1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.285 2006/02/07 14:20:58 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
96 PH/01 Two changes to the default runtime configuration:
98 (1) Move the checks for relay_from_hosts and authenticated clients from
99 after to before the (commented out) DNS black list checks.
101 (2) Add control=submission to the relay_from_hosts and authenticated
102 clients checks, on the grounds that messages accepted by these
103 statements are most likely to be submissions.
105 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
107 (1) Generate an error if the third argument for the ${prvs expansion is
110 (2) Treat a missing third argument of ${prvscheck as if it were an empty
113 (3) Reset the variables that are obtained from the first argument of
114 ${prvscheck and used in the second argument before leaving the code,
115 because their memory is reclaimed, so using them afterwards may do
118 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
119 one (it's much easier than Tom thought :-).
121 (5) Because of (4), we can now allow for the use of $prvscheck_result
122 inside the third argument.
124 PH/03 For some reason, the default setting of PATH when running a command from
125 a pipe transport was just "/usr/bin". I have changed it to
128 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
129 anything to be listed in the output from -bV.
131 PH/05 When a filter generated an autoreply, the entire To: header line was
132 quoted in the delivery log line, like this:
134 => >A.N.Other <ano@some.domain> <original@ddress> ...
136 This has been changed so that it extracts the operative address. There
137 may be more than one such address. If so, they are comma-separated, like
140 => >ano@some.domain,ona@other.domain <original@ddress> ...
142 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
143 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
144 not being looked up in the rDNS to get a host name, Exim was showing the
145 IP address twice in Received: lines, even though the IP addresses were
146 identical. For example:
148 Received: from [1.2.3.4] (helo=[1.2.3.4])
150 However, if the real host name was known, it was omitting the HELO data
151 if it matched the actual IP address. This has been tidied up so that it
152 doesn't show the same IP address twice.
154 PH/07 When both +timestamp and +memory debugging was on, the value given by
155 $tod_xxx expansions could be wrong, because the tod_stamp() function was
156 called by the debug printing, thereby overwriting the timestamp buffer.
157 Debugging no longer uses the tod_stamp() function when +timestamp is set.
159 PH/08 When the original message was included in an autoreply transport, it
160 always said "this is a copy of the message, including all the headers",
161 even if body_only or headers_only was set. It now gives an appropriate
164 PH/09 Applied a patch from the Sieve maintainer which:
166 o fixes some comments
167 o adds the (disabled) notify extension core
168 o adds some debug output for the result of if/elsif tests
169 o points to the current vacation draft in the documentation
170 and documents the missing references header update
174 o fixes a bug in processing the envelope test (when testing
175 multiple envelope elements, the last element determinted the
178 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
179 Electronic Mail") by including:
181 Auto-submitted: auto-generated
183 in the messages that it generates (bounce messages and others, such as
184 warnings). In the case of bounce messages for non-SMTP mesages, there was
185 also a typo: it was using "Auto_submitted" (underscore instead of
186 hyphen). Since every message generated by Exim is necessarily in response
187 to another message, thes have all been changed to:
189 Auto-Submitted: auto-replied
191 in accordance with these statements in the RFC:
193 The auto-replied keyword:
195 - SHOULD be used on messages sent in direct response to another
196 message by an automatic process,
198 - MUST NOT be used on manually-generated messages,
200 - MAY be used on Delivery Status Notifications (DSNs) and Message
201 Disposition Notifications (MDNs),
203 - MUST NOT be used on messages generated by automatic or periodic
204 processes, except for messages which are automatic responses to
207 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
208 to the default Received: header definition.
210 PH/12 Added log selector acl_warn_skipped (default on).
212 PH/13 After a successful wildlsearch lookup, discard the values of numeric
213 variables because (a) they are in the wrong storage pool and (b) even if
214 they were copied, it wouldn't work properly because of the caching.
216 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
217 checking when decoding. Apparently there are clients that generate
218 overlong encoded strings. Why am I not surprised?
220 PH/15 If the first argument of "${if match_address" was not empty, but did not
221 contain an "@" character, Exim crashed. Now it writes a panic log message
222 and treats the condition as false.
224 PH/16 In autoreply, treat an empty string for "once" the same as unset.
226 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
227 extension "envelope-auth". The code is finished and in agreement with
228 other implementations, but there is no documentation so far and in fact,
229 nobody wrote the draft yet. This extension is currently #undef'ed, thus
230 not changing the active code.
232 Print executed "if" and "elsif" statements when debugging is used. This
233 helps a great deal to understand what a filter does.
235 Document more things not specified clearly in RFC3028. I had all this
236 sorted out, when out of a sudden new issues came to my mind. Oops."
238 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
241 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
242 canonical form (as documented). However, after a host name lookup from
243 the IP address, check_host() was doing a simple string comparison with
244 addresses acquired from the DNS when checking that the found name did
245 have the original IP as one of its addresses. Since any found IPv6
246 addresses are likely to be in abbreviated form, the comparison could
247 fail. Luckily, there already exists a function for doing the comparison
248 by converting both addresses to binary, so now that is used instead of
251 PH/20 There was another similar case to PH/19, when a complete host name was
252 given in a host list; looking up its IP address could give an abbreviated
253 form, whereas the current host's name might or might not be abbreviated.
254 The same fix has been applied.
260 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
261 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
264 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
265 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
267 PH/03 Typo: missing ".o" in src/pcre/Makefile.
269 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
270 header line, restrict the check to what is listed in RFCs 2369 and 2929.
271 Also, for "Auto-Submitted", treat anything other than "no" as
272 non-personal, in accordance with RFC 3834. (Previously it treated
273 anything starting "auto-" as non-personal.)
275 TF/01 The control=submission/name=... option had a problem with syntax
276 errors if the name included a slash character. The /name= option
277 now slurps the rest of the string, so it can include any characters
278 but it must come last in the list of options (after /sender_retain
281 PH/05 Some modifications to the interface to the fake nameserver for the new
289 TK/01 Added the "success_on_redirect" address verification option. See
290 NewStuff for rationale and an example.
292 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
294 PH/02 Patch to exigrep to allow it to work on syslog lines.
296 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
297 fread() to skip over the body file's header line, because in Cygwin the
298 header line is locked and is inaccessible.
300 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
301 co-exist for some time) to make it clear that it is the Exim ID that is
302 referenced, not the Message-ID: header line.
304 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
305 string_format() function, because snprintf() does not exist on all
308 PH/06 The use of forbid_filter_existstest now also locks out the use of the
309 ${stat: expansion item.
311 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
312 protocol synchronization error", to keep the pedants happy.
314 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
315 well as for IRIX systems, when gcc is being used. See the host.c source
318 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
320 PH/10 Named domain lists were not working if used in a queue_smtp_domains
323 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
324 transport and to the smtp transport in LMTP mode.
326 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
328 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
329 run a filter in a subprocess. This could lead to confusion in subsequent
330 lookups in the parent process. There should also be a search_tidyup() at
331 the end of the subprocess.
333 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
334 only if the host matched helo_try_verify_hosts, which caused the
335 verification to occur when the EHLO/HELO command was issued. The ACL just
336 tested the remembered result. Now, if a previous verification attempt has
337 not happened, "verify = helo" does it there and then.
339 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
341 TK/03 Fix log output including CR from clamd.
343 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
344 exist provoked a memory error which could cause a segfault.
346 PH/15 Installed PCRE 6.2
348 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
350 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
351 of the problem. Specifically, suggested +O2 rather than +O1 for the
354 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
356 PH/20 If a delivery was routed to a non-standard port by means of an SRV
357 record, the port was not correctly logged when the outgoing_port log
358 selector was set (it logged the transort's default port).
360 PH/21 Added support for host-specific ports to manualroute, queryprogram,
361 fallback_hosts, and "hosts" in the smtp transport.
363 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
364 host errors such as "Connection refused".
366 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
367 authentication with radiusclient 0.4.9:
369 - Error returned from rc_read_config was caught wrongly
370 - Username/password not passed on to radius server due to wrong length.
372 The presumption is that some radiusclient API changes for 4.51/PH/17
373 were not taken care of correctly. The code is still untested by me (my
374 Linux distribution still has 0.3.2 of radiusclient), but it was
375 contributed by a Radius user.
377 PH/24 When doing a callout, the value of $domain wasn't set correctly when
378 expanding the "port" option of the smtp transport.
380 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
381 while reading a MIME header. Thanks to Tom Hughes for a patch.
383 PH/24 Include config.h inside local_scan.h so that configuration settings are
386 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
387 that in an ACL for RCPT (for example), you can examine exactly what was
390 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
391 commands, but it was not correctly comparing the address with the actual
392 client host address. Thus, it would show the EHLO address in Received:
393 header lines when this was not necessary.
395 PH/27 Added the % operator to ${eval:}.
397 PH/28 Exim tries to create and chdir to its spool directory when it starts;
398 it should be ignoring failures (because with -C, for example, it has lost
399 privilege). It wasn't ignoring creation failures other than "already
402 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
405 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
406 because an input line was too long, either on its own, or by virtue of
407 too many continuations, the temporary file was not being removed, and the
408 return code was incorrect.
410 PH/31 Missing "BOOL" in function definition in filtertest.c.
412 PH/32 Applied Sieve patches from the maintainer.
414 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
416 PH/33 Added "verify = not_blind".
418 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
419 Local/Makefile (with some defaults set). These are used in built scripts
420 such as exicyclog, but they have never been used in the exim_install
421 script (though there are many overriding facilities there). I have
422 arranged that the exim_install script now takes note of these two
425 PH/35 Installed configuration files for Dragonfly.
427 PH/36 When a locally submitted message by a trusted user did not contain a
428 From: header, and the sender address was obtained from -f or from an SMTP
429 MAIL command, and the trusted user did not use -F to supply a sender
430 name, $originator_name was incorrectly used when constructing a From:
431 header. Furthermore, $originator_name was used for submission mode
432 messages from external hosts without From: headers in a similar way,
433 which is clearly wrong.
435 PH/37 Added control=suppress_local_fixups.
437 PH/38 When log_selector = +received_sender was set, and the addition of the
438 sender made the log line's construction buffer exactly full, or one byte
439 less than full, an overflow happened when the terminating "\n" was
442 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
443 when the result of a list match is failure because a DNS lookup failed.
445 PH/40 RM_COMMAND is now used in the building process.
447 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
448 the "build-* directories that it finds.
450 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
451 address in a domain literal was a prefix of an interface address.
453 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
454 when verifying a sender address, unless rewrite_headers is false.
456 PH/44 Wrote a long comment about why errors_to addresses are verified as
457 recipients, not senders.
459 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
460 the ratelimit ACL was added.
462 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
464 PH/46 Added extra information about PostgreSQL errors to the error string.
466 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
467 avoiding the need to install special zones in a real server. This is
468 backwards compatible; if it can't find the fake resolver, it drops back.
469 Thus, both old and new test suites can be run.
471 TF/02 Added util/ratelimit.pl
473 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
474 clock is set back in time.
476 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
477 Candler <B.Candler@pobox.com>.
479 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
480 OK for addresses that are the result of redirections.
482 PH/48 A number of further additions for the benefit of the new test suite,
483 including a fake gethostbyname() that interfaces to the fake DNS resolver
486 TF/06 The fix for widen_domains has also been applied to qualify_single and
487 search_parents which are the other dnslookup options that can cause
490 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
493 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
495 TF/07 Exim produced the error message "an SRV record indicated no SMTP
496 service" if it encountered an MX record with an empty target hostname.
497 The message is now "an MX or SRV record indicated no SMTP service".
499 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
500 if the DNS of the sending site is misconfigured. This is quite a
501 common situation. This change restores the behaviour of treating a
502 helo verification defer as a failure.
504 PH/51 If self=fail was set on a router, the bounce message did not include the
505 actual error message.
511 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
513 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
514 command itself ended in error, the underlying message about the transport
515 filter timeout was being overwritten with the pipe command error. Now the
516 underlying error message should be appended to the second error message.
518 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
520 PH/02 Reduce the amount of output that "make" produces by default. Full output
521 can still be requested.
523 PH/03 The warning log line about a condition test deferring for a "warn" verb
524 was being output only once per connection, rather than after each
525 occurrence (because it was using the same function as for successful
526 "warn" verbs). This seems wrong, so I have changed it.
528 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
529 it should not have, which might have caused a crash in the right
530 circumstances, but probably never did.
532 PH/04 Installed a modified version of Tony Finch's patch to make submission
533 mode fix the return path as well as the Sender: header line, and to
534 add a /name= option so that you can make the user's friendly name appear
537 TF/03 Added the control = fakedefer ACL modifier.
539 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
540 Mark Lowes for thorough testing.
542 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
544 TK/03 Merged latest SRS patch from Miles Wilton.
546 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
547 with the definition in sysexits.h (which is #included earlier).
548 Fortunately, Exim does not actually use EX_OK. The code used to try to
549 preserve the sysexits.h value, by assumimg that macro definitions were
550 scanned for macro replacements. I have been disabused of this notion,
551 so now the code just undefines EX_OK before #including unistd.h.
553 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
554 in the smtp transport. When a block could not be written in a single
555 write() function, the timeout was being re-applied to each part-write.
556 This seems wrong - if the receiver was accepting one byte at a time it
557 would take for ever. The timeout is now adjusted when this happens. It
558 doesn't have to be particularly precise.
560 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
561 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
563 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
564 without a domain if the check to <postmaster@domain> fails.
566 SC/01 Eximstats: added -xls and the ability to specify output files
567 (patch written by Frank Heydlauf).
569 SC/02 Eximstats: use FileHandles for outputing results.
571 SC/03 Eximstats: allow any combination of xls, txt, and html output.
573 SC/04 Eximstats: fixed display of large numbers with -nvr option
575 SC/05 Eximstats: fixed merging of reports with empty tables.
577 SC/06 Eximstats: added the -include_original_destination flag
579 SC/07 Eximstats: removed tabs and trailing whitespace.
581 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
583 TK/06 MBOX spool code: Add real "From " MBOX separator line
584 so the .eml file is really in mbox format (even though
585 most programs do not really care). Patch from Alex Miller.
587 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
588 The latter is generated from $received_to and is only set if the
589 message has one envelope recipient. SA can use these headers,
590 obviously out-of-the-box. Patch from Alex Miller.
592 PH/08 The ${def test on a variable was returning false if the variable's
593 value was "0", contrary to what the specification has always said!
594 The result should be true unless the variable is empty.
596 PH/09 The syntax error of a character other than { following "${if
597 def:variable_name" (after optional whitespace) was not being diagnosed.
598 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
599 accidental colon was present, for example, could give incorrect results.
601 PH/10 Tidied the code in a number of places where the st_size field of a stat()
602 result is used (not including appendfile, where other changes are about
605 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
606 This involved changing a lot of size variables from int to off_t. It
607 should work with maildirs and everything.
609 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
610 spamd dying while we are connected to it.
612 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
615 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
616 function for BDB 4.3.
618 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
620 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
621 the sources for PCRE have been re-arranged and more files are now
624 PH/15 The code I had for printing potentially long long variables in PH/11
625 above was not the best (it lost precision). The length of off_t variables
626 is now inspected at build time, and an appropriate printing format (%ld
627 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
628 to be "long long int" or "long int". This is needed for the internal
629 formatting function string_vformat().
631 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
632 the configuration file to be ":syslog", then the script "guesses" where
633 the logs files are, rather than using the compiled in default. In our
634 case the guess is not the same as the compiled default, so the script
635 suddenly stopped working when I started to use syslog. The patch checks
636 to see if log_file_path is "". If so, it attempts to read it from exim
637 with no configuration file to get the compiled in version, before it
638 falls back to the previous guessing code."
640 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
641 implementing BATV in an Exim configuration. See NewStuff for the gory
644 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
645 Makefile that are specific to HP-UX.
647 PH/18 If the "use_postmaster" option was set for a recipient callout together
648 with the "random" option, the postmaster address was used as the MAIL
649 FROM address for the random test, but not for the subsequent recipient
650 test. It is now used for both.
652 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
653 patch removes a few documentation additions to RFC 3028, because the
654 latest draft now contains them. It adds the new en;ascii-case comparator
655 and a new error check for 8bit text in MIME parts. Comparator and
656 require names are now matched exactly. I enabled the subaddress
657 extension, but it is not well tested yet (read: it works for me)."
659 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
660 rework some of the code of TK/09 above to avoid the hardwired use of
661 "%lld" and "long long". Replaced the call to snprintf() with a call to
664 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
665 records point to non-existent hosts", "retry timeout exceeded", and
666 "retry time not reached for any host after a long failure period".
668 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
669 experimental DomainKeys support:
671 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
672 (2) On an error such as an illegally used "control", the wrong name for
673 the control was given.
675 These problems did NOT occur unless DomainKeys support was compiled.
677 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
679 PH/24 Added ${if match_ip condition.
681 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
682 databases so that it will be absolutely obvious if a crash occurs in the
683 DB library. This is a regular occurrence (often caused by mis-matched
686 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
687 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
688 picked up on a user's system that detects such things. There doesn't seem
689 to be a gcc warning option for this - only an attribute that has to be
690 put on the function's prototype. It seems that in Fedora Core 4 they have
691 set this on a number of new functions. No doubt there will be more in due
694 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
695 specify a transport. However, if an address that was verified by such a
696 router was the subject of a callout, Exim crashed because it tried to
697 read the rcpt_include_affixes from the non-existent transport. Now it
698 just assumes that the setting of that option is false. This bug was
699 introduced by 4.51/PH/31.
701 PH/28 Changed -d+all to exclude +memory, because that information is very
702 rarely of interest, but it makes the output a lot bigger. People tend to
703 do -d+all out of habit.
705 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
706 code in os-type was giving problems when libc.so lives in lib64, like on
709 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
710 aren't the modern standard, and it seems that some systems' include files
711 don't always have them. Exim was already checking for some of the newer
712 ones like T_AAAA, and defining it itself. I've added checks for all the
713 record types that Exim uses.
715 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
716 not automatically generating a new one, as it is supposed to. This
717 prevented TLS from working. If the file did exist, but contained invalid
718 data, a new version was generated, as expected. It was only the case of a
719 non-existent file that was broken.
721 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
722 with a change in libdomainkeys > 0.64.
724 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
725 from DNS. If the selector record carries the flag, it now has
726 precedence over the domain-wide flag.
728 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
730 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
731 the use of a port name that isn't defined in /etc/services) occurred, the
732 message was deferred as in a normal delivery, and thus remained on the
733 spool, instead of being failed because of the mua_wrapper setting. This
734 is now fixed, and I tidied up some of the mua_wrapper messages at the
737 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
738 the messages in a hash of arrays rather than using individual hashes.
739 This is a bit cleaner and results in dramatic memory savings, albeit
740 at a slight CPU cost.
742 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
743 as requested by Marc Sherman.
745 SC/10 Eximstats: added histograms for user specified patterns as requested
748 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
750 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
751 fopen() in the content-scanning modules that did not already have it.
757 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
758 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
760 TK/02 Fix ACL "control" statement not being available in MIME ACL.
762 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
764 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
765 to test Sieve filters that use "vacation".
767 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
768 that changes the way the GnuTLS parameters are stored in the cache file.
769 The new format can be generated externally. For backward compatibility,
770 if the data in the cache doesn't make sense, Exim assumes it has read an
771 old-format file, and it generates new data and writes a new file. This
772 means that you can't go back to an older release without removing the
775 PH/03 A redirect router that has both "unseen" and "one_time" set does not
776 work if there are any delivery delays because "one_time" forces the
777 parent to be marked "delivered", so its unseen clone is never tried
778 again. For this reason, Exim now forbids the simultaneous setting of
781 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
782 redirected to themselves ("homonym" addresses). Read the long ChangeLog
783 entry if you want to know the details. The fix, however, neglected to
784 consider the case when local delivery batching is involved. The test for
785 "previously delivered" was not happening when checking to see if an
786 address could be batched with a previous (undelivered) one; under
787 certain circumstances this could lead to multiple deliveries to the same
790 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
791 in its include files, and this causes problems building Exim.
793 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
794 header_syntax) but Exim was just ignoring anything given after a slash.
795 In particular, this caused confusion with an attempt to use "verify =
796 reverse_host_lookup/defer_ok". An error is now given when options are
797 supplied for verify items that do not have them. (Maybe reverse_host_
798 lookup should have a defer_ok option, but that's a different point.)
800 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
801 defined by RFC 821) to 2048, because there were problems with some AUTH
802 commands, and RFC 1869 says the size should be increased for extended
803 SMTP commands that take arguments.
805 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
808 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
809 "unknown" error; now it says that the functionality isn't in the binary.
811 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
812 an address' error message when a string expansion fails (syntax or
813 whatever). Otherwise the password may appear in the log. Following change
814 PH/42 below, there is no longer a chance of it appearing in a bounce
817 PH/11 Installed exipick version 20050225.0 from John Jetmore.
819 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
820 of its addresses was ever tried. (Bugzilla bug #2.)
822 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
823 the result incorrectly in the debug output. (It correctly added a newline
824 to what was transported.)
826 TF/01 Added $received_time.
828 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
829 commented out examples of how to interface to a virus scanner and to
830 SpamAssassin. Also added commented examples of av_scanner and
831 spamd_address settings.
833 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
834 and controls are allowed in which ACLs. There were a couple of minor
835 errors. Some of the entries in the conditions table (which is a table of
836 where they are NOT allowed) were getting very unwieldy; rewrote them as a
837 negation of where the condition IS allowed.
839 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
841 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
842 header file does not have a version number, so I've had to invent a new
843 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
844 API. The code is untested by me (my Linux distribution still has 0.3.2 of
845 radiusclient), but it was contributed by a Radius user.
847 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
848 files or directories, for OpenSSL.
850 PH/19 When an Exim process that is running as root has to create an Exim log
851 file, it does so in a subprocess that runs as exim:exim so as to get the
852 ownership right at creation (otherwise, other Exim processes might see
853 the file with the wrong ownership). There was no test for failure of this
854 fork() call, which would lead to the process getting stuck as it waited
855 for a non-existent subprocess. Forks do occasionally fail when resources
856 run out. I reviewed all the other calls to fork(); they all seem to check
859 PH/20 When checking for unexpected SMTP input at connect time (before writing
860 the banner), Exim was not dealing correctly with a non-positive return
861 from the read() function. If the client had disconnected by this time,
862 the result was a log entry for a synchronization error with an empty
863 string after "input=" when read() returned zero. If read() returned -1
864 (an event I could not check), uninitialized data bytes were printed.
865 There were reports of junk text (parts of files, etc) appearing after
868 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
870 PH/22 Added support for macro redefinition, and (re)definition in between
871 driver and ACL definitions.
873 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
874 forgetting to use the resulting value; it was using the unexpanded value.
876 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
877 hadn't been configured. The fix is from Juergen Kreileder, who
878 understands it better than I do:
880 "Here's what I see happening with three configured cyrus_sasl
881 authenticators configured (plain, login, cram-md5):
883 On startup auth_cyrus_sasl_init() gets called for each of these.
884 This means three calls to sasl_listmech() without a specified mech_list.
885 => SASL tests which mechs of all available mechs actually work
886 => three warnings about OTP not working
887 => the returned list contains: plain, login, cram-md5, digest-md5, ...
889 With the patch, sasl_listmech() also gets called three times. But now
890 SASL's mech_list option is set to the server_mech specified in the the
891 authenticator. Or in other words, the answer from sasl_listmech()
892 gets limited to just the mech you're testing for (which is different
894 => the return list contains just 'plain' or 'login', 'cram-md5' or
895 nothing depending on the value of ob->server_mech.
897 I've just tested the patch: Authentication still works fine,
898 unavailable mechs specified in the exim configuration are still
899 caught, and the auth.log warnings about OTP are gone."
901 PH/25 When debugging is enabled, the contents of the command line are added
902 to the debugging output, even when log_selector=+arguments is not
905 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
906 answer is "GNU", and only if the return is "GNU/something" is the answer
909 PH/27 $acl_verify_message is now set immediately after the failure of a
910 verification in an ACL, and so is available in subsequent modifiers. In
911 particular, the message can be preserved by coding like this:
913 warn !verify = sender
914 set acl_m0 = $acl_verify_message
916 Previously, $acl_verify_message was set only while expanding "message"
917 and "log_message" when a very denied access.
919 PH/28 Modified OS/os.c-Linux with
921 -#ifndef OS_LOAD_AVERAGE
922 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
924 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
925 nomenclature these days.)
927 PH/29 Installed patch from the Sieve maintainer that adds the options
928 sieve_useraddress and sieve_subaddress to the redirect router.
930 PH/30 In these circumstances:
931 . Two addresses routed to the same list of hosts;
932 . First host does not offer TLS;
933 . First host accepts first address;
934 . First host gives temporary error to second address;
935 . Second host offers TLS and a TLS session is established;
936 . Second host accepts second address.
937 Exim incorrectly logged both deliveries with the TLS parameters (cipher
938 and peerdn, if requested) that were in fact used only for the second
941 PH/31 When doing a callout as part of verifying an address, Exim was not paying
942 attention to any local part prefix or suffix that was matched by the
943 router that accepted the address. It now behaves in the same way as it
944 does for delivery: the affixes are removed from the local part unless
945 rcpt_include_affixes is set on the transport.
947 PH/32 Add the sender address, as F=<...>, to the log line when logging a
948 timeout during the DATA phase of an incoming message.
950 PH/33 Sieve envelope tests were broken for match types other than :is. I have
951 applied a patch sanctioned by the Sieve maintainer.
953 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
954 the uid or gid is negative. A case of a negative gid caused this to be
955 noticed. The fix allows for either to be negative.
957 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
958 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
961 PH/36 The change PH/12 above was broken. Fixed it.
963 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
964 the grounds that routing the same address twice would always produce the
965 same answer. This might have been true once, but it is certainly no
966 longer true now. Routing a child address may depend on the previous
967 routing that produced that child. Some complicated redirection strategies
968 went wrong when messages had multiple recipients, and made Exim's
969 behaviour dependent on the order in which the addresses were given.
971 I have moved the duplicate checking until after the routing is complete.
972 Exim scans the addresses that are assigned to local and remote
973 transports, and removes any duplicates. This means that more work will be
974 done, as duplicates will always all be routed, but duplicates are
975 presumably rare, so I don't expect this is of any significance.
977 For deliveries to pipes, files, and autoreplies, the duplicate checking
978 still happens during the routing process, since they are not going to be
981 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
982 It corrects a timeout issue with spamd. This is Ian's comment: "The
983 background is that sometimes spamd either never reads data from a
984 connection it has accepted, or it never writes response data. The exiscan
985 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
986 blindly assumes that writes won't block so it may never time out."
988 PH/39 Allow G after quota size as well as K and M.
990 PH/40 The value set for $authenticated_id in an authenticator may not contain
991 binary zeroes or newlines because the value is written to log lines and
992 to spool files. There was no check on this. Now the value is run through
993 the string_printing() function so that such characters are converted to
994 printable escape sequences.
996 PH/41 $message_linecount is a new variable that contains the total number of
997 lines in the message. Compare $body_linecount, which is the count for the
1000 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1001 bounce and delay warning messages, except in certain special cases, which
1004 (a) An SMTP error message from a remote host;
1005 (b) A message specified in a :fail: redirection;
1006 (c) A message specified in a "fail" command in a system filter;
1007 (d) A message specified in a FAIL return from the queryprogram router;
1008 (e) A message specified by the cannot_route_message router option.
1010 In these cases only, Exim does include the error details in bounce and
1011 warning messages. There are also a few cases where bland messages such
1012 as "unrouteable address" or "local delivery error" are given.
1014 PH/43 $value is now also set for the "else" part of a ${run expansion.
1016 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1017 being worked on, but at least Exim now implements the latest version to
1020 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1021 process to complete was treated as a delivery failure, a timeout while
1022 writing the message to the pipe was logged, but erroneously treated as a
1023 successful delivery. Such timeouts include transport filter timeouts. For
1024 consistency with the overall process timeout, these timeouts are now
1025 treated as errors, giving rise to delivery failures by default. However,
1026 there is now a new Boolean option for the pipe transport called
1027 timeout_defer, which, if set TRUE, converts the failures into defers for
1028 both kinds of timeout. A transport filter timeout is now identified in
1031 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1032 systems where "make" and "gmake" are different, calling "gmake" at top
1033 level broke things. I've arranged for the value of $(MAKE) to be passed
1034 from the Makefile to this script so that it can call the same version of
1038 A note about Exim versions 4.44 and 4.50
1039 ----------------------------------------
1041 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1042 changes of various kinds. As a consequence, a big documentation update was
1043 needed. This delayed the release for rather longer than seemed good, especially
1044 in the light of a couple of (minor) security issues. Therefore, the changes
1045 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1046 release. So 4.44 and 4.50 are in effect two different branches that both start
1049 I have left the 4.50 change log unchanged; it contains all the changes since
1050 4.43. The change log for 4.44 is below; many of its items are identical to
1051 those for 4.50. This seems to be the most sensible way to preserve the
1052 historical information.
1058 1. Minor wording change to the doc/README.SIEVE file.
1060 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1061 computation of the current number of files was incorrect.
1063 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1064 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1067 4. Give more explanation in the error message when the command for a transport
1068 filter fails to execute.
1070 5. There are several places where Exim runs a non-Exim command in a
1071 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1072 was being done only for the command run by the queryprogram router. It is
1073 now done for all such subprocesses. The other cases are: ${run, transport
1074 filters, and the commands run by the lmtp and pipe transports.
1076 6. Added CONFIGURE_GROUP build-time option.
1078 7. Some older OS have a limit of 256 on the maximum number of file
1079 descriptors. Exim was using setrlimit() to set 1000 as a large value
1080 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1081 systems. I've change it so that if it can't get 1000, it tries for 256.
1083 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1084 was an oversight, and furthermore, ever since the addition of extra
1085 controls (e.g. 4.43/32), the checks on when to allow different forms of
1086 "control" were broken. There should now be diagnostics for all cases when a
1087 control that does not make sense is encountered.
1089 9. Added the /retain_sender option to "control=submission".
1091 10. $recipients is now available in the predata ACL (oversight).
1093 11. Tidy the search cache before the fork to do a delivery from a message
1094 received from the command line. Otherwise the child will trigger a lookup
1095 failure and thereby defer the delivery if it tries to use (for example) a
1096 cached ldap connection that the parent has called unbind on.
1098 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1099 of $address_data from the recipient verification was clobbered by the
1100 sender verification.
1102 13. The value of address_data from a sender verification is now available in
1103 $sender_address_data in subsequent conditions in the ACL statement.
1105 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1107 15. Added a new option "connect=<time>" to callout options, to set a different
1110 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1111 was its contents. (It was OK if the option was not defined at all.)
1113 17. A "Completed" log line is now written for messages that are removed from
1114 the spool by the -Mrm option.
1116 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1117 information about exactly what failed.
1119 19. Added -dd to debug only the daemon process.
1121 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1122 handles timeouts, both on the server side and network timeouts. Renamed the
1123 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1125 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1126 It is now set to "smtps".
1128 22. $host_address is now set to the target address during the checking of
1129 ignore_target_hosts.
1131 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1132 being passed; this would have caused $sender_host_name to have been used if
1133 matching the list had actually called for a host name (not very likely,
1134 since this list is usually IP addresses). A host name is now passed as
1137 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1138 code that specifies a non-restarting handler (typically sigaction() in
1139 modern systems) in an attempt to fix a rare and obscure crash bug.
1141 25. Narrowed the window for a race in the daemon that could cause it to ignore
1142 SIGCHLD signals. This is not a major problem, because they are used only to
1143 wake it up if nothing else does.
1145 26. A malformed maildirsize file could cause Exim to calculate negative values
1146 for the mailbox size or file count. Odd effects could occur as a result.
1147 The maildirsize information is now recalculated if the size or filecount
1150 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1151 support for a long time. Removed HAVE_SYS_VFS_H.
1153 28. Installed the latest version of exipick from John Jetmore.
1155 29. In an address list, if the pattern was not a regular expression, an empty
1156 subject address (from a bounce message) matched only if the pattern was an
1157 empty string. Non-empty patterns were not even tested. This was the wrong
1158 because it is perfectly reasonable to use an empty address as part of a
1159 database query. An empty address is now tested by patterns that are
1160 lookups. However, all the other forms of pattern expect the subject to
1161 contain a local part and a domain, and therefore, for them, an empty
1162 address still always fails if the pattern is not itself empty.
1164 30. Exim went into a mad DNS loop when attempting to do a callout where the
1165 host was specified on an smtp transport, and looking it up yielded more
1166 than one IP address.
1168 31. Re-factored the code for checking spool and log partition space into a
1169 function that finds that data and another that does the check. The former
1170 is then used to implement four new variables: $spool_space, $log_space,
1171 $spool_inodes, and $log_inodes.
1173 32. The RFC2047 encoding function was originally intended for short strings
1174 such as real names; it was not keeping to the 75-character limit for
1175 encoded words that the RFC imposes. It now respects the limit, and
1176 generates multiple encoded words if necessary. To be on the safe side, I
1177 have increased the buffer size for the ${rfc2047: expansion operator from
1180 33. It is now permitted to omit both strings after an "if" condition; if the
1181 condition is true, the result is "true". As before, when the second string
1182 is omitted, a false condition yields an empty string. This makes it less
1183 cumbersome to write custom ACL and router conditions.
1185 34. Failure to deliver a bounce message always caused it to be frozen, even if
1186 there was an errors_to setting on the router. The errors_to setting is now
1189 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1190 canonical form (fully expanded) before being placed in
1191 $sender_host_address.
1193 36. The table in the code that translates DNS record types into text (T_A to
1194 "A" for instance) was missing entries for NS and CNAME. It is just possible
1195 that this could have caused confusion if both these types were looked up
1196 for the same domain, because the text type is used as part of Exim's
1197 per-process caching. But the chance of anyone hitting this buglet seems
1200 37. The dnsdb lookup has been extended in a number of ways.
1202 (1) There is a new type, "zns", which walks up the domain tree until it
1203 finds some nameserver records. It should be used with care.
1205 (2) There is a new type, "mxh", which is like "mx" except that it returns
1206 just the host names, not the priorities.
1208 (3) It is now possible to give a list of domains (or IP addresses) to be
1209 looked up. The behaviour when one of the lookups defers can be
1210 controlled by a keyword.
1212 (4) It is now possible to specify the separator character for use when
1213 multiple records are returned.
1215 38. The dnslists ACL condition has been extended: it is now possible to supply
1216 a list of IP addresses and/or domains to be looked up in a particular DNS
1219 39. Added log_selector=+queue_time_overall.
1221 40. When running the queue in the test harness, wait just a tad after forking a
1222 delivery process, to get repeatability of debugging output.
1224 41. Include certificate and key file names in error message when GnuTLS fails
1225 to set them up, because the GnuTLS error message doesn't include the name
1226 of the failing file when there is a problem reading it.
1228 42. Allow both -bf and -bF in the same test run.
1230 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1232 44. The "Exiscan patch" is now merged into the mainline Exim source.
1234 45. Sometimes the final signoff response after QUIT could fail to get
1235 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1236 < 0 before doing a fflush(). This bug looks as though it goes back to the
1237 introduction of TLS in release 3.20, but "sometimes" must have been rare
1238 because the tests only now provoked it.
1240 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1241 (this can affect the format of dates).
1243 47. exim_tidydb, when checking for the continued existence of a message for
1244 which it has found a message-specific retry record, was not finding
1245 messages that were in split spool directories. Consequently, it was
1246 deleting retry records that should have stayed in existence.
1248 48. Steve fixed some bugs in eximstats.
1250 49. The SPA authentication driver was not abandoning authentication and moving
1251 on to the next authenticator when an expansion was forced to fail,
1252 contradicting the general specification for all authenticators. Instead it
1253 was generating a temporary error. It now behaves as specified.
1255 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1256 (the order specifies the preference for clients). The order is now AES256,
1257 AES128, 3DES, ARCFOUR128.
1259 51. Small patch to Sieve code - explicitly set From: when generating an
1262 52. Exim crashed if a remote delivery caused a very long error message to be
1263 recorded - for instance if somebody sent an entire SpamAssassin report back
1264 as a large number of 550 error lines. This bug was coincidentally fixed by
1265 increasing the size of one of Exim's internal buffers (big_buffer) that
1266 happened as part of the Exiscan merge. However, to be on the safe side, I
1267 have made the code more robust (and fixed the comments that describe what
1270 53. Now that there can be additional text after "Completed" in log lines (if
1271 the queue_time_overall log selector is set), a one-byte patch to exigrep
1272 was needed to allow it to recognize "Completed" as not the last thing in
1275 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1276 patch that reportedly fixes this has been added. I am not expert enough to
1277 create a test for it. This is what the patch creator wrote:
1279 "I found a little strange behaviour of ldap code when working with
1280 Windows 2003 AD Domain, where users was placed in more than one
1281 Organization Units. When I tried to give exim partial DN, the exit code
1282 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1283 But simultaneously result of request was absolutely normal ldap result,
1284 so I produce this patch..."
1286 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1287 REFERENCE, so I have modified the code to exclude the patch when that macro
1290 55. Some experimental protocols are using DNS PTR records for new purposes. The
1291 keys for these records are domain names, not reversed IP addresses. The
1292 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1293 leaves it alone. Component reversal etc. now happens only for IP addresses.
1296 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1298 57. Double the size of the debug message buffer (to 2048) so that more of very
1299 long debug lines gets shown.
1301 58. The exicyclog utility now does better if the number of log files to keep
1302 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1304 59. Two changes related to the smtp_active_hostname option:
1306 (1) $smtp_active_hostname is now available as a variable.
1307 (2) The default for smtp_banner uses $smtp_active_hostname instead
1308 of $primary_hostname.
1310 60. The host_aton() function is supposed to be passed a string that is known
1311 to be a valid IP address. However, in the case of IPv6 addresses, it was
1312 not checking this. This is a hostage to fortune. Exim now panics and dies
1313 if the condition is not met. A case was found where this could be provoked
1314 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1315 components; fortuitously, this particular loophole had already been fixed
1316 by change 4.50/55 above.
1318 If there are any other similar loopholes, the new check in host_aton()
1319 itself should stop them being exploited. The report I received stated that
1320 data on the command line could provoke the exploit when Exim was running as
1321 exim, but did not say which command line option was involved. All I could
1322 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1323 running as the user.
1326 61. There was a buffer overflow vulnerability in the SPA authentication code
1327 (which came originally from the Samba project). I have added a test to the
1328 spa_base64_to_bits() function which I hope fixes it.
1331 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1332 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1334 63. The daemon start-up calls getloadavg() while still root for those OS that
1335 need the first call to be done as root, but it missed one case: when
1336 deliver_queue_load_max is set with deliver_drop_privilege. This is
1337 necessary for the benefit of the queue runner, because there is no re-exec
1338 when deliver_drop_privilege is set.
1340 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1341 This has been fixed.
1343 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1344 was in use, was not putting the data itself into the right store pool;
1345 consequently, it could be overwritten for a subsequent message in the same
1346 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1349 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1351 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1352 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1353 the function were treating the return as a boolean value, which happened to
1354 work because 0=false and not-0=true, but is not correct code.
1356 68. The host_aton() function was not handling scoped IPv6 addresses (those
1357 with, for example, "%eth0" on the end) correctly.
1359 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1360 negated items (that is, ~something) in unsigned ints. Some compilers
1361 apparently mutter when there is no cast.
1363 70. If an address verification called from an ACL failed, and did not produce a
1364 user-specific message (i.e. there was only a "system" message), nothing was
1365 put in $acl_verify_message. In this situation, it now puts the system
1368 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1369 session; change 4.31/43 added the unwanted input to the log line - except
1370 that it did not do this in the start of session case. It now does.
1372 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1373 This is wrong and can cause the other end to generate a synchronization
1374 error if it is another Exim or anything else that does the synchronization
1375 check. A QUIT command is no longer sent after a timeout.
1377 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1378 during host lookups.
1380 74. The defer_ok option of callout verification was not working if it was used
1381 when verifying addresses in header lines, that is, for this case:
1383 verify = header_sender/callout=defer_ok
1385 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1386 those file descriptors could be used for SMTP connections. If anything
1387 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1388 could be sent to the SMTP client, causing chaos. The daemon now opens
1389 stdin, stdout, and stderr to /dev/null when it puts itself into the
1392 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1393 log by default. The user can override this with suitable Perl magic.
1395 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1396 the log message when discard triggers, was not working for the DATA ACL or
1397 for the non-SMTP ACL.
1399 78. Error message wording change in sieve.c.
1401 79. If smtp_accept_max_per_host was set, the number of connections could be
1402 restricted to fewer than expected, because the daemon was trying to set up
1403 a new connection before checking whether the processes handling previous
1404 connections had finished. The check for completed processes is now done
1405 earlier. On busy systems, this bug wouldn't be noticed because something
1406 else would have woken the daemon, and it would have reaped the completed
1409 80. If a message was submitted locally by a user whose login name contained one
1410 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1411 It caused a spool format error. I have fixed the spool reading code. A
1412 related problem was that the "from" clause in the Received: line became
1413 illegal because of the space(s). It is now covered by ${quote_local_part.
1415 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1418 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1420 83. Patch from Sieve maintainer to fix unterminated string problem in
1421 "vacation" handling.
1423 84. Some minor changes to the Linux configuration files to help with other
1424 OS variants using glibc.
1426 85. One more patch for Sieve to update vacation handling to latest spec.
1429 ----------------------------------------------------
1430 See the note above about the 4.44 and 4.50 releases.
1431 ----------------------------------------------------
1437 1. Change 4.43/35 introduced a bug that caused file counts to be
1438 incorrectly computed when quota_filecount was set in an appendfile
1441 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1442 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1445 3. Give more explanation in the error message when the command for a transport
1446 filter fails to execute.
1448 4. There are several places where Exim runs a non-Exim command in a
1449 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1450 was being done only for the command run by the queryprogram router. It is
1451 now done for all such subprocesses. The other cases are: ${run, transport
1452 filters, and the commands run by the lmtp and pipe transports.
1454 5. Some older OS have a limit of 256 on the maximum number of file
1455 descriptors. Exim was using setrlimit() to set 1000 as a large value
1456 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1457 systems. I've change it so that if it can't get 1000, it tries for 256.
1459 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1460 was an oversight, and furthermore, ever since the addition of extra
1461 controls (e.g. 4.43/32), the checks on when to allow different forms of
1462 "control" were broken. There should now be diagnostics for all cases when a
1463 control that does not make sense is encountered.
1465 7. $recipients is now available in the predata ACL (oversight).
1467 8. Tidy the search cache before the fork to do a delivery from a message
1468 received from the command line. Otherwise the child will trigger a lookup
1469 failure and thereby defer the delivery if it tries to use (for example) a
1470 cached ldap connection that the parent has called unbind on.
1472 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1473 of $address_data from the recipient verification was clobbered by the
1474 sender verification.
1476 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1477 was its contents. (It was OK if the option was not defined at all.)
1479 11. A "Completed" log line is now written for messages that are removed from
1480 the spool by the -Mrm option.
1482 12. $host_address is now set to the target address during the checking of
1483 ignore_target_hosts.
1485 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1486 being passed; this would have caused $sender_host_name to have been used if
1487 matching the list had actually called for a host name (not very likely,
1488 since this list is usually IP addresses). A host name is now passed as
1491 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1492 code that specifies a non-restarting handler (typically sigaction() in
1493 modern systems) in an attempt to fix a rare and obscure crash bug.
1495 15. Narrowed the window for a race in the daemon that could cause it to ignore
1496 SIGCHLD signals. This is not a major problem, because they are used only to
1497 wake it up if nothing else does.
1499 16. A malformed maildirsize file could cause Exim to calculate negative values
1500 for the mailbox size or file count. Odd effects could occur as a result.
1501 The maildirsize information is now recalculated if the size or filecount
1504 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1505 support for a long time. Removed HAVE_SYS_VFS_H.
1507 18. Updated exipick to current release from John Jetmore.
1509 19. Allow an empty sender to be matched against a lookup in an address list.
1510 Previously the only cases considered were a regular expression, or an
1513 20. Exim went into a mad DNS lookup loop when doing a callout where the
1514 host was specified on the transport, if the DNS lookup yielded more than
1517 21. The RFC2047 encoding function was originally intended for short strings
1518 such as real names; it was not keeping to the 75-character limit for
1519 encoded words that the RFC imposes. It now respects the limit, and
1520 generates multiple encoded words if necessary. To be on the safe side, I
1521 have increased the buffer size for the ${rfc2047: expansion operator from
1524 22. Failure to deliver a bounce message always caused it to be frozen, even if
1525 there was an errors_to setting on the router. The errors_to setting is now
1528 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1529 canonical form (fully expanded) before being placed in
1530 $sender_host_address.
1532 24. Updated eximstats to version 1.33
1534 25. Include certificate and key file names in error message when GnuTLS fails
1535 to set them up, because the GnuTLS error message doesn't include the name
1536 of the failing file when there is a problem reading it.
1538 26. Expand error message when OpenSSL has problems setting up cert/key files.
1541 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1542 (this can affect the format of dates).
1544 28. exim_tidydb, when checking for the continued existence of a message for
1545 which it has found a message-specific retry record, was not finding
1546 messages that were in split spool directories. Consequently, it was
1547 deleting retry records that should have stayed in existence.
1549 29. eximstats updated to version 1.35
1550 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1551 1.35 - bugfix such that pie charts by volume are generated correctly
1553 30. The SPA authentication driver was not abandoning authentication and moving
1554 on to the next authenticator when an expansion was forced to fail,
1555 contradicting the general specification for all authenticators. Instead it
1556 was generating a temporary error. It now behaves as specified.
1558 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1559 (the order specifies the preference for clients). The order is now AES256,
1560 AES128, 3DES, ARCFOUR128.
1562 31. Small patch to Sieve code - explicitly set From: when generating an
1565 32. Exim crashed if a remote delivery caused a very long error message to be
1566 recorded - for instance if somebody sent an entire SpamAssassin report back
1567 as a large number of 550 error lines. This bug was coincidentally fixed by
1568 increasing the size of one of Exim's internal buffers (big_buffer) that
1569 happened as part of the Exiscan merge. However, to be on the safe side, I
1570 have made the code more robust (and fixed the comments that describe what
1573 33. Some experimental protocols are using DNS PTR records for new purposes. The
1574 keys for these records are domain names, not reversed IP addresses. The
1575 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1576 leaves it alone. Component reversal etc. now happens only for IP addresses.
1579 34. The host_aton() function is supposed to be passed a string that is known
1580 to be a valid IP address. However, in the case of IPv6 addresses, it was
1581 not checking this. This is a hostage to fortune. Exim now panics and dies
1582 if the condition is not met. A case was found where this could be provoked
1583 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1584 components; fortuitously, this particular loophole had already been fixed
1585 by change 4.50/55 or 4.44/33 above.
1587 If there are any other similar loopholes, the new check in host_aton()
1588 itself should stop them being exploited. The report I received stated that
1589 data on the command line could provoke the exploit when Exim was running as
1590 exim, but did not say which command line option was involved. All I could
1591 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1592 running as the user.
1595 35. There was a buffer overflow vulnerability in the SPA authentication code
1596 (which came originally from the Samba project). I have added a test to the
1597 spa_base64_to_bits() function which I hope fixes it.
1600 36. The daemon start-up calls getloadavg() while still root for those OS that
1601 need the first call to be done as root, but it missed one case: when
1602 deliver_queue_load_max is set with deliver_drop_privilege. This is
1603 necessary for the benefit of the queue runner, because there is no re-exec
1604 when deliver_drop_privilege is set.
1606 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1607 was in use, was not putting the data itself into the right store pool;
1608 consequently, it could be overwritten for a subsequent message in the same
1609 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1612 38. Sometimes the final signoff response after QUIT could fail to get
1613 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1614 < 0 before doing a fflush(). This bug looks as though it goes back to the
1615 introduction of TLS in release 3.20, but "sometimes" must have been rare
1616 because the tests only now provoked it.
1622 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1623 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1624 Now it returns an integer. A number of calls were still expecting a T/F
1625 return. Fortuitously, in all cases, the tests worked in OK situations,
1626 which is the norm. However, things would have gone wrong on any write
1627 failures on the smtp file descriptor. This function is used when sending
1628 messages over SMTP and also when doing verify callouts.
1630 2. When Exim is called to do synchronous delivery of a locally submitted
1631 message (the -odf or -odi options), it no longer closes stderr before doing
1634 3. Implemented the mua_wrapper option.
1636 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1638 5. Implemented the functions header_remove(), header_testname(),
1639 header_add_at_position(), and receive_remove_recipient(), and exported them
1642 6. If an ACL "warn" statement specified the addition of headers, Exim already
1643 inserted X-ACL-Warn: at the start if there was no header name. However, it
1644 was not making this test for the second and subsequent header lines if
1645 there were newlines in the string. This meant that an invalid header could
1646 be inserted if Exim was badly configured.
1648 7. Allow an ACL "warn" statement to add header lines at the start or after all
1649 the Received: headers, as well as at the end.
1651 8. Added the rcpt_4xx retry error code.
1653 9. Added postmaster_mailfrom=xxx to callout verification option.
1655 10. Added mailfrom=xxxx to the callout verification option, for verify=
1658 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1659 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1661 12. Inserted some casts to stop certain compilers warning when using pointer
1662 differences as field lengths or precisions in printf-type calls (mostly
1663 affecting debugging statements).
1665 13. Added optional readline() support for -be (dynamically loaded).
1667 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1668 same clock tick as a message's arrival, so that its received time was the
1669 same as the "first fail" time on the retry record, and that message
1670 remained on the queue past the ultimate address timeout, every queue runner
1671 would try a delivery (because it was past the ultimate address timeout) but
1672 after another failure, the ultimate address timeout, which should have then
1673 bounced the address, did not kick in. This was a "< instead of <=" error;
1674 in most cases the first failure would have been in the next clock tick
1675 after the received time, and all would be well.
1677 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1678 being recognized when the domain list was tested by the match_domain
1679 condition in an expansion string.
1681 16. Added the ${str2b64: operator.
1683 17. Exim was always calling setrlimit() to set a large limit for the number of
1684 processes, without checking whether the existing limit was already
1685 adequate. (It did check for the limit on file descriptors.) Furthermore,
1686 errors from getrlimit() and setrlimit() were being ignored. Now they are
1687 logged to the main and panic logs, but Exim does carry on, to try to do its
1688 job under whatever limits there are.
1690 18. Imported PCRE 5.0.
1692 19. Trivial typo in log message " temporarily refused connection" (the leading
1695 20. If the log selector return_path_on_delivery was set and an address was
1696 redirected to /dev/null, the delivery process crashed because it assumed
1697 that a return path would always be set for a "successful" delivery. In this
1698 case, the whole delivery is bypassed as an optimization, and therefore no
1701 21. Internal re-arrangement: the function for sending a challenge and reading
1702 a response while authentication was assuming a zero-terminated challenge
1703 string. It's now changed to take a pointer and a length, to allow for
1704 binary data in such strings.
1706 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1708 23. Exim was not respecting finduser_retries when seeking the login of the
1709 uid under which it was called; it was always trying 10 times. (The default
1710 setting of finduser_retries is zero.) Also, it was sleeping after the final
1711 failure, which is pointless.
1713 24. Implemented tls_on_connect_ports.
1715 25. Implemented acl_smtp_predata.
1717 26. If the domain in control=submission is set empty, Exim assumes that the
1718 authenticated id is a complete email address when it generates From: or
1719 Sender: header lines.
1721 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1722 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1723 chown and chgrp in /bin and hostname in /usr/bin.
1725 28. Exim was keeping the "process log" file open after each use, just as it
1726 does for the main log. This opens the possibility of it remaining open for
1727 long periods when the USR1 signal hits a daemon. Occasional processlog
1728 errors were reported, that could have been caused by this. Anyway, it seems
1729 much more sensible not to leave this file open at all, so that is what now
1732 29. The long-running daemon process does not normally write to the log once it
1733 has entered its main loop, and it closes the log before doing so. This is
1734 so that log files can straightforwardly be renamed and moved. However,
1735 there are a couple of unusual error situations where the daemon does write
1736 log entries, and I had neglected to close the log afterwards.
1738 30. The text of an SMTP error response that was received during a remote
1739 delivery was being truncated at 512 bytes. This is too short for some of
1740 the long messages that one sometimes sees. I've increased the limit to
1743 31. It is now possible to make retry rules that apply only when a message has a
1744 specific sender, in particular, an empty sender.
1746 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1747 it possible to be selective about when SMTP synchronization is enforced.
1749 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1751 32. Implemented hosts_connection_nolog.
1753 33. Added an ACL for QUIT.
1755 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1758 35. Added mailbox_size and mailbox_filecount to appendfile.
1760 36. Added control = no_multiline_responses to ACLs.
1762 37. There was a bug in the logic of the code that waits for the clock to tick
1763 in the case where the clock went backwards by a substantial amount such
1764 that the microsecond fraction of "now" was more than the microsecond
1765 fraction of "then" (but the whole seconds number was less).
1767 38. Added support for the libradius Radius client library this is found on
1768 FreeBSD (previously only the radiusclient library was supported).
1774 1. When certain lookups returned multiple values in the form name=value, the
1775 quoting of the values was not always being done properly. Specifically:
1776 (a) If the value started with a double quote, but contained no whitespace,
1778 (b) If the value contained whitespace other than a space character (i.e.
1779 tabs or newlines or carriage returns) it was not quoted.
1780 This fix has been applied to the mysql and pgsql lookups by writing a
1781 separate quoting function and calling it from the lookup code. The fix
1782 should probably also be applied to nisplus, ibase and oracle lookups, but
1783 since I cannot test any of those, I have not disturbed their existing code.
1785 2. A hit in the callout cache for a specific address caused a log line with no
1786 reason for rejecting RCPT. Now it says "Previous (cached) callout
1787 verification failure".
1789 3. There was an off-by-one bug in the queryprogram router. An over-long
1790 return line was truncated at 256 instead of 255 characters, thereby
1791 overflowing its buffer with the terminating zero. As well as fixing this, I
1792 have increased the buffer size to 1024 (and made a note to document this).
1794 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1795 when Exim is waiting for an SMTP response from a remote server, Exim
1796 restarts its select() call on the socket, thereby resetting its timeout.
1797 This is not a problem when such interrupts are rare. Somebody set up a cron
1798 job to run exiwhat every 2 minutes, which is less than the normal select()
1799 timeout (5 or 10 minutes). This meant that the select() timeout never
1800 kicked in because it was always reset. I have fixed this by comparing the
1801 time when an interrupt arrives with the time at the start of the first call
1802 to select(). If more time than the timeout has elapsed, the interrupt is
1803 treated as a timeout.
1805 5. Some internal re-factoring in preparation for the addition of Sieve
1806 extensions (by MH). In particular, the "personal" test is moved to a
1807 separate function, and given an option for scanning Cc: and Bcc: (which is
1808 not set for Exim filters).
1810 6. When Exim created an email address using the login of the caller as the
1811 local part (e.g. when creating a From: or Sender: header line), it was not
1812 quoting the local part when it contained special characters such as @.
1814 7. Installed new OpenBSD configuration files.
1816 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1817 try to make them clearer.
1819 9. Callout options, other than the timeout value, were being ignored when
1820 verifying sender addresses in header lines. For example, when using
1822 verify = header_sender/callout=no_cache
1824 the cache was (incorrectly) being used.
1826 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1827 only the Cygwin environment.
1829 11. When return_path_on_delivery was set as a log selector, if different remote
1830 addresses in the same message used different return paths and parallel
1831 remote delivery occurred, the wrong values would sometimes be logged.
1832 (Whenever a remote delivery process finished, the return path value from
1833 the most recently started remote delivery process was logged.)
1835 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1836 lines when AUTH and/or TLS are in use. This is the "received protocol"
1837 field. Exim used to use "asmtp" for authenticated SMTP, without any
1838 indication (in the protocol name) for TLS use. Now it follows the RFC and
1839 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1840 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1841 names appear in log lines as well as in Received: header lines.
1843 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1844 extensions, and comparison tests, and to fix some bugs.
1846 14. Changes to the "personal" filter test:
1848 (1) The test was buggy in that it was just doing the equivalent of
1849 "contains" tests on header lines. For example, if a user's address was
1850 anne@some.where, the "personal" test would incorrectly be true for
1852 To: susanne@some.where
1854 This test is now done by extracting each address from the header in turn,
1855 and checking the entire address. Other tests that are part of "personal"
1856 are now done using regular expressions (for example, to check local parts
1857 of addresses in From: header lines).
1859 (2) The list of non-personal local parts in From: addresses has been
1860 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1861 taken from the Sieve specification recommendations.
1863 (3) If the message contains any header line starting with "List-" it is
1864 treated as non-personal.
1866 (4) The test for "circular" in the Subject: header line has been removed
1867 because it now seems ill-conceived.
1869 15. Minor typos in src/EDITME comments corrected.
1871 16. Installed latest exipick from John Jetmore.
1873 17. If headers_add on a router specified a text string that was too long for
1874 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1875 of string_sprintf() is now avoided.
1877 18. $message_body_size was not set (it was always zero) when running the DATA
1878 ACL and the local_scan() function.
1880 19. For the "mail" command in an Exim filter, no default was being set for
1881 the once_repeat time, causing a random time value to be used if "once" was
1882 specified. (If the value happened to be <= 0, no repeat happened.) The
1883 default is now 0s, meaning "never repeat". The "vacation" command was OK
1884 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1885 (I found it when inspecting the code).
1887 20. There is now an overall timeout for performing a callout verification. It
1888 defaults to 4 times the callout timeout, which applies to individual SMTP
1889 commands during the callout. The overall timeout applies when there is more
1890 than one host that can be tried. The timeout is checked before trying the
1891 next host. This prevents very long delays if there are a large number of
1892 hosts and all are timing out (e.g. when the network connections are timing
1893 out). The value of the overall timeout can be changed by specifying an
1894 additional sub-option for "callout", called "maxwait". For example:
1896 verify = sender/callout=5s,maxwait=20s
1898 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1899 to the end before writing, but this should make it even safer).
1901 22. Exim was forgetting that it had advertised PIPELINING for the second and
1902 subsequent messages on an SMTP connection. It was also not resetting its
1903 memory on STARTTLS and an internal HELO.
1905 23. When Exim logs an SMTP synchronization error within a session, it now
1906 records whether PIPELINING has been advertised or not.
1908 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1909 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1910 rather than long int.
1912 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1914 26. Added the never_mail option to autoreply.
1920 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1921 crash if the getsockname() call failed; this can happen if a connection is
1922 closed very soon after it is established. The problem was simply in the
1923 order in which certain operations were done, causing Exim to try to write
1924 to the SMTP stream before it had set up the file descriptor. The bug has
1925 been fixed by making things happen in the correct order.
1931 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1932 before the connection was closed, thus losing the rejection response.
1934 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1935 some early Solaris releases, but causes trouble in current releases where
1936 socklen_t is defined.
1938 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1941 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1944 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1945 that releasing the top of it at the end releases what was used for sub-
1946 expansions (unless the block got too big). However, discard this block if
1947 the first thing is a variable or header, so that we can use its block when
1948 it is dynamic (useful for very large $message_headers, for example).
1950 6. Lookups now cache *every* query, not just the most recent. A new, separate
1951 store pool is used for this. It can be recovered when all lookup caches are
1952 flushed. Lookups now release memory at the end of their result strings.
1953 This has involved some general refactoring of the lookup sources.
1955 7. Some code has been added to the store_xxx() functions to reduce the amount
1956 of flapping under certain conditions.
1958 8. log_incoming_interface used to affect only the <= reception log lines. Now
1959 it causes the local interface and port to be added to several more SMTP log
1960 lines, for example "SMTP connection from", and rejection lines.
1962 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1964 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1966 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1967 could be overwritten at the end of the current message (or the start of a
1968 new message if it was set in a HELO ACL). The value is now preserved for
1969 the duration of the SMTP connection.
1971 12. If a transport had a headers_rewrite setting, and a matching header line
1972 contained an unqualified address, that address was qualified, even if it
1973 did not match any rewriting rules. The underlying bug was that the values
1974 of the flags that permit the existence of unqualified sender and recipient
1975 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1976 non-local messages, and by -bnq for local messages) were not being
1977 preserved with the message after it was received.
1979 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1980 "next input=" as part of the text comprising the host identity instead of
1981 the correct text. The code was using the same buffer for two different
1982 strings. However, depending on which order the printing function evaluated
1983 its arguments, the bug did not always show up. Under Linux, for example, my
1984 test suite worked just fine.
1986 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1987 doesn't work with some older versions of Perl. It has been changed to "my",
1988 which in any case is probably the better facility to use.
1990 15. A really picky compiler found some instances of statements for creating
1991 error messages that either had too many or two few arguments for the format
1994 16. The size of the buffer for calls to the DNS resolver has been increased
1995 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1996 for addresses that have a lot of PTR records. This alleviates a problem; it
1997 does not fully solve it.
1999 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2000 buffer now truncates the list and logs the incident, which is the same
2001 action as happens when Exim is looking up a host name and its aliases.
2002 Previously in this situation something unpredictable would happen;
2003 sometimes it was "internal error: store_reset failed".
2005 18. If a server dropped the connection unexpectedly when an Exim client was
2006 using GnuTLS and trying to read a response, the client delivery process
2007 crashed while trying to generate an error log message.
2009 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2010 string, for example:
2012 warn message = H1: something\nH2: something
2014 the text was added as a single header line from Exim's point of view
2015 though it ended up OK in the delivered message. However, searching for the
2016 second and subsequent header lines using $h_h2: did not work. This has been
2017 fixed. Similarly, if a system filter added multiple headers in this way,
2018 the routers could not see them.
2020 20. Expanded the error message when iplsearch is called with an invalid key to
2021 suggest using net-iplsearch in a host list.
2023 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2024 ACLs are no longer actually imposed (and a message to that effect is
2027 22. If a "gecos" field in a passwd entry contained escaped characters, in
2028 particular, if it contained a \" sequence, Exim got it wrong when building
2029 a From: or a Sender: header from that name. A second bug also caused
2030 incorrect handling when an unquoted " was present following a character
2031 that needed quoting.
2033 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2034 was not being matched caselessly.
2036 24. Arranged for all hyphens in the exim.8 source to be escaped with
2039 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2040 properly. Recipient callout cache records were still being keyed to include
2041 the sender, even when use_sender was set false. This led to far more
2042 callouts that were necessary. The sender is no longer included in the key
2043 when use_sender is false.
2045 26. Added "control = submission" modifier to ACLs.
2047 27. Added the ${base62d: operator to decode base 62 numbers.
2049 28. dnsdb lookups can now access SRV records.
2051 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2052 the configuration file.
2054 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2055 (-v) mode. This makes the output for a verbose queue run more intelligible.
2057 31. Added a use_postmaster feature to recipient callouts.
2059 32. Added the $body_zerocount variable, containing the number of binary zero
2060 bytes in the message body.
2062 33. The time of last modification of the "new" subdirectory is now used as the
2063 "mailbox time last read" when there is a quota error for a maildir
2066 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2068 35. Added +ignore_unknown as a special item in host lists.
2070 36. Code for decoding IPv6 addresses in host lists is now included, even if
2071 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2072 address was recognized as an IP address, but was then not correctly decoded
2073 into binary, causing unexpected and incorrect effects when compared with
2080 1. Very minor rewording of debugging text in manualroute to say "list of
2081 hosts" instead of "hostlist".
2083 2. If verify=header_syntax was set, and a header line with an unqualified
2084 address (no domain) and a large number of spaces between the end of the
2085 name and the colon was received, the reception process suffered a buffer
2086 overflow, and (when I tested it) crashed. This was caused by some obsolete
2087 code that should have been removed. The fix is to remove it!
2089 3. When running in the test harness, delay a bit after writing a bounce
2090 message to get a bit more predictability in the log output.
2092 4. Added a call to search_tidyup() just before forking a reception process. In
2093 theory, someone could use a lookup in the expansion of smtp_accept_max_
2094 per_host which, without the tidyup, could leave open a database connection.
2096 5. Added the variables $recipient_data and $sender_data which get set from a
2097 lookup success in an ACL "recipients" or "senders" condition, or a router
2098 "senders" option, similar to $domain_data and $local_part_data.
2100 6. Moved the writing of debug_print from before to after the "senders" test
2103 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2104 problems for message scanning, either using a data ACL, or using
2105 local_scan() because the Received: header was not generated till after they
2106 were called (in order to set the time as the time of reception completion).
2107 I have revised the way this works. The header is now generated after the
2108 body is received, but before the ACL or local_scan() are called. After they
2109 are run, the timestamp in the header is updated.
2115 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2116 before starting a queue runner without re-exec. This happened only when
2117 deliver_drop_privilege was set or when the Exim user was set to root. The
2118 effect of the bug was that timeouts during subsequent deliveries caused
2119 crashes instead of being properly handled. The handler is now left at its
2120 default (and expected) setting.
2122 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2123 message, again when deliver_drop_privilege is set or Exim is run as root.
2124 The bug described in (1) was not present in this case, but the tidying up
2125 of the other signals was missing. I have made the two cases consistent.
2127 3. The ignore_target_hosts setting on a manualroute router was being ignored
2128 for hosts that were looked up using the /MX notation.
2130 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2133 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2134 operated on the sender address. After changing the $sender_address to <>
2135 for the sender address verify, Exim was re-instated it as the original
2136 (before rewriting) address, but remembering that it had rewritten it, so it
2137 wasn't rewriting it again. This bug also had the effect of breaking the
2138 sender address verification caching when the sender address was rewritten.
2140 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2141 This has been changed so that if the ip literal address matches
2142 ignore_target_hosts, the router declines.
2144 7. Added expansion conditions match_domain, match_address, and match_local_
2145 part (NOT match_host).
2147 8. The placeholder for the Received: header didn't have a length field set.
2149 9. Added code to Exim itself and to exim_lock to test for a specific race
2150 condition that could lead to file corruption when using MBX delivery. The
2151 issue is with the lockfile that is created in /tmp. If this file is removed
2152 after a process has opened it but before that process has acquired a lock,
2153 there is the potential for a second process to recreate the file and also
2154 acquire a lock. This could lead to two Exim processes writing to the file
2155 at the same time. The added code performs the same test as UW imapd; it
2156 checks after acquiring the lock that its file descriptor still refers to
2157 the same named file.
2159 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2160 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2163 11. Added the smtp_active_hostname option. If used, this will typically be made
2164 to depend on the incoming interface address. Because $interface_address is
2165 not set up until the daemon has forked a reception process, error responses
2166 that can happen earlier (such as "too many connections") no longer contain
2169 12. If an expansion in a condition on a "warn" statement fails because a lookup
2170 defers, the "warn" statement is abandoned, and the next ACL statement is
2171 processed. Previously this caused the whole ACL to be aborted.
2173 13. Added the iplsearch lookup type.
2175 14. Added ident_timeout as a log selector.
2177 15. Added tls_certificate_verified as a log selector.
2179 16. Added a global option tls_require_ciphers (compare the smtp transport
2180 option of the same name). This controls incoming TLS connections.
2182 17. I finally figured out how to make tls_require_ciphers do a similar thing
2183 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2184 before starting the TLS session.
2186 18. Tabs are now shown as \t in -bP output.
2188 19. If the log selector return_path_on_delivery was set, Exim crashed when
2189 bouncing a message because it had too many Received: header lines.
2191 20. If two routers both had headers_remove settings, and the first one included
2192 a superfluous trailing colon, the final name in the first list and the
2193 first name in the second list were incorrectly joined into one item (with a
2194 colon in the middle).
2200 1. Added -C and -D options to the exinext utility, mainly to make it easier
2201 to include in the automated testing, but these could be helpful when
2202 multiple configurations are in use.
2204 2. The exinext utility was not formatting the output nicely when there was
2205 an alternate port involved in the retry record key, nor when there was a
2206 message id as well (for retries that were specific to a specific message
2207 and a specific host). It was also confused by IPv6 addresses, because of
2208 the additional colons they contain. I have fixed the IPv4 problem, and
2209 patched it up to do a reasonable job for IPv6.
2211 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2212 delivery, the log line now contains "pipelined" if PIPELINING was used.
2214 4. An SMTP transport process used to panic and die if the bind() call to set
2215 an explicit outgoing interface failed. This has been changed; it is now
2216 treated in the same way as a connect() failure.
2218 5. A reference to $sender_host_name in the part of a conditional expansion
2219 that was being skipped was still causing a DNS lookup. This no longer
2222 6. The def: expansion condition was not recognizing references to header lines
2223 that used bh_ and bheader_.
2225 7. Added the _cache feature to named lists.
2227 8. The code for checking quota_filecount in the appendfile transport was
2228 allowing one more file than it should have been.
2230 9. For compatibility with Sendmail, the command line option
2238 and sets the incoming protocol and host name (for trusted callers). The
2239 host name and its colon can be omitted when only the protocol is to be set.
2240 Note the Exim already has two private options, -pd and -ps, that refer to
2241 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2242 "s", but I don't think that's a major issue.
2244 10. A number of refactoring changes to the code, none of which should affect
2247 (a) The number of logging options was getting close to filling up the
2248 32-bit word that was used as a bit map. I have split them into two classes:
2249 those that are passed in the argument to log_write(), and those that are
2250 only ever tested independently outside of that function. These are now in
2251 separate 32-bit words, so there is plenty of room for expansion again.
2252 There is no change in the user interface or the logging behaviour.
2254 (b) When building, for example, log lines, the code previously used a
2255 macro that called string_cat() twice, in order to add two strings. This is
2256 not really sufficiently general. Furthermore, there was one instance where
2257 it was actually wrong because one of the argument was used twice, and in
2258 one call a function was used. (As it happened, calling the function twice
2259 did not affect the overall behaviour.) The macro has been replaced by a
2260 function that can join an arbitrary number of extra strings onto a growing
2263 (c) The code for expansion conditions now uses a table and a binary chop
2264 instead of a serial search (which was left over from when there were very
2265 few conditions). Also, it now recognizes conditions like "pam" even when
2266 the relevant support is not compiled in: a suitably worded error message is
2267 given if an attempt is made to use such a condition.
2269 11. Added ${time_interval:xxxxx}.
2271 12. A bug was causing one of the ddress fields not to be passed back correctly
2272 from remote delivery subprocesses. The field in question was not being
2273 subsequently used, so this caused to problems in practice.
2275 13. Added new log selectors queue_time and deliver_time.
2277 14. Might have fixed a bug in maildirsizefile handling that threw up
2278 "unexpected character" debug warnings, and recalculated the data
2279 unnecessarily. In any case, I expanded the warning message to give more
2282 15. Added the message "Restricted characters in address" to the statements in
2283 the default ACL that block characters like @ and % in local parts.
2285 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2286 Three changes have been made:
2288 (a) There was a serious bug; a negative response to MAIL caused the whole
2289 recipient domain to be cached as invalid, thereby blocking all messages
2290 to all local parts at the same domain, from all senders. This bug has
2291 been fixed. The domain is no longer cached after a negative response to
2292 MAIL if the sender used is not empty.
2294 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2297 (c) A new callout option, "use_sender" has been added for people who want
2298 the modified behaviour.
2304 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2307 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2308 indeed breaks things for older releases.
2310 3. Added additional logging to the case where there is a problem reading data
2311 from a filter that is running in a subprocess using a pipe, in order to
2312 try to track down a specific problem.
2314 4. Testing facility fudge: when running in the test harness and attempting
2315 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2316 getting "No route to host". Convert this to a timeout.
2318 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2321 6. Some OS don't have socklen_t but use size_t instead. This affects the
2322 fifth argument of getsockopt() amongst other things. This is now
2323 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2324 can be set for individual OS. I have set it for SunOS5, OSF1, and
2325 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2326 some earlier ones do not.
2328 7. Change 4.30/15 was not doing the test caselessly.
2330 8. The standard form for an IPv6 address literal was being rejected by address
2331 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2332 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2333 this, as well as the form without the "IPv6" on the front (but only when
2334 address literals are enabled, of course).
2336 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2338 10. Exim crashed if a message with an empty sender address specified by -f
2339 encountered a router with an errors_to setting. This could be provoked only
2340 by a command such as
2344 where an empty string was supplied; "<>" did not hit this bug.
2346 11. Installed PCRE release 4.5.
2348 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2349 remained set. It is now erased.
2351 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2352 times from message ids (which are base 36 rather than the normal 62).
2354 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2355 were being counted as actual protocol errors, and logged if the log
2356 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2357 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2358 and DATA following a set of rejected RCPTs do not count as protocol errors.
2359 In other words, Exim assumes they were pipelined, though this may not
2360 actually be the case. Of course, in all cases the client gets an
2361 appropriate error code.
2363 15. If a lookup fails in an ACL condition, a message about the failure may
2364 be available; it is used if testing the ACL cannot continue, because most
2365 such messages specify what the cause of the deferral is. However, some
2366 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2367 that caused an old message to be retained and used if a later statement
2368 caused a defer, replacing the real cause of the deferral.
2370 16. If an IP address had so many PTR records that the DNS lookup buffer
2371 was not large enough to hold them, Exim could crash while trying to process
2372 the truncated data. It now detects and logs this case.
2374 17. Further to 4.21/58, another change has been made: if (and only if) the
2375 first line of a message (the first header line) ends with CRLF, a bare LF
2376 in a subsequent header line has a space inserted after it, so as not to
2377 terminate the header.
2379 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2380 unnecessarily, used atoi() instead of strtol(), and didn't check the
2381 termination when getting file sizes from file names by regex.
2383 19. Completely re-implemented the support for maildirsize files, in the light
2384 of a number of problems with the previous contributed implementation
2385 (4.30/29). In particular:
2387 . If the quota is zero, the maildirsize file is maintained, but no quota is
2390 . If the maildir directory does not exist, it is created before any attempt
2391 to write a maildirsize file.
2393 . The quota value in the file is just a cache; if the quota is changed in
2394 the transport, the new value overrides.
2396 . A regular expression is available for excluding directories from the
2399 20. The autoreply transport checks the characters in options that define the
2400 message's headers; it allows continued headers, but it was checking with
2401 isspace() after an embedded newline instead of explicitly looking for a
2404 21. If all the "regular" hosts to which an address was routed had passed their
2405 expiry times, and had not reached their retry times, the address was
2406 bounced, even if fallback hosts were defined. Now Exim should go on to try
2409 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2410 equivalent code in the SMTP transport. Some hosts send humungous responses
2411 to HELO/EHLO, more than 1024 it seems.
2413 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2414 gives compiler warnings in some environments. I've now done it "properly",
2417 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2418 (because of problems with the built-in one) was declared to return uschar *
2419 instead of char *, causing compiler failure.
2421 25. Fixed a file descriptor leak when processing alias/forward files.
2423 26. Fixed a minor format string issue in dbfn.c.
2425 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2427 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2428 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2431 29. When a sender address is verified, it is cached, to save repeating the test
2432 when there is more than one recipient in a message. However, when the
2433 verification involves a callout, it is possible for different callout
2434 options to be set for different recipients. It is too complicated to keep
2435 track of this in the cache, so now Exim always runs a verification when a
2436 callout is required, relying on the callout cache for the optimization.
2437 The overhead is duplication of the address routing, but this should not be
2440 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2441 to be verified with callout=postmaster, and the main callout worked but the
2442 postmaster check failed, the verification correctly failed. However, if a
2443 subsequent RCPT command asked for sender verification *without* the
2444 postmaster check, incorrect caching caused this verification also to fail,
2447 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2448 it was not caching the DNS options (qualify_single, search_parents) that
2449 were used when the lookup failed. A subsequent lookup with different
2450 options therefore always gave the same answer, though there were cases
2451 where it should not have. (Example: a "domains = !$mx_any" option on a
2452 dnslookup router: the "domains" option is always processed without any
2453 widening, but the router might have qualify_single set.) Now Exim uses the
2454 cached value only when the same options are set.
2456 32. Added John Jetmore's "exipick" utility to the distribution.
2458 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2459 than a timeout (e.g. a certificate is required, and is not provided), an
2460 Exim server now closes the connection immediately. Previously it waited for
2461 the client to close - but if the client is SSL, it seems that they each
2462 wait for each other, leading to a delay before one of them times out.
2464 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2465 maintained 0.8.x compatibility because I don't think many are using it, and
2466 it is clearly obsolete.
2468 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2471 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2472 client certificate was expired. A simple patch fixes this, though I don't
2473 understand the full logic of why the verify callback is called multiple
2476 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2477 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2478 which causes problems with some clients (such as the Certicom SSL Plus
2479 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2480 disables the coutermeasure allowing Eudora to connect."
2482 38. Exim was not checking that a write() to a log file succeeded. This could
2483 lead to Bad Things if a log got too big, in particular if it hit a file
2484 size limit. Exim now panics and dies if it cannot write to a log file, just
2485 as it does if it cannot open a log file.
2487 39. Modified OS/Makefile-Linux so that it now contains
2489 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2491 The two -D definitions ensure that Exim is compiled with large file
2492 support, which makes it possible to handle log files that are bigger than
2495 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2496 instance) a domain was checked against a named list that involved a lookup,
2497 causing $domain_data to be set, then another domain was checked against the
2498 same list, then the first domain was re-checked, the value of $domain_data
2499 after the final check could be wrong. In particular, if the second check
2500 failed, it could be set empty. This bug probably also applied to
2503 41. The strip_trailing_dot option was not being applied to the address given
2504 with the -f command-line option.
2506 42. The code for reading a message's header from the spool was incrementing
2507 $received_count, but never initializing it. This meant that the value was
2508 incorrect (doubled) while delivering a message in the same process in which
2509 it was received. In the most common configuration of Exim, this never
2510 happens - a fresh exec is done - but it can happen when
2511 deliver_drop_privilege is set.
2513 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2514 it now includes up to 150 characters of the unexpected data in the log
2517 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2518 and building data strings. The size of both of these buffers was 10 000
2519 bytes - far larger than anybody would *ever* want, thought I. Needless to
2520 say, somebody hit the limit. I have increased the maximum line length to
2521 20 000 and the maximum data length of concatenated lines to 100 000. I have
2522 also fixed two bugs, because there was no checking on these buffers. Tsk,
2523 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2524 buffer is too small.
2526 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2527 lsearch lookups. Now it does.
2529 46. When parsing a route_list item in a manualroute router, a fixed-length
2530 buffer was used for the list of hosts. I made this 1024 bytes long,
2531 thinking that nobody would ever have a list of hosts that long. Wrong.
2532 Somebody had a whole pile of complicated expansion conditions, and the
2533 string was silently truncated, leading to an expansion error. It turns out
2534 that it is easier to change to an unlimited length (owing to other changes
2535 that have happened since this code was originally written) than to build
2536 structure for giving a limitation error. The length of the item that
2537 expands into the list of hosts is now unlimited.
2539 47. The lsearch lookup could not handle data where the length of text line was
2540 more than 4095 characters. Such lines were truncated, leading to shortened
2541 data being returned. It should now handle lines of any length.
2543 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2544 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2547 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2548 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2549 no longer happens in comment lines. A list of replacements is now placed
2550 at the head of all of the source files, except those whose only change is
2551 to replace PERL_COMMAND in the very first #! line.
2553 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2554 messages on the queue, with a bottom-up merge sort, using code contributed
2555 by Michael Haardt. This should make operations like -bp somewhat faster on
2556 large queues. It won't affect queue runners, except when queue_run_in_order
2559 51. Installed eximstats 1.31 in the distribution.
2561 52. Added support for SRV lookups to the dnslookup router.
2563 53. If an ACL referred to $message_body or $message_body_end, the value was not
2564 reset for any messages that followed in the same SMTP session.
2566 54. The store-handling optimization for building very long strings was not
2567 differentiating between the different store pools. I don't think this
2568 actually made any difference in practice, but I've tidied it.
2570 55. While running the routers to verify a sender address, $sender_address
2571 was still set to the sender address. This is wrong, because when routing to
2572 send a bounce to the sender, it would be empty. Therefore, I have changed
2573 it so that, while verifying a sender address, $sender_address is set to <>.
2574 (There is no change to what happens when verifying a recipient address.)
2576 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2577 target A or AAAA records (if not already returned) without resetting the
2578 qualify_single or search_parents options of the DNS resolver. These are
2579 inappropriate in this case because the targets of MX and SRV records must
2580 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2581 target that, when qualified, matched something in the local domain. These
2582 two options are now turned off when doing these lookups.
2584 57. It seems that at least some releases of Reiserfs (which does not have the
2585 concept of a fixed number of inodes) returns zero and not -1 for the
2586 number of available inodes. This interacted badly with check_spool_inodes,
2587 which assumed that -1 was the "no such thing" setting. What I have done is
2588 to check that the total number of inodes is greater than zero before doing
2589 the test of how many are available.
2591 58. When a "warn" ACL statement has a log_message modifier, the message is
2592 remembered, and not repeated. This is to avoid a lot of repetition when a
2593 message has many recipients that cause the same warning to be written.
2594 Howewer, Exim was preserving the list of already written lines for an
2595 entire SMTP session, which doesn't seem right. The memory is now reset if a
2596 new message is started.
2598 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2599 debugging output unless log_rewrite was also set.
2601 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2602 of (int)(handle) when we know that handle contains (void *)(-1).
2604 61. The Exim daemon panic-logs an error return when it closes the incoming
2605 connection. However "connection reset by peer" seems to be common, and
2606 isn't really an error worthy of noting specially, so that particular error
2609 62. When Exim is trying to find all the local interfaces, it used to panic and
2610 die if the ioctl to get the interface flags failed. However, it seems that
2611 on at least one OS (Solaris 9) it is possible to have an interface that is
2612 included in the list of interfaces, but for which you get a failure error
2613 for this call. This happens when the interface is not "plumbed" into a
2614 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2615 failure of the "get flags" call assumes that the interface is down.
2617 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2618 makes life easier for people who are doing arithmetic on fields extracted
2619 from dates, where you often get leading zeros that should not be
2620 interpreted as octal.
2622 64. Added qualify_domain to the redirect router, to override the global
2625 65. If a pathologically long header line contained very many addresses (the
2626 report of this problem mentioned 10 000) and each of them was rewritten,
2627 Exim could use up a very large amount of memory. (It kept on making new
2628 copies of the header line as it rewrote, and never released the old ones.)
2629 At the expense of a bit more processing, the header rewriting function has
2630 been changed so that it no longer eats memory in this way.
2632 66. The generation of the Received: header has been moved from the time that a
2633 message starts to be received, to the time that it finishes. The timestamp
2634 in the Received: header should now be very close to that of the <= log
2635 line. There are two side-effects of this change:
2637 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2638 logged header lines no longer include the local Received: line, because
2639 it has not yet been created. The same applies to a copy of the message
2640 that is returned to a non-SMTP sender when a message is rejected.
2642 (b) When a filter file is tested using -bf, no additional Received: header
2643 is added to the test message. After some thought, I decided that this
2646 This change does not affect the value of $received_for. It is still set
2647 after address rewriting, but before local_scan() is called.
2649 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2651 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2652 gave an unhelpful panic error message, and a defer error. I have managed to
2653 change this behaviour so that it now rejects any supplied certificate,
2654 which seems right, as the list of acceptable certificates is empty.
2656 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2657 gave an unhelpful defer error. I have not managed to make this reject any
2658 supplied certificates, but the error message it gives is "no certificate
2659 supplied", which is not helpful.
2661 70. exigrep's output now also includes lines that are not associated with any
2662 message, but which match the given pattern. Implemented by a patch from
2663 Martin Sluka, which also tidied up the Perl a bit.
2665 71. Recipient callout verification, like sender verification, was using <> in
2666 the MAIL FROM command. This isn't really the right thing, since the actual
2667 sender may affect whether the remote host accepts the recipient or not. I
2668 have changed it to use the actual sender in the callout; this means that
2669 the cache record is now keyed on a recipient/sender pair, not just the
2670 recipient address. There doesn't seem to be a real danger of callout loops,
2671 since a callout by the remote host to check the sender would use <>.
2672 [SEE ABOVE: changed after hitting problems.]
2674 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2675 temporary errors. However, in the case of such a code being given after
2676 the end of a data transmission (i.e. after ".") Exim was failing to write
2677 a retry record for the message. (Yes, there was some broken host that was
2678 actually sending 8xx at this point.)
2680 73. An unknown lookup type in a host list could cause Exim to panic-die when
2681 the list was checked. (An example that provoked this was putting <; in the
2682 middle of a list instead of at the start.) If this happened during a DATA
2683 ACL check, a -D file could be left lying around. This kind of configuration
2684 error no longer causes Exim to die; instead it causes a defer errror. The
2685 incident is still logged to the main and panic logs.
2687 74. Buglet left over from Exim 3 conversion. The message "too many messages
2688 in one connection" was written to the rejectlog but not the mainlog, except
2689 when address rewriting (yes!) was being logged.
2691 75. Added write_rejectlog option.
2693 76. When a system filter was run not as root (that is, when system_filter_user
2694 was set), the values of the $n variables were not being returned to the
2695 main process; thus, they were not subsequently available in the $sn
2698 77. Added +return_path_on_delivery log selector.
2700 78. A connection timeout was being treated differently from recipients deferred
2701 when testing hosts_max_try with a message that was older than the host's
2702 retry timeout. (The host should not be counted, thus allowing all hosts to
2703 be tried at least once before bouncing.) This may have been the cause of an
2704 occasionally reported bug whereby a message would remain on the queue
2705 longer than the retry timeout, but would be bounced if a delivery was
2706 forced. I say "may" because I never totally pinned down the problem;
2707 setting up timeout/retry tests is difficult. See also the next item.
2709 79. The ultimate address timeout was not being applied to errors that involved
2710 a combination of host plus message (for example, a timeout on a MAIL
2711 command). When an address resolved to a number of possible hosts, and they
2712 were not all tried for each delivery (e.g. because of hosts_max_try), a
2713 message could remain on the queue longer than the retry timeout.
2715 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2718 81. Fixed an obscure SMTP outgoing bug which required at least the following
2719 conditions: (a) there was another message waiting for the same server;
2720 (b) the server returned 5xx to all RCPT commands in the first message so
2721 that the message was not completed; (c) the server dropped the connection
2722 or gave a negative response to the RSET that Exim sends to abort the
2723 transaction. The observed case was a dropped connection after DATA that had
2724 been sent in pipelining mode. That is, the server had advertised PIPELINING
2725 but was not implementing it correctly. The effect of the bug was incorrect
2726 behaviour, such as trying another host, and this could lead to a crash.
2732 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2733 and daemon.c were passed as pointers to ints; they should have been
2734 pointers to socklen_t variables (which are typically unsigned ints).
2736 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2739 3. Fixed a really odd bug that affected only the testing scheme; patching a
2740 certain fixed string in the binary changed the value of another string that
2741 happened to be identical to the end of the original first string.
2743 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2744 name", it returns that address as the IP address. On some operating
2745 systems (e.g. Solaris), it also passes back the IP address string as the
2746 "host name". However, on others (e.g. Linux), it passes back an empty
2747 string. Exim wasn't checking for this, and was changing the host name to an
2748 empty string, assuming it had been canonicized.
2750 5. Although rare, it is permitted to have more than one PTR record for a given
2751 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2752 all the names associated with an address, because they do in Solaris.
2753 However, it seems that they do not in Linux for data that comes from the
2754 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2755 I found this out when I moved to a new Linux workstation and tried to run
2756 the Exim test suite.
2758 To get round this problem I have changed the code so that it now does its
2759 own call to the DNS to look up PTR records when searching for a host name.
2760 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2761 addresses that are only in /etc/hosts are still found.
2763 This behaviour is, however, controlled by an option called host_lookup_
2764 order, which defaults to "bydns:byaddr". If people want to use the other
2765 order, or indeed, just use one or the other means of lookup, they can
2766 specify it in this variable.
2768 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2769 some operating systems, this comes back from gethostbyaddr() as an empty
2770 string, and this is what Exim used to test for. However, it seems that in
2771 other systems, "." is yielded. Exim now tests for this case too.
2773 7. The values of check_spool_space and check_log_space are now held internally
2774 as a number of kilobytes instead of an absolute number of bytes. If a
2775 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2776 kilobyte. This means that much larger values can be stored.
2778 8. Exim monitor: an attempt to get the action menu when not actually pointing
2779 at a message produces an empty menu entitled "No message selected". This
2780 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2781 no entries in it ("Shell widget menu has zero width and/or height"). So I
2782 have added a single, blank menu entry in this case.
2784 9. Added ${quote_local_part.
2786 10. MIME decoding is now applied to the contents of Subject: header lines when
2789 11. Now that a reference to $sender_host_address automatically causes a reverse
2790 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2791 host lookup before query-style lookups in lists that might use this
2792 variable. This has therefore been abolished, and the "net-" prefix is no
2793 longer necessary for query-style lookups.
2795 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2796 have been a typo for LFLAGS, so it has been changed.
2798 13. The install script calls Exim with "-C /dev/null" in order to find the
2799 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2800 to be output. Howeve, since Exim outputs its version number before the
2801 error, it didn't break the script. It just looked ugly. I fixed this by
2802 always allowing "-C /dev/null" if the caller is root.
2804 14. Ignore overlarge ACL variable number when reading spool file - insurance
2805 against a later release with more variables having written the file.
2807 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2808 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2809 this, as well as the form without the "IPv6" on the front.
2811 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2812 OS/Makefile-Darwin file.
2814 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2815 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2817 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2818 variable. However, some OS use ints for inodes. I've added cast to long int
2819 to get rid of the compiler warning.
2821 19. I had forgotten to lock out "/../" in configuration file names when
2822 ALT_CONFIG_PREFIX was set.
2824 20. Routers used for verification do not need to specify transports. However,
2825 if such a router generated a host list, and callout was configured, Exim
2826 crashed, because it could not find a port number from the (non-existent)
2827 transport. It now assumes port 25 in this circumstance.
2829 21. Added the -t option to exigrep.
2831 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2832 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2833 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2834 them is set, LOOKUP_LSEARCH is forced.
2836 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2838 24. Added sender and host information to the "rejected by local_scan()" log
2839 line; previously there was no indication of these.
2841 25. Added .include_if_exists.
2843 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2844 Linux. It turns out that not all file systems support this. Apparently some
2845 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2846 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2847 not supported on the file descriptor, is now ignored when Exim is trying to
2848 sync a directory. This applies only to Linux.
2850 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2852 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2853 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2856 29. Added support for maildirsize files from supplied patch (modified a bit).
2858 30. The use of :fail: followed by an empty string could lead Exim to respond to
2859 sender verification failures with (e.g.):
2861 550 Verification failed for <xxx>
2862 550 Sender verify failed
2864 where the first response line was missing the '-' that indicates it is not
2865 the final line of the response.
2867 31. The loop for finding the name of the user that called Exim had a hardwired
2868 limit of 10; it now uses the value of finduser_retries, which is used for
2869 all other user lookups.
2871 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2874 33. Exim was neglecting to zero errno before one call of strtol() when
2875 expanding a string and expecting an integer value. On some systems this
2876 resulted in spurious "integer overflow" errors. Also, it was casting the
2877 result into an int without checking.
2879 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2880 did not work. The code looks as if it has *never* worked, though it appears
2881 to have been documented since at least releast 1.62. I have made it work.
2883 35. The "timeout_DNS" error in retry rules, also documented since at least
2884 1.62, also never worked. As it isn't clear exactly what this means, and
2885 clearly it isn't a major issue, I have abolished the feature by treating it
2886 as "timeout", and writing a warning to the main and panic logs.
2888 36. The display of retry rules for -brt wasn't always showing the error code
2891 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2892 timeout_connect_A, timeout_connect_MX.
2894 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2895 to the empty sender.
2897 39. The daemon was not analysing the content of -oX till after it had closed
2898 stderr and disconnected from the controlling terminal. This meant that any
2899 syntax errors were only noted on the panic log, and the return code from
2900 the command was 0. By re-arranging the code a little, I've made the
2901 decoding happen first, so such errors now appear on stderr, and the return
2902 code is 1. However, the actual setting up of the sockets still happens in
2903 the disconnected process, so errors there are still only recorded on the
2906 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2907 connections (as happens on some IP stacks) was logged at start up time as
2908 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2909 it from "IPv6 and IPv4", which means that two separate sockets are being
2912 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2913 says whether AF_INET or AF_INET6 was passed as an argument.
2915 42. Exiwhat output was messed up when time zones were included in log
2918 43. Exiwhat now gives more information about the daemon's listening ports,
2919 and whether -tls-on-connect was used.
2921 44. The "port" option of the smtp transport is now expanded.
2923 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2924 silently ignored. Now an error message is written to the main and panic
2927 46. There's a new ACL modifier called "logwrite" which writes to a log file
2928 as soon as it is encountered.
2930 47. Added $local_user_uid and $local_user_gid at routing time.
2932 48. Exim crashed when trying to verify a sender address that was being
2935 49. Exim was recognizing only a space character after ".include". It now also
2936 recognizes a tab character.
2938 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2939 extracting the relevant information from the specification. The man page no
2940 longer contains scrambled data for the -d option, and I've added a section
2941 at the front about calling Exim under different names.
2943 51. Added "extra_headers" argument to the "mail" command in filter files.
2945 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2948 53. Installed eximstats 1.29.
2950 54. Added transport_filter_timeout as a generic transport option.
2952 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2953 Cc: header lines. This was required by RFC 822, but it not required by RFC
2956 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2957 incoming messages that did not have them. Now it does so only if the
2958 message originates locally, that is, if there is no associated remote host
2959 address. When Resent- header lines are present, this applies to the Resent-
2960 lines rather than the non-Resent- lines.
2962 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2963 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2965 58. Messages for configuration errors now include the name of the main
2966 configuration file - useful now that there may be more than one file in a
2967 list (.included file names were always shown).
2969 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2970 for those rare installations that do not start the daemon as root or run it
2971 setuid root. I've cut out the call to initgroups() if the daemon is not
2974 60. The Exim user and group can now be bound into the binary as text strings
2975 that are looked up at the start of Exim's processing.
2977 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2979 62. Added $mailstore_basename variable.
2981 63. Installed patch to sieve.c from Michael Haardt.
2983 64. When Exim failed to open the panic log after failing to open the main log,
2984 the original message it was trying to log was written to stderr and debug
2985 output, but if they were not available (the usual case in production), it
2986 was lost. Now it is written to syslog before the two lines that record the
2987 failures to open the logs.
2989 65. Users' Exim filters run in subprocesses under the user's uid. It is
2990 possible for a "deliver" command or an alias in a "personal" command to
2991 provoke an address rewrite. If logging of address rewriting is configured,
2992 this fails because the process is not running as root or exim. There may be
2993 a better way of dealing with this, but for the moment (because 4.30 needs
2994 to be released), I have disabled address rewrite logging when running a
2995 filter in a non-root, non-exim process.
3001 1. The buildconfig auxiliary program wasn't quoting the value set for
3002 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3003 not defined. This bug was masked in 4.22 by the effect that was fixed in
3006 2. Some messages that were rejected after a message id was allocated were
3007 shown as "incomplete" by exigrep. It no longer does this for messages that
3008 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3010 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3011 have allow_domain_literals set, the ID did not get logged in the <= line.
3012 Domain literals are now always recognized in Message-ID: header lines.
3014 4. The first argument for a ${extract expansion item is the key name or field
3015 number. Leading and trailing spaces in this item were not being ignored,
3016 causing some misleading effects.
3018 5. When deliver_drop_privilege was set, single queue runner processes started
3019 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3020 same command in the process it spins off) were not dropping privilege.
3022 6. When the daemon running as "exim" started a queue runner, it always
3023 re-executed Exim in the spun-off process. This is a waste of effort when
3024 deliver_drop_privilege is set. The new process now just calls the
3025 queue-runner function directly.
3031 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3034 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3035 ignored. Though the use of -oP was forcing the writing of a pid file, it
3036 was always written to the default place.
3038 3. If the message "no IP address found for host xxxx" is generated during
3039 incoming verification, it is now followed by identification of the incoming
3040 connection (so you can more easily find what provoked it).
3042 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3044 5. Added some features to "harden" Exim a bit more against certain attacks:
3046 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3047 be put in Local/Makefile. This is like the never_users runtime option,
3048 but it cannot be overridden. The default setting is "root".
3050 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3051 prefix string with which any file named in a -C command line option
3054 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3055 is retained for -C and -D only if the caller of Exim is root. Without
3056 it, the exim user may also use -C and -D and retain privilege.
3058 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3059 command line option is disabled.
3061 6. Macro names set by the -D option must start with an upper case letter, just
3062 like macro names defined in the configuration file.
3064 7. Added "dereference=" facility to LDAP.
3066 8. Two instances of the typo "uknown" in the source files are fixed.
3068 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3069 the Configure-Makefile script screwed up while processing it.
3071 10. Incorporated PCRE 4.4.
3073 11. The SMTP synchronization check was not operating right at the start of an
3074 SMTP session. For example, it could not catch a HELO sent before the client
3075 waited for the greeting. There is now a check for outstanding input at the
3076 point when the greeting is written. Because of the duplex, asynchronous
3077 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3078 way, but not yet received, when the check is performed.
3080 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3081 on TCP/IP sockets, because this apparently causes some broken clients to
3084 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3085 unchanged) from the Cygwin maintainer.
3087 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3088 was supported instead of testing for mbx. Effectively a typo.
3090 15. The spa authenticator server code was not checking that the input it
3091 received was valid base64.
3093 16. The debug output line for the "set" modifier in ACLs was not showing the
3094 name of the variable that was being set.
3096 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3098 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3099 lookup on its own. Something else was needed to trigger the lookup. For
3100 example, a match in host_lookup or the need for a host name in a host list.
3101 Now, if $sender_host_name is referenced and the host name has not yet been
3102 looked up, a lookup is performed. If the lookup fails, the variable remains
3103 empty, and $host_lookup_failed is set to "1".
3105 19. Added "eqi" as a case-independent comparison operator.
3107 20. The saslauthd authentication condition could segfault if neither service
3108 nor realm was specified.
3110 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3111 error message that was logged was misleading, and incoming SMTP
3112 connections were dropped. The message is now more accurate, and temporary
3113 errors are given to SMTP connections.
3115 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3116 (except RSET) with a 421 error, until QUIT is received. However, it was
3117 failing to send a response to QUIT.
3119 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3120 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3121 ACL is now run after the helo_try_verify_hosts code.
3123 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3124 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3125 case-independent, so other case variants are also recognized). Apparently
3126 some systems use these upper case variants.
3128 25. If more than two messages were waiting for the same host, and a transport
3129 filter was specified for the transport, Exim sent two messages over the
3130 same TCP/IP connection, and then failed with "socket operation on non-
3131 socket" when it tried to send the third.
3133 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3135 27. The extern definition of crypt16() in expand.c was not being excluded when
3136 the OS had its own crypt16() function.
3138 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3139 as a preferred synonym for return_size_limit, both as an option and as an
3142 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3144 30. Changed the default configuration ACL to relax the local part checking rule
3145 for addresses that are not in any local domains. For these addresses,
3146 slashes and pipe symbols are allowed within local parts, but the sequence
3147 /../ is explicitly forbidden.
3149 31. SPA server authentication was not clearing the challenge buffer before
3152 32. log_message in a "warn" ACL statement was writing to the reject log as
3153 well as to the main log, which contradicts the documentation and doesn't
3154 seem right (because no rejection is happening). So I have stopped it.
3156 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3157 However, I am unable to do any testing of this.
3159 34. Fixed an infelicity in the appendfile transport. When checking directories
3160 for a mailbox, to see if any needed to be created, it was accidentally
3161 using path names with one or more superfluous leading slashes; tracing
3162 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3164 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3165 contents are added to the log line that is written for every discarded
3166 recipient. (Previously a log_message setting was ignored.)
3168 36. The ${quote: operator now quotes the string if it is empty.
3170 37. The install script runs exim in order to find its version number. If for
3171 some reason other than non-existence or emptiness, which it checks, it
3172 could not run './exim', it was installing it with an empty version number,
3173 i.e. as "exim-". This error state is now caught, and the installation is
3176 38. An argument was missing from the function that creates an error message
3177 when Exim fails to connect to the socket for saslauthd authentication.
3178 This could cause Exim to crash, or give a corrupted message.
3180 39. Added isip, isip4, and isip6 to ${if conditions.
3182 40. The ACL variables $acl_xx are now saved with the message, and can be
3183 accessed later in routers, transports, and filters.
3185 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3186 strings in the file are not string-expanded.
3188 42. If a MAIL command specified a SIZE value that was too large to fit into an
3189 int variable, the check against message_size_limit failed. Such values are
3190 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3191 day this will have to be increased, but I don't think I want to be around
3192 when emails are that large.
3199 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3200 iconv() is not standard in FreeBSD.
3202 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3203 IPv6 enabled. The observed symptom was a segmentation fault on return from
3204 the function os_common_find_running_interfaces() in src/os.c.
3206 3. In the check_special_case() function in daemon.c I had used "errno" as an
3207 argument name, which causes warnings on some systems. This was basically a
3208 typo, since it was named "eno" in the comments!
3210 4. The code that waits for the clock to tick (at a resolution of some fraction
3211 of a second) so as to ensure message-id uniqueness was always waiting for
3212 at least one whole tick, when it could have waited for less. [This is
3213 almost certainly not relevant at current processor speeds, where it is
3214 unlikely to ever wait at all. But we try to future-proof.]
3216 5. The function that sleeps for a time interval that includes fractions of a
3217 second contained a race. It did not block SIGALRM between setting the
3218 timer, and suspending (a couple of lines later). If the interval was short
3219 and the sigsuspend() was delayed until after it had expired, the suspension
3220 never ended. On busy systems this could lead to processes getting stuck for
3223 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3224 process, before it forks any delivery processes. The open lookup caching
3225 mechanism meant that the open file or database connection was passed into
3226 the delivery process. The problem was that delivery processes always tidy
3227 up cached lookup data. This could cause a problem for the next delivery
3228 process started by the queue runner, because the external queue runner
3229 process does not know about the closure. So the next delivery process
3230 still has data in the lookup cache. In the case of a file lookup, there was
3231 no problem because closing a file descriptor in a subprocess doesn't affect
3232 the parent. However, if the lookup was caching a connection to a database,
3233 the connection was closed, and the second delivery process was likely to
3234 see errors such as "PGSQL: query failed: server closed the connection
3235 unexpectedly". The problem has been fixed by closing all cached lookups
3236 in a queue runner before running a delivery process.
3238 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3239 seem to have the "const" qualifier which it has on other OS. I've
3242 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3243 *of the same type* (client or server) with the same public name that an
3244 error should be diagnosed.
3246 9. When Exim looked up a host name for an IP address, but failed to find the
3247 original IP address when looking up the host name (a safety check), it
3248 output the message "<ip address> does not match any IP for NULL", which was
3249 confusing, to say the least. The bug was that the host name should have
3250 appeared instead of "NULL".
3252 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3253 user that is built into the binary, and the -C or -D options is used, root
3254 privilege is dropped before the configuration file is read. In addition,
3255 logging is switched to stderr instead of the normal log files. If the
3256 configuration then re-defines the Exim user, the unprivileged environment
3257 is probably not what is expected, so Exim logs a panic warning message (but
3260 However, if deliver_drop_privilege is set, the unprivileged state may well
3261 be exactly what is intended, so the warning has been cut out in that case,
3262 and Exim is allowed to try to write to its normal log files.
3268 1. smtp_return_error_details was not giving details for temporary sender
3269 or receiver verification errors.
3271 2. Diagnose a configuration error if two authenticators have the same public
3274 3. Exim used not to create the message log file for a message until the first
3275 delivery attempt. This could be confusing when incoming messages were held
3276 for policy or load reasons. The message log file is now created at the time
3277 the message is received, and an initial "Received" line is written to it.
3279 4. The automatically generated man page for command line options had a minor
3280 bug that caused no ill effects; however, a more serious problem was that
3281 the procedure for building the man page automatically didn't always
3282 operate. Consequently, release 4.20 contains an out-of-date version. This
3283 shouldn't happen again.
3285 5. When building Exim with embedded Perl support, the script that builds the
3286 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3287 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3289 6. The freeze_tell option was not being used for messages that were frozen on
3290 arrival, either by an ACL or by local_scan().
3292 7. Added the smtp_incomplete_transaction log selector.
3294 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3295 was accepting AUTH without a new EHLO.
3297 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3298 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3301 10. Logging of TCP/IP connections (when configured) now happens in the main
3302 daemon process instead of the child process, so that the TCP/IP connection
3303 count is more accurate (but it can never be perfect).
3305 11. The use of "drop" in a nested ACL was not being handled correctly in the
3306 outer ACL. Now, if condition failure induced by the nested "drop" causes
3307 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3308 or "require"), the connection is dropped.
3310 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3311 that yield "discard" can now be used with an "accept" or a "discard" verb,
3312 but an error is generated for any others (because I can't see a useful way
3313 to define what should happen).
3315 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3316 are now processed in the same way as lines in the Exim configuration file.
3317 In particular, continuation lines are supported.
3319 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3321 15. Added -ti meaning -t -i.
3323 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3324 domains, and warn by logging if others are found.
3326 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3327 returned from GIFCONF when Exim is trying to find the list of interfaces on
3328 a host. The code in os.c has been modified to copy each ifreq to an aligned
3329 structure in all cases.
3331 Also, in some cases, the returned ifreq's were being copied to a 'struct
3332 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3333 means the last couple of bytes of an IPv6 address could be chopped if the
3334 ifreq contained only a normal sockaddr (14 bytes storage).
3336 18. Named domain lists were not supported in the hosts_treat_as_local option.
3337 An entry such as +xxxx was not recognized, and was treated as a literal
3340 19. Ensure that header lines added by a DATA ACL are included in the reject log
3341 if the ACL subsequently rejects the message.
3343 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3344 MD5 (which is deprecated).
3346 21. When testing a filter file using -bf, Exim was writing a message when it
3347 took the sender from a "From " line in the message, but it was not doing so
3348 when it took $return_path from a Return-Path: header line. It now does.
3350 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3351 with a valid header line field name (a series of printing characters
3352 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3354 23. Changed "disc" in the source to "disk" to conform to the documentation and
3355 the book and for uniformity.
3357 24. Ignore Sendmail's -Ooption=value command line item.
3359 25. When execve() failed while trying to run a command in a pipe transport,
3360 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3361 could be confused with a return value of 69 from the command itself. This
3362 has been changed to 127, the value the shell returns if it is asked to run
3363 a non-existent command. The wording for the related log line suggests a
3364 non-existent command as the problem.
3366 26. If received_header_text expands to an empty string, do not add a Received:
3367 header line to the message. (Well, it adds a token one on the spool, but
3368 marks it "old" so that it doesn't get used or transmitted.)
3370 27. Installed eximstats 1.28 (addition of -nt option).
3372 28. There was no check for failure on the call to getsockname() in the daemon
3373 code. This can fail if there is a shortage of resources on the system, with
3374 ENOMEM, for example. A temporary error is now given on failure.
3376 29. Contrary to the C standard, it seems that in some environments, the
3377 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3378 program. Exim now does this explicitly; it affects the formatting of
3379 timestamps using strftime().
3381 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3382 complaints. I've now initialized all the variables, to avoid this.
3384 32. Header lines added by a system filter were not being "seen" during
3385 transport-time rewrites.
3387 33. The info_callback() function passed to OpenSSL is set up with type void
3388 (*)(SSL *, int, int), as described somewhere. However, when calling the
3389 function (actually a macro) that sets it up, the type void(*)() is
3390 expected. I've put in a cast to prevent warnings from picky compilers.
3392 34. If a DNS black list lookup found a CNAME record, but there were no A
3393 records associated with the domain it pointed at, Exim crashed.
3395 35. If a DNS black list lookup returned more than one A record, Exim ignored
3396 all but the first. It now scans all returned addresses if a particular IP
3397 value is being sought. In this situation, the contents of the
3398 $dnslist_value variable are a list of all the addresses, separated by a
3401 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3402 to accept a host name and all its aliases if the forward lookup for any of
3403 them yielded the IP address of the incoming connection. Now it accepts only
3404 those names whose forward lookup yields the correct IP address. Any other
3405 names are discarded. This closes a loophole whereby a rogue DNS
3406 administrator could create reverse DNS records to break through a
3407 wildcarded host restriction in an ACL.
3409 37. If a user filter or a system filter that ran in a subprocess used any of
3410 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3411 the wrong values were passed to the pipe command ($thisaddress had the
3412 value of $0, $0 had the value of $1, etc). This bug was introduced by
3413 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3415 38. Improved the line breaking for long SMTP error messages from ACLs.
3416 Previously, if there was no break point between 40 and 75 characters, Exim
3417 left the rest of the message alone. Two changes have been made: (a) I've
3418 reduced the minimum length to 35 characters; (b) if it can't find a break
3419 point between 35 and 75 characters, it looks ahead and uses the first one
3420 that it finds. This may give the occasional overlong line, but at least the
3421 remaining text gets split now.
3423 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3424 file descriptors might be low, and that setting 1000 would always raise it.
3425 It turns out that in some environments, the limit is already over 1000 and
3426 that lowering it causes trouble. So now Exim takes care not to decrease it.
3428 40. When delivering a message, the value of $return_path is set to $sender_
3429 address at the start of routing (routers may change the value). By an
3430 oversight, this default was not being set up when an address was tested by
3431 -bt or -bv, which affected the outcome if any router or filter referred to
3434 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3435 log only when "message" or "log_message" are set. However, if one of the
3436 conditions was an address verification, or a call to a nested ACL, the
3437 messages generated by the underlying test were being passed through. This
3438 no longer happens. The underlying message is available in $acl_verify_
3439 message for both "message" and "log_message" expansions, so it can be
3440 passed through if needed.
3442 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3443 new expansion $bh_ to give the encoded byte string without charset
3444 translation. Translation happens only if iconv() is available; HAVE_ICONV
3445 indicates this at build time. HEADERS_CHARSET gives the charset to
3446 translate to; headers_charset can change it in the configuration, and
3447 "headers charset" can change it in an individual filter file.
3449 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3450 that creates RFC 2047 encoded "words" labels them as that charset instead
3451 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3452 expansion operator; (ii) when Exim creates a From: line for a local
3453 message; (iii) when a header line is rewritten to include a "phrase" part.
3455 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3456 buggy, causing it to skip the first lines of messages whose message ID
3457 ended in 'D'. This would not have bitten before Exim release 4.14, because
3458 message IDs were unlikely to end in 'D' before then. The effect was to have
3459 incorrect size information for certain domains.
3461 45. #include "config.h" was missing at the start of the crypt16.c module. This
3462 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3465 46. If there was a timeout during a "random" callout check, Exim treated it as
3466 a failure of the random address, and carried on sending RSET and the real
3467 address. If the delay was just some slowness somewhere, the response to the
3468 original RCPT would be taken as a response to RSET and so on, causing
3469 mayhem of various kinds.
3471 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3472 when I implemented it. It didn't allow for the fact that some option values
3473 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3474 the right test for positive values.
3476 48. Domain names in DNS records are case-independent. Exim always looks them up
3477 in lower case. Some resolvers return domain names in exactly the case they
3478 appear in the zone file, that is, they may contain uppercase letters. Not
3479 all resolvers do this - some return always lower case. Exim was treating a
3480 change of case by a resolver as a change of domain, similar to a widening
3481 of a domain abbreviation. This triggered its re-routing code and so it was
3482 trying to route what was effectively the same domain again. This normally
3483 caused routing to fail (because the router wouldn't handle the domain
3484 twice). Now Exim checks for this case specially, and just changes the
3485 casing of the domain that it ultimately uses when it transmits the message
3488 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3491 50. If a filter generated a file delivery with a non-absolute name (possible if
3492 no home directory exists for the router), the forbid_file option was not
3495 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3496 the existing equality matching.
3498 52. Exim was using ints instead of ino_t variables in some places where it was
3499 dealing with inode numbers.
3501 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3502 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3503 called TMPDIR, and if it finds it is different, it changes its value.
3505 54. The smtp_printf() function is now made available to local_scan() so
3506 additional output lines can be written before returning. There is also an
3507 smtp_fflush() function to enable the detection of a dropped connection.
3508 The variables smtp_input and smtp_batched_input are exported to
3511 55. Changed the default runtime configuration: the message "Unknown user"
3512 has been removed from the ACL, and instead placed on the localuser router,
3513 using the cannot_route_message feature. This means that any verification
3514 failures that generate their own messages won't get overridden. Similarly,
3515 the "Unrouteable address" message that was in the ACL for unverifiable
3516 relay addresses has also been removed.
3518 56. Added hosts_avoid_esmtp to the smtp transport.
3520 57. The exicyclog script was not checking for the esoteric option
3521 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3522 will work only if exicyclog is run under the appropriate euid.
3524 58. Following a discussion on the list, the rules by which Exim recognises line
3525 endings on incoming messages have been changed. The -dropcr and drop_cr
3526 options are now no-ops, retained only for backwards compatibility. The
3527 following line terminators are recognized: LF CRLF CR. However, special
3528 processing applies to CR:
3530 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3531 nor a local message in the state where . is a terminator.
3533 (ii) If a bare CR is encountered in a header line, an extra space is added
3534 after the line terminator so as not to end the header. The reasoning
3535 behind this is that bare CRs in header lines are most likely either
3536 to be mistakes, or people trying to play silly games.
3538 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3539 was being incorrectly given as 18 bytes larger than it should have been.
3540 This is a VOB (very old bug).
3542 60. This may never have affected anything current, but just in case it has:
3543 When the local host is found other than at the start of a list of hosts,
3544 the local host, those with the same MX, and any that follow, are discarded.
3545 When the list in question was part of a longer list of hosts, the following
3546 hosts (not currently being processed) were also being discarded. This no
3547 longer happens. I'm not sure if this situation could ever has previously
3550 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3553 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3554 header. This is something that is recommended in a new Internet Draft, and
3555 is something that is documented as being done by Sendmail. There are two
3556 possible values. For messages generated by the autoreply transport, Exim
3559 Auto-Submitted: auto-replied
3561 whereas for all other generated messages (e.g. bounces) it adds
3563 Auto-Submitted: auto-generated
3565 63. The "personal" condition in filters now includes a test for the
3566 Auto-Submitted: header. If it contains the string "auto-" the message it
3567 not considered personal.
3569 64. Added rcpt_include_affixes as a generic transport option.
3571 65. Added queue_only_override (default true).
3573 66. Added the syslog_duplication option.
3575 67. If what should have been the first header line of a message consisted of
3576 a space followed by a colon, Exim was mis-interpreting it as a header line.
3577 It isn't of course - it is syntactically invalid and should therefore be
3578 treated as the start of the message body. The misbehaviour could have
3579 caused a number of strange effects, including loss of data in subsequent
3580 header lines, and spool format errors.
3582 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3583 client host had authenticated. This control can now be exercised by an ACL
3584 for more flexibility.
3586 69. By default, callouts do not happen when testing with -bh. There is now a
3587 variant, -bhc, which does actually run the callout code, including
3588 consulting and updating the callout cache.
3590 70. Added support for saslauthd authentication, courtesy of Alexander
3593 71. If statvfs() failed on the spool or log directories while checking their
3594 size for availability, Exim confusingly gave the error "space shortage".
3595 Furthermore, in debugging mode it crashed with a floating point exception.
3596 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3597 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3598 serious problem, Exim now writes to the main and panic logs when this
3599 happens, with details of the failure. It then refuses to accept the
3600 incoming message, giving the message "spool directory problem" or "log
3601 directory problem" with a 421 code for SMTP messages.
3603 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3604 0, 1, and 2 exist, because some OS complain for execs without them (see
3605 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3606 descriptors. However, the code omitted to check that the open succeeded,
3607 causing mysterious errors if for some reason the permissions on /dev/null
3608 got screwed. Now Exim writes a message to the main and panic logs, and
3609 bombs out if it can't open /dev/null.
3611 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3612 interact so that it is all more flexible. It is supposed to remain
3613 backwards compatible. Also added extra_local_interfaces.
3615 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3616 to bomb out with an assertion failure - to the client this appears as a
3617 connection drop. This problem occurs in the part of the code that was taken
3618 from the Samba project. Fortunately, the assertion is in a very simple
3619 function, so I have fixed this by reproducing the function inline in the
3620 one place where it is called, and arranging for authentication to fail
3621 instead of killing the process with assert().
3623 75. The SPA client code was not working when the server requested OEM rather
3624 than Unicode encoding.
3626 76. Added code to make require_files with a specific uid setting more usable in
3627 the case where statting the file as root fails - usually a non-root-mounted
3628 NFS file system. When this happens and the failure is EACCES, Exim now
3629 forks a subprocess and does the per-uid checking as the relevant uid.
3631 77. Added process_log_path.
3633 78. If log_file_path was not explicitly set, a setting of check_log_space or
3634 check_log_inodes was ignored.
3636 79. If a space check for the spool or log partitions fails, the incident is now
3637 logged. Of course, in the latter case the data may get lost...
3639 80. Added the %p formatting code to string_format() so that it can be used to
3640 print addresses in debug_print(). Adjusted all the address printing in the
3641 debugging in store.c to use %p rather than %d.
3643 81. There was a concern that a line of code in smtp_in.c could overflow a
3644 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3645 initially expressed, the concern was not well-founded, because trailing
3646 spaces are removed early. However, if the trailing spaces were followed by
3647 a NULL, they did not get removed, so the overflow was possible. Two fixes
3650 (a) I re-wrote the offending code in a cleaner fashion.
3651 (b) If an incoming SMTP command contains a NULL character, it is rejected
3654 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3655 runs initgroups(), so that if the Exim user is in any additional groups,
3656 they will be used during message reception.
3662 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob