling the function twice
2269 did not affect the overall behaviour.) The macro has been replaced by a
2270 function that can join an arbitrary number of extra strings onto a growing
2273 (c) The code for expansion conditions now uses a table and a binary chop
2274 instead of a serial search (which was left over from when there were very
2275 few conditions). Also, it now recognizes conditions like "pam" even when
2276 the relevant support is not compiled in: a suitably worded error message is
2277 given if an attempt is made to use such a condition.
2279 11. Added ${time_interval:xxxxx}.
2281 12. A bug was causing one of the ddress fields not to be passed back correctly
2282 from remote delivery subprocesses. The field in question was not being
2283 subsequently used, so this caused to problems in practice.
2285 13. Added new log selectors queue_time and deliver_time.
2287 14. Might have fixed a bug in maildirsizefile handling that threw up
2288 "unexpected character" debug warnings, and recalculated the data
2289 unnecessarily. In any case, I expanded the warning message to give more
2292 15. Added the message "Restricted characters in address" to the statements in
2293 the default ACL that block characters like @ and % in local parts.
2295 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2296 Three changes have been made:
2298 (a) There was a serious bug; a negative response to MAIL caused the whole
2299 recipient domain to be cached as invalid, thereby blocking all messages
2300 to all local parts at the same domain, from all senders. This bug has
2301 been fixed. The domain is no longer cached after a negative response to
2302 MAIL if the sender used is not empty.
2304 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2307 (c) A new callout option, "use_sender" has been added for people who want
2308 the modified behaviour.
2314 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2317 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2318 indeed breaks things for older releases.
2320 3. Added additional logging to the case where there is a problem reading data
2321 from a filter that is running in a subprocess using a pipe, in order to
2322 try to track down a specific problem.
2324 4. Testing facility fudge: when running in the test harness and attempting
2325 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2326 getting "No route to host". Convert this to a timeout.
2328 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2331 6. Some OS don't have socklen_t but use size_t instead. This affects the
2332 fifth argument of getsockopt() amongst other things. This is now
2333 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2334 can be set for individual OS. I have set it for SunOS5, OSF1, and
2335 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2336 some earlier ones do not.
2338 7. Change 4.30/15 was not doing the test caselessly.
2340 8. The standard form for an IPv6 address literal was being rejected by address
2341 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2342 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2343 this, as well as the form without the "IPv6" on the front (but only when
2344 address literals are enabled, of course).
2346 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2348 10. Exim crashed if a message with an empty sender address specified by -f
2349 encountered a router with an errors_to setting. This could be provoked only
2350 by a command such as
2354 where an empty string was supplied; "<>" did not hit this bug.
2356 11. Installed PCRE release 4.5.
2358 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2359 remained set. It is now erased.
2361 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2362 times from message ids (which are base 36 rather than the normal 62).
2364 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2365 were being counted as actual protocol errors, and logged if the log
2366 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2367 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2368 and DATA following a set of rejected RCPTs do not count as protocol errors.
2369 In other words, Exim assumes they were pipelined, though this may not
2370 actually be the case. Of course, in all cases the client gets an
2371 appropriate error code.
2373 15. If a lookup fails in an ACL condition, a message about the failure may
2374 be available; it is used if testing the ACL cannot continue, because most
2375 such messages specify what the cause of the deferral is. However, some
2376 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2377 that caused an old message to be retained and used if a later statement
2378 caused a defer, replacing the real cause of the deferral.
2380 16. If an IP address had so many PTR records that the DNS lookup buffer
2381 was not large enough to hold them, Exim could crash while trying to process
2382 the truncated data. It now detects and logs this case.
2384 17. Further to 4.21/58, another change has been made: if (and only if) the
2385 first line of a message (the first header line) ends with CRLF, a bare LF
2386 in a subsequent header line has a space inserted after it, so as not to
2387 terminate the header.
2389 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2390 unnecessarily, used atoi() instead of strtol(), and didn't check the
2391 termination when getting file sizes from file names by regex.
2393 19. Completely re-implemented the support for maildirsize files, in the light
2394 of a number of problems with the previous contributed implementation
2395 (4.30/29). In particular:
2397 . If the quota is zero, the maildirsize file is maintained, but no quota is
2400 . If the maildir directory does not exist, it is created before any attempt
2401 to write a maildirsize file.
2403 . The quota value in the file is just a cache; if the quota is changed in
2404 the transport, the new value overrides.
2406 . A regular expression is available for excluding directories from the
2409 20. The autoreply transport checks the characters in options that define the
2410 message's headers; it allows continued headers, but it was checking with
2411 isspace() after an embedded newline instead of explicitly looking for a
2414 21. If all the "regular" hosts to which an address was routed had passed their
2415 expiry times, and had not reached their retry times, the address was
2416 bounced, even if fallback hosts were defined. Now Exim should go on to try
2419 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2420 equivalent code in the SMTP transport. Some hosts send humungous responses
2421 to HELO/EHLO, more than 1024 it seems.
2423 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2424 gives compiler warnings in some environments. I've now done it "properly",
2427 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2428 (because of problems with the built-in one) was declared to return uschar *
2429 instead of char *, causing compiler failure.
2431 25. Fixed a file descriptor leak when processing alias/forward files.
2433 26. Fixed a minor format string issue in dbfn.c.
2435 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2437 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2438 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2441 29. When a sender address is verified, it is cached, to save repeating the test
2442 when there is more than one recipient in a message. However, when the
2443 verification involves a callout, it is possible for different callout
2444 options to be set for different recipients. It is too complicated to keep
2445 track of this in the cache, so now Exim always runs a verification when a
2446 callout is required, relying on the callout cache for the optimization.
2447 The overhead is duplication of the address routing, but this should not be
2450 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2451 to be verified with callout=postmaster, and the main callout worked but the
2452 postmaster check failed, the verification correctly failed. However, if a
2453 subsequent RCPT command asked for sender verification *without* the
2454 postmaster check, incorrect caching caused this verification also to fail,
2457 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2458 it was not caching the DNS options (qualify_single, search_parents) that
2459 were used when the lookup failed. A subsequent lookup with different
2460 options therefore always gave the same answer, though there were cases
2461 where it should not have. (Example: a "domains = !$mx_any" option on a
2462 dnslookup router: the "domains" option is always processed without any
2463 widening, but the router might have qualify_single set.) Now Exim uses the
2464 cached value only when the same options are set.
2466 32. Added John Jetmore's "exipick" utility to the distribution.
2468 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2469 than a timeout (e.g. a certificate is required, and is not provided), an
2470 Exim server now closes the connection immediately. Previously it waited for
2471 the client to close - but if the client is SSL, it seems that they each
2472 wait for each other, leading to a delay before one of them times out.
2474 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2475 maintained 0.8.x compatibility because I don't think many are using it, and
2476 it is clearly obsolete.
2478 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2481 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2482 client certificate was expired. A simple patch fixes this, though I don't
2483 understand the full logic of why the verify callback is called multiple
2486 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2487 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2488 which causes problems with some clients (such as the Certicom SSL Plus
2489 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2490 disables the coutermeasure allowing Eudora to connect."
2492 38. Exim was not checking that a write() to a log file succeeded. This could
2493 lead to Bad Things if a log got too big, in particular if it hit a file
2494 size limit. Exim now panics and dies if it cannot write to a log file, just
2495 as it does if it cannot open a log file.
2497 39. Modified OS/Makefile-Linux so that it now contains
2499 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2501 The two -D definitions ensure that Exim is compiled with large file
2502 support, which makes it possible to handle log files that are bigger than
2505 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2506 instance) a domain was checked against a named list that involved a lookup,
2507 causing $domain_data to be set, then another domain was checked against the
2508 same list, then the first domain was re-checked, the value of $domain_data
2509 after the final check could be wrong. In particular, if the second check
2510 failed, it could be set empty. This bug probably also applied to
2513 41. The strip_trailing_dot option was not being applied to the address given
2514 with the -f command-line option.
2516 42. The code for reading a message's header from the spool was incrementing
2517 $received_count, but never initializing it. This meant that the value was
2518 incorrect (doubled) while delivering a message in the same process in which
2519 it was received. In the most common configuration of Exim, this never
2520 happens - a fresh exec is done - but it can happen when
2521 deliver_drop_privilege is set.
2523 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2524 it now includes up to 150 characters of the unexpected data in the log
2527 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2528 and building data strings. The size of both of these buffers was 10 000
2529 bytes - far larger than anybody would *ever* want, thought I. Needless to
2530 say, somebody hit the limit. I have increased the maximum line length to
2531 20 000 and the maximum data length of concatenated lines to 100 000. I have
2532 also fixed two bugs, because there was no checking on these buffers. Tsk,
2533 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2534 buffer is too small.
2536 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2537 lsearch lookups. Now it does.
2539 46. When parsing a route_list item in a manualroute router, a fixed-length
2540 buffer was used for the list of hosts. I made this 1024 bytes long,
2541 thinking that nobody would ever have a list of hosts that long. Wrong.
2542 Somebody had a whole pile of complicated expansion conditions, and the
2543 string was silently truncated, leading to an expansion error. It turns out
2544 that it is easier to change to an unlimited length (owing to other changes
2545 that have happened since this code was originally written) than to build
2546 structure for giving a limitation error. The length of the item that
2547 expands into the list of hosts is now unlimited.
2549 47. The lsearch lookup could not handle data where the length of text line was
2550 more than 4095 characters. Such lines were truncated, leading to shortened
2551 data being returned. It should now handle lines of any length.
2553 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2554 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2557 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2558 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2559 no longer happens in comment lines. A list of replacements is now placed
2560 at the head of all of the source files, except those whose only change is
2561 to replace PERL_COMMAND in the very first #! line.
2563 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2564 messages on the queue, with a bottom-up merge sort, using code contributed
2565 by Michael Haardt. This should make operations like -bp somewhat faster on
2566 large queues. It won't affect queue runners, except when queue_run_in_order
2569 51. Installed eximstats 1.31 in the distribution.
2571 52. Added support for SRV lookups to the dnslookup router.
2573 53. If an ACL referred to $message_body or $message_body_end, the value was not
2574 reset for any messages that followed in the same SMTP session.
2576 54. The store-handling optimization for building very long strings was not
2577 differentiating between the different store pools. I don't think this
2578 actually made any difference in practice, but I've tidied it.
2580 55. While running the routers to verify a sender address, $sender_address
2581 was still set to the sender address. This is wrong, because when routing to
2582 send a bounce to the sender, it would be empty. Therefore, I have changed
2583 it so that, while verifying a sender address, $sender_address is set to <>.
2584 (There is no change to what happens when verifying a recipient address.)
2586 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2587 target A or AAAA records (if not already returned) without resetting the
2588 qualify_single or search_parents options of the DNS resolver. These are
2589 inappropriate in this case because the targets of MX and SRV records must
2590 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2591 target that, when qualified, matched something in the local domain. These
2592 two options are now turned off when doing these lookups.
2594 57. It seems that at least some releases of Reiserfs (which does not have the
2595 concept of a fixed number of inodes) returns zero and not -1 for the
2596 number of available inodes. This interacted badly with check_spool_inodes,
2597 which assumed that -1 was the "no such thing" setting. What I have done is
2598 to check that the total number of inodes is greater than zero before doing
2599 the test of how many are available.
2601 58. When a "warn" ACL statement has a log_message modifier, the message is
2602 remembered, and not repeated. This is to avoid a lot of repetition when a
2603 message has many recipients that cause the same warning to be written.
2604 Howewer, Exim was preserving the list of already written lines for an
2605 entire SMTP session, which doesn't seem right. The memory is now reset if a
2606 new message is started.
2608 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2609 debugging output unless log_rewrite was also set.
2611 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2612 of (int)(handle) when we know that handle contains (void *)(-1).
2614 61. The Exim daemon panic-logs an error return when it closes the incoming
2615 connection. However "connection reset by peer" seems to be common, and
2616 isn't really an error worthy of noting specially, so that particular error
2619 62. When Exim is trying to find all the local interfaces, it used to panic and
2620 die if the ioctl to get the interface flags failed. However, it seems that
2621 on at least one OS (Solaris 9) it is possible to have an interface that is
2622 included in the list of interfaces, but for which you get a failure error
2623 for this call. This happens when the interface is not "plumbed" into a
2624 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2625 failure of the "get flags" call assumes that the interface is down.
2627 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2628 makes life easier for people who are doing arithmetic on fields extracted
2629 from dates, where you often get leading zeros that should not be
2630 interpreted as octal.
2632 64. Added qualify_domain to the redirect router, to override the global
2635 65. If a pathologically long header line contained very many addresses (the
2636 report of this problem mentioned 10 000) and each of them was rewritten,
2637 Exim could use up a very large amount of memory. (It kept on making new
2638 copies of the header line as it rewrote, and never released the old ones.)
2639 At the expense of a bit more processing, the header rewriting function has
2640 been changed so that it no longer eats memory in this way.
2642 66. The generation of the Received: header has been moved from the time that a
2643 message starts to be received, to the time that it finishes. The timestamp
2644 in the Received: header should now be very close to that of the <= log
2645 line. There are two side-effects of this change:
2647 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2648 logged header lines no longer include the local Received: line, because
2649 it has not yet been created. The same applies to a copy of the message
2650 that is returned to a non-SMTP sender when a message is rejected.
2652 (b) When a filter file is tested using -bf, no additional Received: header
2653 is added to the test message. After some thought, I decided that this
2656 This change does not affect the value of $received_for. It is still set
2657 after address rewriting, but before local_scan() is called.
2659 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2661 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2662 gave an unhelpful panic error message, and a defer error. I have managed to
2663 change this behaviour so that it now rejects any supplied certificate,
2664 which seems right, as the list of acceptable certificates is empty.
2666 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2667 gave an unhelpful defer error. I have not managed to make this reject any
2668 supplied certificates, but the error message it gives is "no certificate
2669 supplied", which is not helpful.
2671 70. exigrep's output now also includes lines that are not associated with any
2672 message, but which match the given pattern. Implemented by a patch from
2673 Martin Sluka, which also tidied up the Perl a bit.
2675 71. Recipient callout verification, like sender verification, was using <> in
2676 the MAIL FROM command. This isn't really the right thing, since the actual
2677 sender may affect whether the remote host accepts the recipient or not. I
2678 have changed it to use the actual sender in the callout; this means that
2679 the cache record is now keyed on a recipient/sender pair, not just the
2680 recipient address. There doesn't seem to be a real danger of callout loops,
2681 since a callout by the remote host to check the sender would use <>.
2682 [SEE ABOVE: changed after hitting problems.]
2684 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2685 temporary errors. However, in the case of such a code being given after
2686 the end of a data transmission (i.e. after ".") Exim was failing to write
2687 a retry record for the message. (Yes, there was some broken host that was
2688 actually sending 8xx at this point.)
2690 73. An unknown lookup type in a host list could cause Exim to panic-die when
2691 the list was checked. (An example that provoked this was putting <; in the
2692 middle of a list instead of at the start.) If this happened during a DATA
2693 ACL check, a -D file could be left lying around. This kind of configuration
2694 error no longer causes Exim to die; instead it causes a defer errror. The
2695 incident is still logged to the main and panic logs.
2697 74. Buglet left over from Exim 3 conversion. The message "too many messages
2698 in one connection" was written to the rejectlog but not the mainlog, except
2699 when address rewriting (yes!) was being logged.
2701 75. Added write_rejectlog option.
2703 76. When a system filter was run not as root (that is, when system_filter_user
2704 was set), the values of the $n variables were not being returned to the
2705 main process; thus, they were not subsequently available in the $sn
2708 77. Added +return_path_on_delivery log selector.
2710 78. A connection timeout was being treated differently from recipients deferred
2711 when testing hosts_max_try with a message that was older than the host's
2712 retry timeout. (The host should not be counted, thus allowing all hosts to
2713 be tried at least once before bouncing.) This may have been the cause of an
2714 occasionally reported bug whereby a message would remain on the queue
2715 longer than the retry timeout, but would be bounced if a delivery was
2716 forced. I say "may" because I never totally pinned down the problem;
2717 setting up timeout/retry tests is difficult. See also the next item.
2719 79. The ultimate address timeout was not being applied to errors that involved
2720 a combination of host plus message (for example, a timeout on a MAIL
2721 command). When an address resolved to a number of possible hosts, and they
2722 were not all tried for each delivery (e.g. because of hosts_max_try), a
2723 message could remain on the queue longer than the retry timeout.
2725 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2728 81. Fixed an obscure SMTP outgoing bug which required at least the following
2729 conditions: (a) there was another message waiting for the same server;
2730 (b) the server returned 5xx to all RCPT commands in the first message so
2731 that the message was not completed; (c) the server dropped the connection
2732 or gave a negative response to the RSET that Exim sends to abort the
2733 transaction. The observed case was a dropped connection after DATA that had
2734 been sent in pipelining mode. That is, the server had advertised PIPELINING
2735 but was not implementing it correctly. The effect of the bug was incorrect
2736 behaviour, such as trying another host, and this could lead to a crash.
2742 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2743 and daemon.c were passed as pointers to ints; they should have been
2744 pointers to socklen_t variables (which are typically unsigned ints).
2746 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2749 3. Fixed a really odd bug that affected only the testing scheme; patching a
2750 certain fixed string in the binary changed the value of another string that
2751 happened to be identical to the end of the original first string.
2753 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2754 name", it returns that address as the IP address. On some operating
2755 systems (e.g. Solaris), it also passes back the IP address string as the
2756 "host name". However, on others (e.g. Linux), it passes back an empty
2757 string. Exim wasn't checking for this, and was changing the host name to an
2758 empty string, assuming it had been canonicized.
2760 5. Although rare, it is permitted to have more than one PTR record for a given
2761 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2762 all the names associated with an address, because they do in Solaris.
2763 However, it seems that they do not in Linux for data that comes from the
2764 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2765 I found this out when I moved to a new Linux workstation and tried to run
2766 the Exim test suite.
2768 To get round this problem I have changed the code so that it now does its
2769 own call to the DNS to look up PTR records when searching for a host name.
2770 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2771 addresses that are only in /etc/hosts are still found.
2773 This behaviour is, however, controlled by an option called host_lookup_
2774 order, which defaults to "bydns:byaddr". If people want to use the other
2775 order, or indeed, just use one or the other means of lookup, they can
2776 specify it in this variable.
2778 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2779 some operating systems, this comes back from gethostbyaddr() as an empty
2780 string, and this is what Exim used to test for. However, it seems that in
2781 other systems, "." is yielded. Exim now tests for this case too.
2783 7. The values of check_spool_space and check_log_space are now held internally
2784 as a number of kilobytes instead of an absolute number of bytes. If a
2785 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2786 kilobyte. This means that much larger values can be stored.
2788 8. Exim monitor: an attempt to get the action menu when not actually pointing
2789 at a message produces an empty menu entitled "No message selected". This
2790 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2791 no entries in it ("Shell widget menu has zero width and/or height"). So I
2792 have added a single, blank menu entry in this case.
2794 9. Added ${quote_local_part.
2796 10. MIME decoding is now applied to the contents of Subject: header lines when
2799 11. Now that a reference to $sender_host_address automatically causes a reverse
2800 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2801 host lookup before query-style lookups in lists that might use this
2802 variable. This has therefore been abolished, and the "net-" prefix is no
2803 longer necessary for query-style lookups.
2805 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2806 have been a typo for LFLAGS, so it has been changed.
2808 13. The install script calls Exim with "-C /dev/null" in order to find the
2809 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2810 to be output. Howeve, since Exim outputs its version number before the
2811 error, it didn't break the script. It just looked ugly. I fixed this by
2812 always allowing "-C /dev/null" if the caller is root.
2814 14. Ignore overlarge ACL variable number when reading spool file - insurance
2815 against a later release with more variables having written the file.
2817 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2818 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2819 this, as well as the form without the "IPv6" on the front.
2821 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2822 OS/Makefile-Darwin file.
2824 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2825 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2827 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2828 variable. However, some OS use ints for inodes. I've added cast to long int
2829 to get rid of the compiler warning.
2831 19. I had forgotten to lock out "/../" in configuration file names when
2832 ALT_CONFIG_PREFIX was set.
2834 20. Routers used for verification do not need to specify transports. However,
2835 if such a router generated a host list, and callout was configured, Exim
2836 crashed, because it could not find a port number from the (non-existent)
2837 transport. It now assumes port 25 in this circumstance.
2839 21. Added the -t option to exigrep.
2841 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2842 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2843 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2844 them is set, LOOKUP_LSEARCH is forced.
2846 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2848 24. Added sender and host information to the "rejected by local_scan()" log
2849 line; previously there was no indication of these.
2851 25. Added .include_if_exists.
2853 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2854 Linux. It turns out that not all file systems support this. Apparently some
2855 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2856 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2857 not supported on the file descriptor, is now ignored when Exim is trying to
2858 sync a directory. This applies only to Linux.
2860 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2862 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2863 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2866 29. Added support for maildirsize files from supplied patch (modified a bit).
2868 30. The use of :fail: followed by an empty string could lead Exim to respond to
2869 sender verification failures with (e.g.):
2871 550 Verification failed for <xxx>
2872 550 Sender verify failed
2874 where the first response line was missing the '-' that indicates it is not
2875 the final line of the response.
2877 31. The loop for finding the name of the user that called Exim had a hardwired
2878 limit of 10; it now uses the value of finduser_retries, which is used for
2879 all other user lookups.
2881 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2884 33. Exim was neglecting to zero errno before one call of strtol() when
2885 expanding a string and expecting an integer value. On some systems this
2886 resulted in spurious "integer overflow" errors. Also, it was casting the
2887 result into an int without checking.
2889 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2890 did not work. The code looks as if it has *never* worked, though it appears
2891 to have been documented since at least releast 1.62. I have made it work.
2893 35. The "timeout_DNS" error in retry rules, also documented since at least
2894 1.62, also never worked. As it isn't clear exactly what this means, and
2895 clearly it isn't a major issue, I have abolished the feature by treating it
2896 as "timeout", and writing a warning to the main and panic logs.
2898 36. The display of retry rules for -brt wasn't always showing the error code
2901 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2902 timeout_connect_A, timeout_connect_MX.
2904 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2905 to the empty sender.
2907 39. The daemon was not analysing the content of -oX till after it had closed
2908 stderr and disconnected from the controlling terminal. This meant that any
2909 syntax errors were only noted on the panic log, and the return code from
2910 the command was 0. By re-arranging the code a little, I've made the
2911 decoding happen first, so such errors now appear on stderr, and the return
2912 code is 1. However, the actual setting up of the sockets still happens in
2913 the disconnected process, so errors there are still only recorded on the
2916 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2917 connections (as happens on some IP stacks) was logged at start up time as
2918 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2919 it from "IPv6 and IPv4", which means that two separate sockets are being
2922 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2923 says whether AF_INET or AF_INET6 was passed as an argument.
2925 42. Exiwhat output was messed up when time zones were included in log
2928 43. Exiwhat now gives more information about the daemon's listening ports,
2929 and whether -tls-on-connect was used.
2931 44. The "port" option of the smtp transport is now expanded.
2933 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2934 silently ignored. Now an error message is written to the main and panic
2937 46. There's a new ACL modifier called "logwrite" which writes to a log file
2938 as soon as it is encountered.
2940 47. Added $local_user_uid and $local_user_gid at routing time.
2942 48. Exim crashed when trying to verify a sender address that was being
2945 49. Exim was recognizing only a space character after ".include". It now also
2946 recognizes a tab character.
2948 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2949 extracting the relevant information from the specification. The man page no
2950 longer contains scrambled data for the -d option, and I've added a section
2951 at the front about calling Exim under different names.
2953 51. Added "extra_headers" argument to the "mail" command in filter files.
2955 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2958 53. Installed eximstats 1.29.
2960 54. Added transport_filter_timeout as a generic transport option.
2962 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2963 Cc: header lines. This was required by RFC 822, but it not required by RFC
2966 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2967 incoming messages that did not have them. Now it does so only if the
2968 message originates locally, that is, if there is no associated remote host
2969 address. When Resent- header lines are present, this applies to the Resent-
2970 lines rather than the non-Resent- lines.
2972 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2973 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2975 58. Messages for configuration errors now include the name of the main
2976 configuration file - useful now that there may be more than one file in a
2977 list (.included file names were always shown).
2979 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2980 for those rare installations that do not start the daemon as root or run it
2981 setuid root. I've cut out the call to initgroups() if the daemon is not
2984 60. The Exim user and group can now be bound into the binary as text strings
2985 that are looked up at the start of Exim's processing.
2987 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2989 62. Added $mailstore_basename variable.
2991 63. Installed patch to sieve.c from Michael Haardt.
2993 64. When Exim failed to open the panic log after failing to open the main log,
2994 the original message it was trying to log was written to stderr and debug
2995 output, but if they were not available (the usual case in production), it
2996 was lost. Now it is written to syslog before the two lines that record the
2997 failures to open the logs.
2999 65. Users' Exim filters run in subprocesses under the user's uid. It is
3000 possible for a "deliver" command or an alias in a "personal" command to
3001 provoke an address rewrite. If logging of address rewriting is configured,
3002 this fails because the process is not running as root or exim. There may be
3003 a better way of dealing with this, but for the moment (because 4.30 needs
3004 to be released), I have disabled address rewrite logging when running a
3005 filter in a non-root, non-exim process.
3011 1. The buildconfig auxiliary program wasn't quoting the value set for
3012 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3013 not defined. This bug was masked in 4.22 by the effect that was fixed in
3016 2. Some messages that were rejected after a message id was allocated were
3017 shown as "incomplete" by exigrep. It no longer does this for messages that
3018 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3020 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3021 have allow_domain_literals set, the ID did not get logged in the <= line.
3022 Domain literals are now always recognized in Message-ID: header lines.
3024 4. The first argument for a ${extract expansion item is the key name or field
3025 number. Leading and trailing spaces in this item were not being ignored,
3026 causing some misleading effects.
3028 5. When deliver_drop_privilege was set, single queue runner processes started
3029 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3030 same command in the process it spins off) were not dropping privilege.
3032 6. When the daemon running as "exim" started a queue runner, it always
3033 re-executed Exim in the spun-off process. This is a waste of effort when
3034 deliver_drop_privilege is set. The new process now just calls the
3035 queue-runner function directly.
3041 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3044 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3045 ignored. Though the use of -oP was forcing the writing of a pid file, it
3046 was always written to the default place.
3048 3. If the message "no IP address found for host xxxx" is generated during
3049 incoming verification, it is now followed by identification of the incoming
3050 connection (so you can more easily find what provoked it).
3052 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3054 5. Added some features to "harden" Exim a bit more against certain attacks:
3056 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3057 be put in Local/Makefile. This is like the never_users runtime option,
3058 but it cannot be overridden. The default setting is "root".
3060 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3061 prefix string with which any file named in a -C command line option
3064 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3065 is retained for -C and -D only if the caller of Exim is root. Without
3066 it, the exim user may also use -C and -D and retain privilege.
3068 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3069 command line option is disabled.
3071 6. Macro names set by the -D option must start with an upper case letter, just
3072 like macro names defined in the configuration file.
3074 7. Added "dereference=" facility to LDAP.
3076 8. Two instances of the typo "uknown" in the source files are fixed.
3078 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3079 the Configure-Makefile script screwed up while processing it.
3081 10. Incorporated PCRE 4.4.
3083 11. The SMTP synchronization check was not operating right at the start of an
3084 SMTP session. For example, it could not catch a HELO sent before the client
3085 waited for the greeting. There is now a check for outstanding input at the
3086 point when the greeting is written. Because of the duplex, asynchronous
3087 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3088 way, but not yet received, when the check is performed.
3090 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3091 on TCP/IP sockets, because this apparently causes some broken clients to
3094 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3095 unchanged) from the Cygwin maintainer.
3097 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3098 was supported instead of testing for mbx. Effectively a typo.
3100 15. The spa authenticator server code was not checking that the input it
3101 received was valid base64.
3103 16. The debug output line for the "set" modifier in ACLs was not showing the
3104 name of the variable that was being set.
3106 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3108 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3109 lookup on its own. Something else was needed to trigger the lookup. For
3110 example, a match in host_lookup or the need for a host name in a host list.
3111 Now, if $sender_host_name is referenced and the host name has not yet been
3112 looked up, a lookup is performed. If the lookup fails, the variable remains
3113 empty, and $host_lookup_failed is set to "1".
3115 19. Added "eqi" as a case-independent comparison operator.
3117 20. The saslauthd authentication condition could segfault if neither service
3118 nor realm was specified.
3120 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3121 error message that was logged was misleading, and incoming SMTP
3122 connections were dropped. The message is now more accurate, and temporary
3123 errors are given to SMTP connections.
3125 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3126 (except RSET) with a 421 error, until QUIT is received. However, it was
3127 failing to send a response to QUIT.
3129 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3130 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3131 ACL is now run after the helo_try_verify_hosts code.
3133 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3134 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3135 case-independent, so other case variants are also recognized). Apparently
3136 some systems use these upper case variants.
3138 25. If more than two messages were waiting for the same host, and a transport
3139 filter was specified for the transport, Exim sent two messages over the
3140 same TCP/IP connection, and then failed with "socket operation on non-
3141 socket" when it tried to send the third.
3143 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3145 27. The extern definition of crypt16() in expand.c was not being excluded when
3146 the OS had its own crypt16() function.
3148 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3149 as a preferred synonym for return_size_limit, both as an option and as an
3152 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3154 30. Changed the default configuration ACL to relax the local part checking rule
3155 for addresses that are not in any local domains. For these addresses,
3156 slashes and pipe symbols are allowed within local parts, but the sequence
3157 /../ is explicitly forbidden.
3159 31. SPA server authentication was not clearing the challenge buffer before
3162 32. log_message in a "warn" ACL statement was writing to the reject log as
3163 well as to the main log, which contradicts the documentation and doesn't
3164 seem right (because no rejection is happening). So I have stopped it.
3166 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3167 However, I am unable to do any testing of this.
3169 34. Fixed an infelicity in the appendfile transport. When checking directories
3170 for a mailbox, to see if any needed to be created, it was accidentally
3171 using path names with one or more superfluous leading slashes; tracing
3172 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3174 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3175 contents are added to the log line that is written for every discarded
3176 recipient. (Previously a log_message setting was ignored.)
3178 36. The ${quote: operator now quotes the string if it is empty.
3180 37. The install script runs exim in order to find its version number. If for
3181 some reason other than non-existence or emptiness, which it checks, it
3182 could not run './exim', it was installing it with an empty version number,
3183 i.e. as "exim-". This error state is now caught, and the installation is
3186 38. An argument was missing from the function that creates an error message
3187 when Exim fails to connect to the socket for saslauthd authentication.
3188 This could cause Exim to crash, or give a corrupted message.
3190 39. Added isip, isip4, and isip6 to ${if conditions.
3192 40. The ACL variables $acl_xx are now saved with the message, and can be
3193 accessed later in routers, transports, and filters.
3195 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3196 strings in the file are not string-expanded.
3198 42. If a MAIL command specified a SIZE value that was too large to fit into an
3199 int variable, the check against message_size_limit failed. Such values are
3200 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3201 day this will have to be increased, but I don't think I want to be around
3202 when emails are that large.
3209 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3210 iconv() is not standard in FreeBSD.
3212 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3213 IPv6 enabled. The observed symptom was a segmentation fault on return from
3214 the function os_common_find_running_interfaces() in src/os.c.
3216 3. In the check_special_case() function in daemon.c I had used "errno" as an
3217 argument name, which causes warnings on some systems. This was basically a
3218 typo, since it was named "eno" in the comments!
3220 4. The code that waits for the clock to tick (at a resolution of some fraction
3221 of a second) so as to ensure message-id uniqueness was always waiting for
3222 at least one whole tick, when it could have waited for less. [This is
3223 almost certainly not relevant at current processor speeds, where it is
3224 unlikely to ever wait at all. But we try to future-proof.]
3226 5. The function that sleeps for a time interval that includes fractions of a
3227 second contained a race. It did not block SIGALRM between setting the
3228 timer, and suspending (a couple of lines later). If the interval was short
3229 and the sigsuspend() was delayed until after it had expired, the suspension
3230 never ended. On busy systems this could lead to processes getting stuck for
3233 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3234 process, before it forks any delivery processes. The open lookup caching
3235 mechanism meant that the open file or database connection was passed into
3236 the delivery process. The problem was that delivery processes always tidy
3237 up cached lookup data. This could cause a problem for the next delivery
3238 process started by the queue runner, because the external queue runner
3239 process does not know about the closure. So the next delivery process
3240 still has data in the lookup cache. In the case of a file lookup, there was
3241 no problem because closing a file descriptor in a subprocess doesn't affect
3242 the parent. However, if the lookup was caching a connection to a database,
3243 the connection was closed, and the second delivery process was likely to
3244 see errors such as "PGSQL: query failed: server closed the connection
3245 unexpectedly". The problem has been fixed by closing all cached lookups
3246 in a queue runner before running a delivery process.
3248 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3249 seem to have the "const" qualifier which it has on other OS. I've
3252 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3253 *of the same type* (client or server) with the same public name that an
3254 error should be diagnosed.
3256 9. When Exim looked up a host name for an IP address, but failed to find the
3257 original IP address when looking up the host name (a safety check), it
3258 output the message "<ip address> does not match any IP for NULL", which was
3259 confusing, to say the least. The bug was that the host name should have
3260 appeared instead of "NULL".
3262 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3263 user that is built into the binary, and the -C or -D options is used, root
3264 privilege is dropped before the configuration file is read. In addition,
3265 logging is switched to stderr instead of the normal log files. If the
3266 configuration then re-defines the Exim user, the unprivileged environment
3267 is probably not what is expected, so Exim logs a panic warning message (but
3270 However, if deliver_drop_privilege is set, the unprivileged state may well
3271 be exactly what is intended, so the warning has been cut out in that case,
3272 and Exim is allowed to try to write to its normal log files.
3278 1. smtp_return_error_details was not giving details for temporary sender
3279 or receiver verification errors.
3281 2. Diagnose a configuration error if two authenticators have the same public
3284 3. Exim used not to create the message log file for a message until the first
3285 delivery attempt. This could be confusing when incoming messages were held
3286 for policy or load reasons. The message log file is now created at the time
3287 the message is received, and an initial "Received" line is written to it.
3289 4. The automatically generated man page for command line options had a minor
3290 bug that caused no ill effects; however, a more serious problem was that
3291 the procedure for building the man page automatically didn't always
3292 operate. Consequently, release 4.20 contains an out-of-date version. This
3293 shouldn't happen again.
3295 5. When building Exim with embedded Perl support, the script that builds the
3296 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3297 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3299 6. The freeze_tell option was not being used for messages that were frozen on
3300 arrival, either by an ACL or by local_scan().
3302 7. Added the smtp_incomplete_transaction log selector.
3304 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3305 was accepting AUTH without a new EHLO.
3307 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3308 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3311 10. Logging of TCP/IP connections (when configured) now happens in the main
3312 daemon process instead of the child process, so that the TCP/IP connection
3313 count is more accurate (but it can never be perfect).
3315 11. The use of "drop" in a nested ACL was not being handled correctly in the
3316 outer ACL. Now, if condition failure induced by the nested "drop" causes
3317 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3318 or "require"), the connection is dropped.
3320 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3321 that yield "discard" can now be used with an "accept" or a "discard" verb,
3322 but an error is generated for any others (because I can't see a useful way
3323 to define what should happen).
3325 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3326 are now processed in the same way as lines in the Exim configuration file.
3327 In particular, continuation lines are supported.
3329 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3331 15. Added -ti meaning -t -i.
3333 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3334 domains, and warn by logging if others are found.
3336 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3337 returned from GIFCONF when Exim is trying to find the list of interfaces on
3338 a host. The code in os.c has been modified to copy each ifreq to an aligned
3339 structure in all cases.
3341 Also, in some cases, the returned ifreq's were being copied to a 'struct
3342 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3343 means the last couple of bytes of an IPv6 address could be chopped if the
3344 ifreq contained only a normal sockaddr (14 bytes storage).
3346 18. Named domain lists were not supported in the hosts_treat_as_local option.
3347 An entry such as +xxxx was not recognized, and was treated as a literal
3350 19. Ensure that header lines added by a DATA ACL are included in the reject log
3351 if the ACL subsequently rejects the message.
3353 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3354 MD5 (which is deprecated).
3356 21. When testing a filter file using -bf, Exim was writing a message when it
3357 took the sender from a "From " line in the message, but it was not doing so
3358 when it took $return_path from a Return-Path: header line. It now does.
3360 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3361 with a valid header line field name (a series of printing characters
3362 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3364 23. Changed "disc" in the source to "disk" to conform to the documentation and
3365 the book and for uniformity.
3367 24. Ignore Sendmail's -Ooption=value command line item.
3369 25. When execve() failed while trying to run a command in a pipe transport,
3370 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3371 could be confused with a return value of 69 from the command itself. This
3372 has been changed to 127, the value the shell returns if it is asked to run
3373 a non-existent command. The wording for the related log line suggests a
3374 non-existent command as the problem.
3376 26. If received_header_text expands to an empty string, do not add a Received:
3377 header line to the message. (Well, it adds a token one on the spool, but
3378 marks it "old" so that it doesn't get used or transmitted.)
3380 27. Installed eximstats 1.28 (addition of -nt option).
3382 28. There was no check for failure on the call to getsockname() in the daemon
3383 code. This can fail if there is a shortage of resources on the system, with
3384 ENOMEM, for example. A temporary error is now given on failure.
3386 29. Contrary to the C standard, it seems that in some environments, the
3387 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3388 program. Exim now does this explicitly; it affects the formatting of
3389 timestamps using strftime().
3391 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3392 complaints. I've now initialized all the variables, to avoid this.
3394 32. Header lines added by a system filter were not being "seen" during
3395 transport-time rewrites.
3397 33. The info_callback() function passed to OpenSSL is set up with type void
3398 (*)(SSL *, int, int), as described somewhere. However, when calling the
3399 function (actually a macro) that sets it up, the type void(*)() is
3400 expected. I've put in a cast to prevent warnings from picky compilers.
3402 34. If a DNS black list lookup found a CNAME record, but there were no A
3403 records associated with the domain it pointed at, Exim crashed.
3405 35. If a DNS black list lookup returned more than one A record, Exim ignored
3406 all but the first. It now scans all returned addresses if a particular IP
3407 value is being sought. In this situation, the contents of the
3408 $dnslist_value variable are a list of all the addresses, separated by a
3411 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3412 to accept a host name and all its aliases if the forward lookup for any of
3413 them yielded the IP address of the incoming connection. Now it accepts only
3414 those names whose forward lookup yields the correct IP address. Any other
3415 names are discarded. This closes a loophole whereby a rogue DNS
3416 administrator could create reverse DNS records to break through a
3417 wildcarded host restriction in an ACL.
3419 37. If a user filter or a system filter that ran in a subprocess used any of
3420 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3421 the wrong values were passed to the pipe command ($thisaddress had the
3422 value of $0, $0 had the value of $1, etc). This bug was introduced by
3423 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3425 38. Improved the line breaking for long SMTP error messages from ACLs.
3426 Previously, if there was no break point between 40 and 75 characters, Exim
3427 left the rest of the message alone. Two changes have been made: (a) I've
3428 reduced the minimum length to 35 characters; (b) if it can't find a break
3429 point between 35 and 75 characters, it looks ahead and uses the first one
3430 that it finds. This may give the occasional overlong line, but at least the
3431 remaining text gets split now.
3433 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3434 file descriptors might be low, and that setting 1000 would always raise it.
3435 It turns out that in some environments, the limit is already over 1000 and
3436 that lowering it causes trouble. So now Exim takes care not to decrease it.
3438 40. When delivering a message, the value of $return_path is set to $sender_
3439 address at the start of routing (routers may change the value). By an
3440 oversight, this default was not being set up when an address was tested by
3441 -bt or -bv, which affected the outcome if any router or filter referred to
3444 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3445 log only when "message" or "log_message" are set. However, if one of the
3446 conditions was an address verification, or a call to a nested ACL, the
3447 messages generated by the underlying test were being passed through. This
3448 no longer happens. The underlying message is available in $acl_verify_
3449 message for both "message" and "log_message" expansions, so it can be
3450 passed through if needed.
3452 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3453 new expansion $bh_ to give the encoded byte string without charset
3454 translation. Translation happens only if iconv() is available; HAVE_ICONV
3455 indicates this at build time. HEADERS_CHARSET gives the charset to
3456 translate to; headers_charset can change it in the configuration, and
3457 "headers charset" can change it in an individual filter file.
3459 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3460 that creates RFC 2047 encoded "words" labels them as that charset instead
3461 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3462 expansion operator; (ii) when Exim creates a From: line for a local
3463 message; (iii) when a header line is rewritten to include a "phrase" part.
3465 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3466 buggy, causing it to skip the first lines of messages whose message ID
3467 ended in 'D'. This would not have bitten before Exim release 4.14, because
3468 message IDs were unlikely to end in 'D' before then. The effect was to have
3469 incorrect size information for certain domains.
3471 45. #include "config.h" was missing at the start of the crypt16.c module. This
3472 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3475 46. If there was a timeout during a "random" callout check, Exim treated it as
3476 a failure of the random address, and carried on sending RSET and the real
3477 address. If the delay was just some slowness somewhere, the response to the
3478 original RCPT would be taken as a response to RSET and so on, causing
3479 mayhem of various kinds.
3481 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3482 when I implemented it. It didn't allow for the fact that some option values
3483 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3484 the right test for positive values.
3486 48. Domain names in DNS records are case-independent. Exim always looks them up
3487 in lower case. Some resolvers return domain names in exactly the case they
3488 appear in the zone file, that is, they may contain uppercase letters. Not
3489 all resolvers do this - some return always lower case. Exim was treating a
3490 change of case by a resolver as a change of domain, similar to a widening
3491 of a domain abbreviation. This triggered its re-routing code and so it was
3492 trying to route what was effectively the same domain again. This normally
3493 caused routing to fail (because the router wouldn't handle the domain
3494 twice). Now Exim checks for this case specially, and just changes the
3495 casing of the domain that it ultimately uses when it transmits the message
3498 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3501 50. If a filter generated a file delivery with a non-absolute name (possible if
3502 no home directory exists for the router), the forbid_file option was not
3505 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3506 the existing equality matching.
3508 52. Exim was using ints instead of ino_t variables in some places where it was
3509 dealing with inode numbers.
3511 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3512 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3513 called TMPDIR, and if it finds it is different, it changes its value.
3515 54. The smtp_printf() function is now made available to local_scan() so
3516 additional output lines can be written before returning. There is also an
3517 smtp_fflush() function to enable the detection of a dropped connection.
3518 The variables smtp_input and smtp_batched_input are exported to
3521 55. Changed the default runtime configuration: the message "Unknown user"
3522 has been removed from the ACL, and instead placed on the localuser router,
3523 using the cannot_route_message feature. This means that any verification
3524 failures that generate their own messages won't get overridden. Similarly,
3525 the "Unrouteable address" message that was in the ACL for unverifiable
3526 relay addresses has also been removed.
3528 56. Added hosts_avoid_esmtp to the smtp transport.
3530 57. The exicyclog script was not checking for the esoteric option
3531 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3532 will work only if exicyclog is run under the appropriate euid.
3534 58. Following a discussion on the list, the rules by which Exim recognises line
3535 endings on incoming messages have been changed. The -dropcr and drop_cr
3536 options are now no-ops, retained only for backwards compatibility. The
3537 following line terminators are recognized: LF CRLF CR. However, special
3538 processing applies to CR:
3540 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3541 nor a local message in the state where . is a terminator.
3543 (ii) If a bare CR is encountered in a header line, an extra space is added
3544 after the line terminator so as not to end the header. The reasoning
3545 behind this is that bare CRs in header lines are most likely either
3546 to be mistakes, or people trying to play silly games.
3548 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3549 was being incorrectly given as 18 bytes larger than it should have been.
3550 This is a VOB (very old bug).
3552 60. This may never have affected anything current, but just in case it has:
3553 When the local host is found other than at the start of a list of hosts,
3554 the local host, those with the same MX, and any that follow, are discarded.
3555 When the list in question was part of a longer list of hosts, the following
3556 hosts (not currently being processed) were also being discarded. This no
3557 longer happens. I'm not sure if this situation could ever has previously
3560 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3563 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3564 header. This is something that is recommended in a new Internet Draft, and
3565 is something that is documented as being done by Sendmail. There are two
3566 possible values. For messages generated by the autoreply transport, Exim
3569 Auto-Submitted: auto-replied
3571 whereas for all other generated messages (e.g. bounces) it adds
3573 Auto-Submitted: auto-generated
3575 63. The "personal" condition in filters now includes a test for the
3576 Auto-Submitted: header. If it contains the string "auto-" the message it
3577 not considered personal.
3579 64. Added rcpt_include_affixes as a generic transport option.
3581 65. Added queue_only_override (default true).
3583 66. Added the syslog_duplication option.
3585 67. If what should have been the first header line of a message consisted of
3586 a space followed by a colon, Exim was mis-interpreting it as a header line.
3587 It isn't of course - it is syntactically invalid and should therefore be
3588 treated as the start of the message body. The misbehaviour could have
3589 caused a number of strange effects, including loss of data in subsequent
3590 header lines, and spool format errors.
3592 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3593 client host had authenticated. This control can now be exercised by an ACL
3594 for more flexibility.
3596 69. By default, callouts do not happen when testing with -bh. There is now a
3597 variant, -bhc, which does actually run the callout code, including
3598 consulting and updating the callout cache.
3600 70. Added support for saslauthd authentication, courtesy of Alexander
3603 71. If statvfs() failed on the spool or log directories while checking their
3604 size for availability, Exim confusingly gave the error "space shortage".
3605 Furthermore, in debugging mode it crashed with a floating point exception.
3606 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3607 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3608 serious problem, Exim now writes to the main and panic logs when this
3609 happens, with details of the failure. It then refuses to accept the
3610 incoming message, giving the message "spool directory problem" or "log
3611 directory problem" with a 421 code for SMTP messages.
3613 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3614 0, 1, and 2 exist, because some OS complain for execs without them (see
3615 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3616 descriptors. However, the code omitted to check that the open succeeded,
3617 causing mysterious errors if for some reason the permissions on /dev/null
3618 got screwed. Now Exim writes a message to the main and panic logs, and
3619 bombs out if it can't open /dev/null.
3621 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3622 interact so that it is all more flexible. It is supposed to remain
3623 backwards compatible. Also added extra_local_interfaces.
3625 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3626 to bomb out with an assertion failure - to the client this appears as a
3627 connection drop. This problem occurs in the part of the code that was taken
3628 from the Samba project. Fortunately, the assertion is in a very simple
3629 function, so I have fixed this by reproducing the function inline in the
3630 one place where it is called, and arranging for authentication to fail
3631 instead of killing the process with assert().
3633 75. The SPA client code was not working when the server requested OEM rather
3634 than Unicode encoding.
3636 76. Added code to make require_files with a specific uid setting more usable in
3637 the case where statting the file as root fails - usually a non-root-mounted
3638 NFS file system. When this happens and the failure is EACCES, Exim now
3639 forks a subprocess and does the per-uid checking as the relevant uid.
3641 77. Added process_log_path.
3643 78. If log_file_path was not explicitly set, a setting of check_log_space or
3644 check_log_inodes was ignored.
3646 79. If a space check for the spool or log partitions fails, the incident is now
3647 logged. Of course, in the latter case the data may get lost...
3649 80. Added the %p formatting code to string_format() so that it can be used to
3650 print addresses in debug_print(). Adjusted all the address printing in the
3651 debugging in store.c to use %p rather than %d.
3653 81. There was a concern that a line of code in smtp_in.c could overflow a
3654 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3655 initially expressed, the concern was not well-founded, because trailing
3656 spaces are removed early. However, if the trailing spaces were followed by
3657 a NULL, they did not get removed, so the overflow was possible. Two fixes
3660 (a) I re-wrote the offending code in a cleaner fashion.
3661 (b) If an incoming SMTP command contains a NULL character, it is rejected
3664 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3665 runs initgroups(), so that if the Exim user is in any additional groups,
3666 they will be used during message reception.
3672 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob