1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.275 2005/12/14 10:00:05 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.
47 PH/01 Two changes to the default runtime configuration:
49 (1) Move the checks for relay_from_hosts and authenticated clients from
50 after to before the (commented out) DNS black list checks.
52 (2) Add control=submission to the relay_from_hosts and authenticated
53 clients checks, on the grounds that messages accepted by these
54 statements are most likely to be submissions.
56 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
58 (1) Generate an error if the third argument for the ${prvs expansion is
61 (2) Treat a missing third argument of ${prvscheck as if it were an empty
64 (3) Reset the variables that are obtained from the first argument of
65 ${prvscheck and used in the second argument before leaving the code,
66 because their memory is reclaimed, so using them afterwards may do
69 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
70 one (it's much easier than Tom thought :-).
72 (5) Because of (4), we can now allow for the use of $prvscheck_result
73 inside the third argument.
75 PH/03 For some reason, the default setting of PATH when running a command from
76 a pipe transport was just "/usr/bin". I have changed it to
79 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
80 anything to be listed in the output from -bV.
82 PH/05 When a filter generated an autoreply, the entire To: header line was
83 quoted in the delivery log line, like this:
85 => >A.N.Other <ano@some.domain> <original@ddress> ...
87 This has been changed so that it extracts the operative address. There
88 may be more than one such address. If so, they are comma-separated, like
91 => >ano@some.domain,ona@other.domain <original@ddress> ...
93 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
94 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
95 not being looked up in the rDNS to get a host name, Exim was showing the
96 IP address twice in Received: lines, even though the IP addresses were
97 identical. For example:
99 Received: from [1.2.3.4] (helo=[1.2.3.4])
101 However, if the real host name was known, it was omitting the HELO data
102 if it matched the actual IP address. This has been tidied up so that it
103 doesn't show the same IP address twice.
105 PH/07 When both +timestamp and +memory debugging was on, the value given by
106 $tod_xxx expansions could be wrong, because the tod_stamp() function was
107 called by the debug printing, thereby overwriting the timestamp buffer.
108 Debugging no longer uses the tod_stamp() function when +timestamp is set.
110 PH/08 When the original message was included in an autoreply transport, it
111 always said "this is a copy of the message, including all the headers",
112 even if body_only or headers_only was set. It now gives an appropriate
115 PH/09 Applied a patch from the Sieve maintainer which:
117 o fixes some comments
118 o adds the (disabled) notify extension core
119 o adds some debug output for the result of if/elsif tests
120 o points to the current vacation draft in the documentation
121 and documents the missing references header update
125 o fixes a bug in processing the envelope test (when testing
126 multiple envelope elements, the last element determinted the
129 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
130 Electronic Mail") by including:
132 Auto-submitted: auto-generated
134 in the messages that it generates (bounce messages and others, such as
135 warnings). In the case of bounce messages for non-SMTP mesages, there was
136 also a typo: it was using "Auto_submitted" (underscore instead of
137 hyphen). Since every message generated by Exim is necessarily in response
138 to another message, thes have all been changed to:
140 Auto-Submitted: auto-replied
142 in accordance with these statements in the RFC:
144 The auto-replied keyword:
146 - SHOULD be used on messages sent in direct response to another
147 message by an automatic process,
149 - MUST NOT be used on manually-generated messages,
151 - MAY be used on Delivery Status Notifications (DSNs) and Message
152 Disposition Notifications (MDNs),
154 - MUST NOT be used on messages generated by automatic or periodic
155 processes, except for messages which are automatic responses to
158 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
159 to the default Received: header definition.
161 PH/12 Added log selector acl_warn_skipped (default on).
163 PH/13 After a successful wildlsearch lookup, discard the values of numeric
164 variables because (a) they are in the wrong storage pool and (b) even if
165 they were copied, it wouldn't work properly because of the caching.
167 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
168 checking when decoding. Apparently there are clients that generate
169 overlong encoded strings. Why am I not surprised?
171 PH/15 If the first argument of "${if match_address" was not empty, but did not
172 contain an "@" character, Exim crashed. Now it writes a panic log message
173 and treats the condition as false.
175 PH/16 In autoreply, treat an empty string for "once" the same as unset.
177 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
178 extension "envelope-auth". The code is finished and in agreement with
179 other implementations, but there is no documentation so far and in fact,
180 nobody wrote the draft yet. This extension is currently #undef'ed, thus
181 not changing the active code.
183 Print executed "if" and "elsif" statements when debugging is used. This
184 helps a great deal to understand what a filter does.
186 Document more things not specified clearly in RFC3028. I had all this
187 sorted out, when out of a sudden new issues came to my mind. Oops."
189 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
192 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
193 canonical form (as documented). However, after a host name lookup from
194 the IP address, check_host() was doing a simple string comparison with
195 addresses acquired from the DNS when checking that the found name did
196 have the original IP as one of its addresses. Since any found IPv6
197 addresses are likely to be in abbreviated form, the comparison could
198 fail. Luckily, there already exists a function for doing the comparison
199 by converting both addresses to binary, so now that is used instead of
202 PH/20 There was another similar case to PH/19, when a complete host name was
203 given in a host list; looking up its IP address could give an abbreviated
204 form, whereas the current host's name might or might not be abbreviated.
205 The same fix has been applied.
211 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
212 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
215 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
216 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
218 PH/03 Typo: missing ".o" in src/pcre/Makefile.
220 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
221 header line, restrict the check to what is listed in RFCs 2369 and 2929.
222 Also, for "Auto-Submitted", treat anything other than "no" as
223 non-personal, in accordance with RFC 3834. (Previously it treated
224 anything starting "auto-" as non-personal.)
226 TF/01 The control=submission/name=... option had a problem with syntax
227 errors if the name included a slash character. The /name= option
228 now slurps the rest of the string, so it can include any characters
229 but it must come last in the list of options (after /sender_retain
232 PH/05 Some modifications to the interface to the fake nameserver for the new
235 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
236 into the default Received: header string.
243 TK/01 Added the "success_on_redirect" address verification option. See
244 NewStuff for rationale and an example.
246 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
248 PH/02 Patch to exigrep to allow it to work on syslog lines.
250 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
251 fread() to skip over the body file's header line, because in Cygwin the
252 header line is locked and is inaccessible.
254 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
255 co-exist for some time) to make it clear that it is the Exim ID that is
256 referenced, not the Message-ID: header line.
258 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
259 string_format() function, because snprintf() does not exist on all
262 PH/06 The use of forbid_filter_existstest now also locks out the use of the
263 ${stat: expansion item.
265 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
266 protocol synchronization error", to keep the pedants happy.
268 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
269 well as for IRIX systems, when gcc is being used. See the host.c source
272 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
274 PH/10 Named domain lists were not working if used in a queue_smtp_domains
277 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
278 transport and to the smtp transport in LMTP mode.
280 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
282 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
283 run a filter in a subprocess. This could lead to confusion in subsequent
284 lookups in the parent process. There should also be a search_tidyup() at
285 the end of the subprocess.
287 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
288 only if the host matched helo_try_verify_hosts, which caused the
289 verification to occur when the EHLO/HELO command was issued. The ACL just
290 tested the remembered result. Now, if a previous verification attempt has
291 not happened, "verify = helo" does it there and then.
293 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
295 TK/03 Fix log output including CR from clamd.
297 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
298 exist provoked a memory error which could cause a segfault.
300 PH/15 Installed PCRE 6.2
302 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
304 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
305 of the problem. Specifically, suggested +O2 rather than +O1 for the
308 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
310 PH/20 If a delivery was routed to a non-standard port by means of an SRV
311 record, the port was not correctly logged when the outgoing_port log
312 selector was set (it logged the transort's default port).
314 PH/21 Added support for host-specific ports to manualroute, queryprogram,
315 fallback_hosts, and "hosts" in the smtp transport.
317 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
318 host errors such as "Connection refused".
320 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
321 authentication with radiusclient 0.4.9:
323 - Error returned from rc_read_config was caught wrongly
324 - Username/password not passed on to radius server due to wrong length.
326 The presumption is that some radiusclient API changes for 4.51/PH/17
327 were not taken care of correctly. The code is still untested by me (my
328 Linux distribution still has 0.3.2 of radiusclient), but it was
329 contributed by a Radius user.
331 PH/24 When doing a callout, the value of $domain wasn't set correctly when
332 expanding the "port" option of the smtp transport.
334 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
335 while reading a MIME header. Thanks to Tom Hughes for a patch.
337 PH/24 Include config.h inside local_scan.h so that configuration settings are
340 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
341 that in an ACL for RCPT (for example), you can examine exactly what was
344 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
345 commands, but it was not correctly comparing the address with the actual
346 client host address. Thus, it would show the EHLO address in Received:
347 header lines when this was not necessary.
349 PH/27 Added the % operator to ${eval:}.
351 PH/28 Exim tries to create and chdir to its spool directory when it starts;
352 it should be ignoring failures (because with -C, for example, it has lost
353 privilege). It wasn't ignoring creation failures other than "already
356 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
359 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
360 because an input line was too long, either on its own, or by virtue of
361 too many continuations, the temporary file was not being removed, and the
362 return code was incorrect.
364 PH/31 Missing "BOOL" in function definition in filtertest.c.
366 PH/32 Applied Sieve patches from the maintainer.
368 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
370 PH/33 Added "verify = not_blind".
372 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
373 Local/Makefile (with some defaults set). These are used in built scripts
374 such as exicyclog, but they have never been used in the exim_install
375 script (though there are many overriding facilities there). I have
376 arranged that the exim_install script now takes note of these two
379 PH/35 Installed configuration files for Dragonfly.
381 PH/36 When a locally submitted message by a trusted user did not contain a
382 From: header, and the sender address was obtained from -f or from an SMTP
383 MAIL command, and the trusted user did not use -F to supply a sender
384 name, $originator_name was incorrectly used when constructing a From:
385 header. Furthermore, $originator_name was used for submission mode
386 messages from external hosts without From: headers in a similar way,
387 which is clearly wrong.
389 PH/37 Added control=suppress_local_fixups.
391 PH/38 When log_selector = +received_sender was set, and the addition of the
392 sender made the log line's construction buffer exactly full, or one byte
393 less than full, an overflow happened when the terminating "\n" was
396 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
397 when the result of a list match is failure because a DNS lookup failed.
399 PH/40 RM_COMMAND is now used in the building process.
401 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
402 the "build-* directories that it finds.
404 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
405 address in a domain literal was a prefix of an interface address.
407 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
408 when verifying a sender address, unless rewrite_headers is false.
410 PH/44 Wrote a long comment about why errors_to addresses are verified as
411 recipients, not senders.
413 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
414 the ratelimit ACL was added.
416 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
418 PH/46 Added extra information about PostgreSQL errors to the error string.
420 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
421 avoiding the need to install special zones in a real server. This is
422 backwards compatible; if it can't find the fake resolver, it drops back.
423 Thus, both old and new test suites can be run.
425 TF/02 Added util/ratelimit.pl
427 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
428 clock is set back in time.
430 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
431 Candler <B.Candler@pobox.com>.
433 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
434 OK for addresses that are the result of redirections.
436 PH/48 A number of further additions for the benefit of the new test suite,
437 including a fake gethostbyname() that interfaces to the fake DNS resolver
440 TF/06 The fix for widen_domains has also been applied to qualify_single and
441 search_parents which are the other dnslookup options that can cause
444 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
447 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
449 TF/07 Exim produced the error message "an SRV record indicated no SMTP
450 service" if it encountered an MX record with an empty target hostname.
451 The message is now "an MX or SRV record indicated no SMTP service".
453 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
454 if the DNS of the sending site is misconfigured. This is quite a
455 common situation. This change restores the behaviour of treating a
456 helo verification defer as a failure.
458 PH/51 If self=fail was set on a router, the bounce message did not include the
459 actual error message.
465 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
467 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
468 command itself ended in error, the underlying message about the transport
469 filter timeout was being overwritten with the pipe command error. Now the
470 underlying error message should be appended to the second error message.
472 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
474 PH/02 Reduce the amount of output that "make" produces by default. Full output
475 can still be requested.
477 PH/03 The warning log line about a condition test deferring for a "warn" verb
478 was being output only once per connection, rather than after each
479 occurrence (because it was using the same function as for successful
480 "warn" verbs). This seems wrong, so I have changed it.
482 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
483 it should not have, which might have caused a crash in the right
484 circumstances, but probably never did.
486 PH/04 Installed a modified version of Tony Finch's patch to make submission
487 mode fix the return path as well as the Sender: header line, and to
488 add a /name= option so that you can make the user's friendly name appear
491 TF/03 Added the control = fakedefer ACL modifier.
493 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
494 Mark Lowes for thorough testing.
496 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
498 TK/03 Merged latest SRS patch from Miles Wilton.
500 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
501 with the definition in sysexits.h (which is #included earlier).
502 Fortunately, Exim does not actually use EX_OK. The code used to try to
503 preserve the sysexits.h value, by assumimg that macro definitions were
504 scanned for macro replacements. I have been disabused of this notion,
505 so now the code just undefines EX_OK before #including unistd.h.
507 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
508 in the smtp transport. When a block could not be written in a single
509 write() function, the timeout was being re-applied to each part-write.
510 This seems wrong - if the receiver was accepting one byte at a time it
511 would take for ever. The timeout is now adjusted when this happens. It
512 doesn't have to be particularly precise.
514 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
515 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
517 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
518 without a domain if the check to <postmaster@domain> fails.
520 SC/01 Eximstats: added -xls and the ability to specify output files
521 (patch written by Frank Heydlauf).
523 SC/02 Eximstats: use FileHandles for outputing results.
525 SC/03 Eximstats: allow any combination of xls, txt, and html output.
527 SC/04 Eximstats: fixed display of large numbers with -nvr option
529 SC/05 Eximstats: fixed merging of reports with empty tables.
531 SC/06 Eximstats: added the -include_original_destination flag
533 SC/07 Eximstats: removed tabs and trailing whitespace.
535 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
537 TK/06 MBOX spool code: Add real "From " MBOX separator line
538 so the .eml file is really in mbox format (even though
539 most programs do not really care). Patch from Alex Miller.
541 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
542 The latter is generated from $received_to and is only set if the
543 message has one envelope recipient. SA can use these headers,
544 obviously out-of-the-box. Patch from Alex Miller.
546 PH/08 The ${def test on a variable was returning false if the variable's
547 value was "0", contrary to what the specification has always said!
548 The result should be true unless the variable is empty.
550 PH/09 The syntax error of a character other than { following "${if
551 def:variable_name" (after optional whitespace) was not being diagnosed.
552 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
553 accidental colon was present, for example, could give incorrect results.
555 PH/10 Tidied the code in a number of places where the st_size field of a stat()
556 result is used (not including appendfile, where other changes are about
559 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
560 This involved changing a lot of size variables from int to off_t. It
561 should work with maildirs and everything.
563 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
564 spamd dying while we are connected to it.
566 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
569 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
570 function for BDB 4.3.
572 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
574 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
575 the sources for PCRE have been re-arranged and more files are now
578 PH/15 The code I had for printing potentially long long variables in PH/11
579 above was not the best (it lost precision). The length of off_t variables
580 is now inspected at build time, and an appropriate printing format (%ld
581 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
582 to be "long long int" or "long int". This is needed for the internal
583 formatting function string_vformat().
585 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
586 the configuration file to be ":syslog", then the script "guesses" where
587 the logs files are, rather than using the compiled in default. In our
588 case the guess is not the same as the compiled default, so the script
589 suddenly stopped working when I started to use syslog. The patch checks
590 to see if log_file_path is "". If so, it attempts to read it from exim
591 with no configuration file to get the compiled in version, before it
592 falls back to the previous guessing code."
594 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
595 implementing BATV in an Exim configuration. See NewStuff for the gory
598 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
599 Makefile that are specific to HP-UX.
601 PH/18 If the "use_postmaster" option was set for a recipient callout together
602 with the "random" option, the postmaster address was used as the MAIL
603 FROM address for the random test, but not for the subsequent recipient
604 test. It is now used for both.
606 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
607 patch removes a few documentation additions to RFC 3028, because the
608 latest draft now contains them. It adds the new en;ascii-case comparator
609 and a new error check for 8bit text in MIME parts. Comparator and
610 require names are now matched exactly. I enabled the subaddress
611 extension, but it is not well tested yet (read: it works for me)."
613 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
614 rework some of the code of TK/09 above to avoid the hardwired use of
615 "%lld" and "long long". Replaced the call to snprintf() with a call to
618 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
619 records point to non-existent hosts", "retry timeout exceeded", and
620 "retry time not reached for any host after a long failure period".
622 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
623 experimental DomainKeys support:
625 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
626 (2) On an error such as an illegally used "control", the wrong name for
627 the control was given.
629 These problems did NOT occur unless DomainKeys support was compiled.
631 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
633 PH/24 Added ${if match_ip condition.
635 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
636 databases so that it will be absolutely obvious if a crash occurs in the
637 DB library. This is a regular occurrence (often caused by mis-matched
640 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
641 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
642 picked up on a user's system that detects such things. There doesn't seem
643 to be a gcc warning option for this - only an attribute that has to be
644 put on the function's prototype. It seems that in Fedora Core 4 they have
645 set this on a number of new functions. No doubt there will be more in due
648 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
649 specify a transport. However, if an address that was verified by such a
650 router was the subject of a callout, Exim crashed because it tried to
651 read the rcpt_include_affixes from the non-existent transport. Now it
652 just assumes that the setting of that option is false. This bug was
653 introduced by 4.51/PH/31.
655 PH/28 Changed -d+all to exclude +memory, because that information is very
656 rarely of interest, but it makes the output a lot bigger. People tend to
657 do -d+all out of habit.
659 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
660 code in os-type was giving problems when libc.so lives in lib64, like on
663 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
664 aren't the modern standard, and it seems that some systems' include files
665 don't always have them. Exim was already checking for some of the newer
666 ones like T_AAAA, and defining it itself. I've added checks for all the
667 record types that Exim uses.
669 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
670 not automatically generating a new one, as it is supposed to. This
671 prevented TLS from working. If the file did exist, but contained invalid
672 data, a new version was generated, as expected. It was only the case of a
673 non-existent file that was broken.
675 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
676 with a change in libdomainkeys > 0.64.
678 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
679 from DNS. If the selector record carries the flag, it now has
680 precedence over the domain-wide flag.
682 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
684 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
685 the use of a port name that isn't defined in /etc/services) occurred, the
686 message was deferred as in a normal delivery, and thus remained on the
687 spool, instead of being failed because of the mua_wrapper setting. This
688 is now fixed, and I tidied up some of the mua_wrapper messages at the
691 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
692 the messages in a hash of arrays rather than using individual hashes.
693 This is a bit cleaner and results in dramatic memory savings, albeit
694 at a slight CPU cost.
696 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
697 as requested by Marc Sherman.
699 SC/10 Eximstats: added histograms for user specified patterns as requested
702 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
704 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
705 fopen() in the content-scanning modules that did not already have it.
711 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
712 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
714 TK/02 Fix ACL "control" statement not being available in MIME ACL.
716 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
718 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
719 to test Sieve filters that use "vacation".
721 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
722 that changes the way the GnuTLS parameters are stored in the cache file.
723 The new format can be generated externally. For backward compatibility,
724 if the data in the cache doesn't make sense, Exim assumes it has read an
725 old-format file, and it generates new data and writes a new file. This
726 means that you can't go back to an older release without removing the
729 PH/03 A redirect router that has both "unseen" and "one_time" set does not
730 work if there are any delivery delays because "one_time" forces the
731 parent to be marked "delivered", so its unseen clone is never tried
732 again. For this reason, Exim now forbids the simultaneous setting of
735 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
736 redirected to themselves ("homonym" addresses). Read the long ChangeLog
737 entry if you want to know the details. The fix, however, neglected to
738 consider the case when local delivery batching is involved. The test for
739 "previously delivered" was not happening when checking to see if an
740 address could be batched with a previous (undelivered) one; under
741 certain circumstances this could lead to multiple deliveries to the same
744 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
745 in its include files, and this causes problems building Exim.
747 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
748 header_syntax) but Exim was just ignoring anything given after a slash.
749 In particular, this caused confusion with an attempt to use "verify =
750 reverse_host_lookup/defer_ok". An error is now given when options are
751 supplied for verify items that do not have them. (Maybe reverse_host_
752 lookup should have a defer_ok option, but that's a different point.)
754 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
755 defined by RFC 821) to 2048, because there were problems with some AUTH
756 commands, and RFC 1869 says the size should be increased for extended
757 SMTP commands that take arguments.
759 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
762 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
763 "unknown" error; now it says that the functionality isn't in the binary.
765 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
766 an address' error message when a string expansion fails (syntax or
767 whatever). Otherwise the password may appear in the log. Following change
768 PH/42 below, there is no longer a chance of it appearing in a bounce
771 PH/11 Installed exipick version 20050225.0 from John Jetmore.
773 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
774 of its addresses was ever tried. (Bugzilla bug #2.)
776 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
777 the result incorrectly in the debug output. (It correctly added a newline
778 to what was transported.)
780 TF/01 Added $received_time.
782 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
783 commented out examples of how to interface to a virus scanner and to
784 SpamAssassin. Also added commented examples of av_scanner and
785 spamd_address settings.
787 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
788 and controls are allowed in which ACLs. There were a couple of minor
789 errors. Some of the entries in the conditions table (which is a table of
790 where they are NOT allowed) were getting very unwieldy; rewrote them as a
791 negation of where the condition IS allowed.
793 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
795 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
796 header file does not have a version number, so I've had to invent a new
797 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
798 API. The code is untested by me (my Linux distribution still has 0.3.2 of
799 radiusclient), but it was contributed by a Radius user.
801 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
802 files or directories, for OpenSSL.
804 PH/19 When an Exim process that is running as root has to create an Exim log
805 file, it does so in a subprocess that runs as exim:exim so as to get the
806 ownership right at creation (otherwise, other Exim processes might see
807 the file with the wrong ownership). There was no test for failure of this
808 fork() call, which would lead to the process getting stuck as it waited
809 for a non-existent subprocess. Forks do occasionally fail when resources
810 run out. I reviewed all the other calls to fork(); they all seem to check
813 PH/20 When checking for unexpected SMTP input at connect time (before writing
814 the banner), Exim was not dealing correctly with a non-positive return
815 from the read() function. If the client had disconnected by this time,
816 the result was a log entry for a synchronization error with an empty
817 string after "input=" when read() returned zero. If read() returned -1
818 (an event I could not check), uninitialized data bytes were printed.
819 There were reports of junk text (parts of files, etc) appearing after
822 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
824 PH/22 Added support for macro redefinition, and (re)definition in between
825 driver and ACL definitions.
827 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
828 forgetting to use the resulting value; it was using the unexpanded value.
830 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
831 hadn't been configured. The fix is from Juergen Kreileder, who
832 understands it better than I do:
834 "Here's what I see happening with three configured cyrus_sasl
835 authenticators configured (plain, login, cram-md5):
837 On startup auth_cyrus_sasl_init() gets called for each of these.
838 This means three calls to sasl_listmech() without a specified mech_list.
839 => SASL tests which mechs of all available mechs actually work
840 => three warnings about OTP not working
841 => the returned list contains: plain, login, cram-md5, digest-md5, ...
843 With the patch, sasl_listmech() also gets called three times. But now
844 SASL's mech_list option is set to the server_mech specified in the the
845 authenticator. Or in other words, the answer from sasl_listmech()
846 gets limited to just the mech you're testing for (which is different
848 => the return list contains just 'plain' or 'login', 'cram-md5' or
849 nothing depending on the value of ob->server_mech.
851 I've just tested the patch: Authentication still works fine,
852 unavailable mechs specified in the exim configuration are still
853 caught, and the auth.log warnings about OTP are gone."
855 PH/25 When debugging is enabled, the contents of the command line are added
856 to the debugging output, even when log_selector=+arguments is not
859 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
860 answer is "GNU", and only if the return is "GNU/something" is the answer
863 PH/27 $acl_verify_message is now set immediately after the failure of a
864 verification in an ACL, and so is available in subsequent modifiers. In
865 particular, the message can be preserved by coding like this:
867 warn !verify = sender
868 set acl_m0 = $acl_verify_message
870 Previously, $acl_verify_message was set only while expanding "message"
871 and "log_message" when a very denied access.
873 PH/28 Modified OS/os.c-Linux with
875 -#ifndef OS_LOAD_AVERAGE
876 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
878 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
879 nomenclature these days.)
881 PH/29 Installed patch from the Sieve maintainer that adds the options
882 sieve_useraddress and sieve_subaddress to the redirect router.
884 PH/30 In these circumstances:
885 . Two addresses routed to the same list of hosts;
886 . First host does not offer TLS;
887 . First host accepts first address;
888 . First host gives temporary error to second address;
889 . Second host offers TLS and a TLS session is established;
890 . Second host accepts second address.
891 Exim incorrectly logged both deliveries with the TLS parameters (cipher
892 and peerdn, if requested) that were in fact used only for the second
895 PH/31 When doing a callout as part of verifying an address, Exim was not paying
896 attention to any local part prefix or suffix that was matched by the
897 router that accepted the address. It now behaves in the same way as it
898 does for delivery: the affixes are removed from the local part unless
899 rcpt_include_affixes is set on the transport.
901 PH/32 Add the sender address, as F=<...>, to the log line when logging a
902 timeout during the DATA phase of an incoming message.
904 PH/33 Sieve envelope tests were broken for match types other than :is. I have
905 applied a patch sanctioned by the Sieve maintainer.
907 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
908 the uid or gid is negative. A case of a negative gid caused this to be
909 noticed. The fix allows for either to be negative.
911 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
912 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
915 PH/36 The change PH/12 above was broken. Fixed it.
917 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
918 the grounds that routing the same address twice would always produce the
919 same answer. This might have been true once, but it is certainly no
920 longer true now. Routing a child address may depend on the previous
921 routing that produced that child. Some complicated redirection strategies
922 went wrong when messages had multiple recipients, and made Exim's
923 behaviour dependent on the order in which the addresses were given.
925 I have moved the duplicate checking until after the routing is complete.
926 Exim scans the addresses that are assigned to local and remote
927 transports, and removes any duplicates. This means that more work will be
928 done, as duplicates will always all be routed, but duplicates are
929 presumably rare, so I don't expect this is of any significance.
931 For deliveries to pipes, files, and autoreplies, the duplicate checking
932 still happens during the routing process, since they are not going to be
935 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
936 It corrects a timeout issue with spamd. This is Ian's comment: "The
937 background is that sometimes spamd either never reads data from a
938 connection it has accepted, or it never writes response data. The exiscan
939 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
940 blindly assumes that writes won't block so it may never time out."
942 PH/39 Allow G after quota size as well as K and M.
944 PH/40 The value set for $authenticated_id in an authenticator may not contain
945 binary zeroes or newlines because the value is written to log lines and
946 to spool files. There was no check on this. Now the value is run through
947 the string_printing() function so that such characters are converted to
948 printable escape sequences.
950 PH/41 $message_linecount is a new variable that contains the total number of
951 lines in the message. Compare $body_linecount, which is the count for the
954 PH/42 Exim no longer gives details of delivery errors for specific addresses in
955 bounce and delay warning messages, except in certain special cases, which
958 (a) An SMTP error message from a remote host;
959 (b) A message specified in a :fail: redirection;
960 (c) A message specified in a "fail" command in a system filter;
961 (d) A message specified in a FAIL return from the queryprogram router;
962 (e) A message specified by the cannot_route_message router option.
964 In these cases only, Exim does include the error details in bounce and
965 warning messages. There are also a few cases where bland messages such
966 as "unrouteable address" or "local delivery error" are given.
968 PH/43 $value is now also set for the "else" part of a ${run expansion.
970 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
971 being worked on, but at least Exim now implements the latest version to
974 PH/45 In a pipe transport, although a timeout while waiting for the pipe
975 process to complete was treated as a delivery failure, a timeout while
976 writing the message to the pipe was logged, but erroneously treated as a
977 successful delivery. Such timeouts include transport filter timeouts. For
978 consistency with the overall process timeout, these timeouts are now
979 treated as errors, giving rise to delivery failures by default. However,
980 there is now a new Boolean option for the pipe transport called
981 timeout_defer, which, if set TRUE, converts the failures into defers for
982 both kinds of timeout. A transport filter timeout is now identified in
985 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
986 systems where "make" and "gmake" are different, calling "gmake" at top
987 level broke things. I've arranged for the value of $(MAKE) to be passed
988 from the Makefile to this script so that it can call the same version of
992 A note about Exim versions 4.44 and 4.50
993 ----------------------------------------
995 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
996 changes of various kinds. As a consequence, a big documentation update was
997 needed. This delayed the release for rather longer than seemed good, especially
998 in the light of a couple of (minor) security issues. Therefore, the changes
999 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1000 release. So 4.44 and 4.50 are in effect two different branches that both start
1003 I have left the 4.50 change log unchanged; it contains all the changes since
1004 4.43. The change log for 4.44 is below; many of its items are identical to
1005 those for 4.50. This seems to be the most sensible way to preserve the
1006 historical information.
1012 1. Minor wording change to the doc/README.SIEVE file.
1014 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1015 computation of the current number of files was incorrect.
1017 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1018 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1021 4. Give more explanation in the error message when the command for a transport
1022 filter fails to execute.
1024 5. There are several places where Exim runs a non-Exim command in a
1025 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1026 was being done only for the command run by the queryprogram router. It is
1027 now done for all such subprocesses. The other cases are: ${run, transport
1028 filters, and the commands run by the lmtp and pipe transports.
1030 6. Added CONFIGURE_GROUP build-time option.
1032 7. Some older OS have a limit of 256 on the maximum number of file
1033 descriptors. Exim was using setrlimit() to set 1000 as a large value
1034 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1035 systems. I've change it so that if it can't get 1000, it tries for 256.
1037 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1038 was an oversight, and furthermore, ever since the addition of extra
1039 controls (e.g. 4.43/32), the checks on when to allow different forms of
1040 "control" were broken. There should now be diagnostics for all cases when a
1041 control that does not make sense is encountered.
1043 9. Added the /retain_sender option to "control=submission".
1045 10. $recipients is now available in the predata ACL (oversight).
1047 11. Tidy the search cache before the fork to do a delivery from a message
1048 received from the command line. Otherwise the child will trigger a lookup
1049 failure and thereby defer the delivery if it tries to use (for example) a
1050 cached ldap connection that the parent has called unbind on.
1052 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1053 of $address_data from the recipient verification was clobbered by the
1054 sender verification.
1056 13. The value of address_data from a sender verification is now available in
1057 $sender_address_data in subsequent conditions in the ACL statement.
1059 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1061 15. Added a new option "connect=<time>" to callout options, to set a different
1064 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1065 was its contents. (It was OK if the option was not defined at all.)
1067 17. A "Completed" log line is now written for messages that are removed from
1068 the spool by the -Mrm option.
1070 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1071 information about exactly what failed.
1073 19. Added -dd to debug only the daemon process.
1075 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1076 handles timeouts, both on the server side and network timeouts. Renamed the
1077 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1079 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1080 It is now set to "smtps".
1082 22. $host_address is now set to the target address during the checking of
1083 ignore_target_hosts.
1085 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1086 being passed; this would have caused $sender_host_name to have been used if
1087 matching the list had actually called for a host name (not very likely,
1088 since this list is usually IP addresses). A host name is now passed as
1091 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1092 code that specifies a non-restarting handler (typically sigaction() in
1093 modern systems) in an attempt to fix a rare and obscure crash bug.
1095 25. Narrowed the window for a race in the daemon that could cause it to ignore
1096 SIGCHLD signals. This is not a major problem, because they are used only to
1097 wake it up if nothing else does.
1099 26. A malformed maildirsize file could cause Exim to calculate negative values
1100 for the mailbox size or file count. Odd effects could occur as a result.
1101 The maildirsize information is now recalculated if the size or filecount
1104 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1105 support for a long time. Removed HAVE_SYS_VFS_H.
1107 28. Installed the latest version of exipick from John Jetmore.
1109 29. In an address list, if the pattern was not a regular expression, an empty
1110 subject address (from a bounce message) matched only if the pattern was an
1111 empty string. Non-empty patterns were not even tested. This was the wrong
1112 because it is perfectly reasonable to use an empty address as part of a
1113 database query. An empty address is now tested by patterns that are
1114 lookups. However, all the other forms of pattern expect the subject to
1115 contain a local part and a domain, and therefore, for them, an empty
1116 address still always fails if the pattern is not itself empty.
1118 30. Exim went into a mad DNS loop when attempting to do a callout where the
1119 host was specified on an smtp transport, and looking it up yielded more
1120 than one IP address.
1122 31. Re-factored the code for checking spool and log partition space into a
1123 function that finds that data and another that does the check. The former
1124 is then used to implement four new variables: $spool_space, $log_space,
1125 $spool_inodes, and $log_inodes.
1127 32. The RFC2047 encoding function was originally intended for short strings
1128 such as real names; it was not keeping to the 75-character limit for
1129 encoded words that the RFC imposes. It now respects the limit, and
1130 generates multiple encoded words if necessary. To be on the safe side, I
1131 have increased the buffer size for the ${rfc2047: expansion operator from
1134 33. It is now permitted to omit both strings after an "if" condition; if the
1135 condition is true, the result is "true". As before, when the second string
1136 is omitted, a false condition yields an empty string. This makes it less
1137 cumbersome to write custom ACL and router conditions.
1139 34. Failure to deliver a bounce message always caused it to be frozen, even if
1140 there was an errors_to setting on the router. The errors_to setting is now
1143 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1144 canonical form (fully expanded) before being placed in
1145 $sender_host_address.
1147 36. The table in the code that translates DNS record types into text (T_A to
1148 "A" for instance) was missing entries for NS and CNAME. It is just possible
1149 that this could have caused confusion if both these types were looked up
1150 for the same domain, because the text type is used as part of Exim's
1151 per-process caching. But the chance of anyone hitting this buglet seems
1154 37. The dnsdb lookup has been extended in a number of ways.
1156 (1) There is a new type, "zns", which walks up the domain tree until it
1157 finds some nameserver records. It should be used with care.
1159 (2) There is a new type, "mxh", which is like "mx" except that it returns
1160 just the host names, not the priorities.
1162 (3) It is now possible to give a list of domains (or IP addresses) to be
1163 looked up. The behaviour when one of the lookups defers can be
1164 controlled by a keyword.
1166 (4) It is now possible to specify the separator character for use when
1167 multiple records are returned.
1169 38. The dnslists ACL condition has been extended: it is now possible to supply
1170 a list of IP addresses and/or domains to be looked up in a particular DNS
1173 39. Added log_selector=+queue_time_overall.
1175 40. When running the queue in the test harness, wait just a tad after forking a
1176 delivery process, to get repeatability of debugging output.
1178 41. Include certificate and key file names in error message when GnuTLS fails
1179 to set them up, because the GnuTLS error message doesn't include the name
1180 of the failing file when there is a problem reading it.
1182 42. Allow both -bf and -bF in the same test run.
1184 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1186 44. The "Exiscan patch" is now merged into the mainline Exim source.
1188 45. Sometimes the final signoff response after QUIT could fail to get
1189 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1190 < 0 before doing a fflush(). This bug looks as though it goes back to the
1191 introduction of TLS in release 3.20, but "sometimes" must have been rare
1192 because the tests only now provoked it.
1194 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1195 (this can affect the format of dates).
1197 47. exim_tidydb, when checking for the continued existence of a message for
1198 which it has found a message-specific retry record, was not finding
1199 messages that were in split spool directories. Consequently, it was
1200 deleting retry records that should have stayed in existence.
1202 48. Steve fixed some bugs in eximstats.
1204 49. The SPA authentication driver was not abandoning authentication and moving
1205 on to the next authenticator when an expansion was forced to fail,
1206 contradicting the general specification for all authenticators. Instead it
1207 was generating a temporary error. It now behaves as specified.
1209 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1210 (the order specifies the preference for clients). The order is now AES256,
1211 AES128, 3DES, ARCFOUR128.
1213 51. Small patch to Sieve code - explicitly set From: when generating an
1216 52. Exim crashed if a remote delivery caused a very long error message to be
1217 recorded - for instance if somebody sent an entire SpamAssassin report back
1218 as a large number of 550 error lines. This bug was coincidentally fixed by
1219 increasing the size of one of Exim's internal buffers (big_buffer) that
1220 happened as part of the Exiscan merge. However, to be on the safe side, I
1221 have made the code more robust (and fixed the comments that describe what
1224 53. Now that there can be additional text after "Completed" in log lines (if
1225 the queue_time_overall log selector is set), a one-byte patch to exigrep
1226 was needed to allow it to recognize "Completed" as not the last thing in
1229 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1230 patch that reportedly fixes this has been added. I am not expert enough to
1231 create a test for it. This is what the patch creator wrote:
1233 "I found a little strange behaviour of ldap code when working with
1234 Windows 2003 AD Domain, where users was placed in more than one
1235 Organization Units. When I tried to give exim partial DN, the exit code
1236 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1237 But simultaneously result of request was absolutely normal ldap result,
1238 so I produce this patch..."
1240 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1241 REFERENCE, so I have modified the code to exclude the patch when that macro
1244 55. Some experimental protocols are using DNS PTR records for new purposes. The
1245 keys for these records are domain names, not reversed IP addresses. The
1246 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1247 leaves it alone. Component reversal etc. now happens only for IP addresses.
1250 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1252 57. Double the size of the debug message buffer (to 2048) so that more of very
1253 long debug lines gets shown.
1255 58. The exicyclog utility now does better if the number of log files to keep
1256 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1258 59. Two changes related to the smtp_active_hostname option:
1260 (1) $smtp_active_hostname is now available as a variable.
1261 (2) The default for smtp_banner uses $smtp_active_hostname instead
1262 of $primary_hostname.
1264 60. The host_aton() function is supposed to be passed a string that is known
1265 to be a valid IP address. However, in the case of IPv6 addresses, it was
1266 not checking this. This is a hostage to fortune. Exim now panics and dies
1267 if the condition is not met. A case was found where this could be provoked
1268 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1269 components; fortuitously, this particular loophole had already been fixed
1270 by change 4.50/55 above.
1272 If there are any other similar loopholes, the new check in host_aton()
1273 itself should stop them being exploited. The report I received stated that
1274 data on the command line could provoke the exploit when Exim was running as
1275 exim, but did not say which command line option was involved. All I could
1276 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1277 running as the user.
1280 61. There was a buffer overflow vulnerability in the SPA authentication code
1281 (which came originally from the Samba project). I have added a test to the
1282 spa_base64_to_bits() function which I hope fixes it.
1285 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1286 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1288 63. The daemon start-up calls getloadavg() while still root for those OS that
1289 need the first call to be done as root, but it missed one case: when
1290 deliver_queue_load_max is set with deliver_drop_privilege. This is
1291 necessary for the benefit of the queue runner, because there is no re-exec
1292 when deliver_drop_privilege is set.
1294 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1295 This has been fixed.
1297 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1298 was in use, was not putting the data itself into the right store pool;
1299 consequently, it could be overwritten for a subsequent message in the same
1300 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1303 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1305 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1306 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1307 the function were treating the return as a boolean value, which happened to
1308 work because 0=false and not-0=true, but is not correct code.
1310 68. The host_aton() function was not handling scoped IPv6 addresses (those
1311 with, for example, "%eth0" on the end) correctly.
1313 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1314 negated items (that is, ~something) in unsigned ints. Some compilers
1315 apparently mutter when there is no cast.
1317 70. If an address verification called from an ACL failed, and did not produce a
1318 user-specific message (i.e. there was only a "system" message), nothing was
1319 put in $acl_verify_message. In this situation, it now puts the system
1322 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1323 session; change 4.31/43 added the unwanted input to the log line - except
1324 that it did not do this in the start of session case. It now does.
1326 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1327 This is wrong and can cause the other end to generate a synchronization
1328 error if it is another Exim or anything else that does the synchronization
1329 check. A QUIT command is no longer sent after a timeout.
1331 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1332 during host lookups.
1334 74. The defer_ok option of callout verification was not working if it was used
1335 when verifying addresses in header lines, that is, for this case:
1337 verify = header_sender/callout=defer_ok
1339 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1340 those file descriptors could be used for SMTP connections. If anything
1341 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1342 could be sent to the SMTP client, causing chaos. The daemon now opens
1343 stdin, stdout, and stderr to /dev/null when it puts itself into the
1346 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1347 log by default. The user can override this with suitable Perl magic.
1349 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1350 the log message when discard triggers, was not working for the DATA ACL or
1351 for the non-SMTP ACL.
1353 78. Error message wording change in sieve.c.
1355 79. If smtp_accept_max_per_host was set, the number of connections could be
1356 restricted to fewer than expected, because the daemon was trying to set up
1357 a new connection before checking whether the processes handling previous
1358 connections had finished. The check for completed processes is now done
1359 earlier. On busy systems, this bug wouldn't be noticed because something
1360 else would have woken the daemon, and it would have reaped the completed
1363 80. If a message was submitted locally by a user whose login name contained one
1364 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1365 It caused a spool format error. I have fixed the spool reading code. A
1366 related problem was that the "from" clause in the Received: line became
1367 illegal because of the space(s). It is now covered by ${quote_local_part.
1369 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1372 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1374 83. Patch from Sieve maintainer to fix unterminated string problem in
1375 "vacation" handling.
1377 84. Some minor changes to the Linux configuration files to help with other
1378 OS variants using glibc.
1380 85. One more patch for Sieve to update vacation handling to latest spec.
1383 ----------------------------------------------------
1384 See the note above about the 4.44 and 4.50 releases.
1385 ----------------------------------------------------
1391 1. Change 4.43/35 introduced a bug that caused file counts to be
1392 incorrectly computed when quota_filecount was set in an appendfile
1395 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1396 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1399 3. Give more explanation in the error message when the command for a transport
1400 filter fails to execute.
1402 4. There are several places where Exim runs a non-Exim command in a
1403 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1404 was being done only for the command run by the queryprogram router. It is
1405 now done for all such subprocesses. The other cases are: ${run, transport
1406 filters, and the commands run by the lmtp and pipe transports.
1408 5. Some older OS have a limit of 256 on the maximum number of file
1409 descriptors. Exim was using setrlimit() to set 1000 as a large value
1410 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1411 systems. I've change it so that if it can't get 1000, it tries for 256.
1413 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1414 was an oversight, and furthermore, ever since the addition of extra
1415 controls (e.g. 4.43/32), the checks on when to allow different forms of
1416 "control" were broken. There should now be diagnostics for all cases when a
1417 control that does not make sense is encountered.
1419 7. $recipients is now available in the predata ACL (oversight).
1421 8. Tidy the search cache before the fork to do a delivery from a message
1422 received from the command line. Otherwise the child will trigger a lookup
1423 failure and thereby defer the delivery if it tries to use (for example) a
1424 cached ldap connection that the parent has called unbind on.
1426 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1427 of $address_data from the recipient verification was clobbered by the
1428 sender verification.
1430 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1431 was its contents. (It was OK if the option was not defined at all.)
1433 11. A "Completed" log line is now written for messages that are removed from
1434 the spool by the -Mrm option.
1436 12. $host_address is now set to the target address during the checking of
1437 ignore_target_hosts.
1439 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1440 being passed; this would have caused $sender_host_name to have been used if
1441 matching the list had actually called for a host name (not very likely,
1442 since this list is usually IP addresses). A host name is now passed as
1445 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1446 code that specifies a non-restarting handler (typically sigaction() in
1447 modern systems) in an attempt to fix a rare and obscure crash bug.
1449 15. Narrowed the window for a race in the daemon that could cause it to ignore
1450 SIGCHLD signals. This is not a major problem, because they are used only to
1451 wake it up if nothing else does.
1453 16. A malformed maildirsize file could cause Exim to calculate negative values
1454 for the mailbox size or file count. Odd effects could occur as a result.
1455 The maildirsize information is now recalculated if the size or filecount
1458 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1459 support for a long time. Removed HAVE_SYS_VFS_H.
1461 18. Updated exipick to current release from John Jetmore.
1463 19. Allow an empty sender to be matched against a lookup in an address list.
1464 Previously the only cases considered were a regular expression, or an
1467 20. Exim went into a mad DNS lookup loop when doing a callout where the
1468 host was specified on the transport, if the DNS lookup yielded more than
1471 21. The RFC2047 encoding function was originally intended for short strings
1472 such as real names; it was not keeping to the 75-character limit for
1473 encoded words that the RFC imposes. It now respects the limit, and
1474 generates multiple encoded words if necessary. To be on the safe side, I
1475 have increased the buffer size for the ${rfc2047: expansion operator from
1478 22. Failure to deliver a bounce message always caused it to be frozen, even if
1479 there was an errors_to setting on the router. The errors_to setting is now
1482 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1483 canonical form (fully expanded) before being placed in
1484 $sender_host_address.
1486 24. Updated eximstats to version 1.33
1488 25. Include certificate and key file names in error message when GnuTLS fails
1489 to set them up, because the GnuTLS error message doesn't include the name
1490 of the failing file when there is a problem reading it.
1492 26. Expand error message when OpenSSL has problems setting up cert/key files.
1495 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1496 (this can affect the format of dates).
1498 28. exim_tidydb, when checking for the continued existence of a message for
1499 which it has found a message-specific retry record, was not finding
1500 messages that were in split spool directories. Consequently, it was
1501 deleting retry records that should have stayed in existence.
1503 29. eximstats updated to version 1.35
1504 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1505 1.35 - bugfix such that pie charts by volume are generated correctly
1507 30. The SPA authentication driver was not abandoning authentication and moving
1508 on to the next authenticator when an expansion was forced to fail,
1509 contradicting the general specification for all authenticators. Instead it
1510 was generating a temporary error. It now behaves as specified.
1512 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1513 (the order specifies the preference for clients). The order is now AES256,
1514 AES128, 3DES, ARCFOUR128.
1516 31. Small patch to Sieve code - explicitly set From: when generating an
1519 32. Exim crashed if a remote delivery caused a very long error message to be
1520 recorded - for instance if somebody sent an entire SpamAssassin report back
1521 as a large number of 550 error lines. This bug was coincidentally fixed by
1522 increasing the size of one of Exim's internal buffers (big_buffer) that
1523 happened as part of the Exiscan merge. However, to be on the safe side, I
1524 have made the code more robust (and fixed the comments that describe what
1527 33. Some experimental protocols are using DNS PTR records for new purposes. The
1528 keys for these records are domain names, not reversed IP addresses. The
1529 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1530 leaves it alone. Component reversal etc. now happens only for IP addresses.
1533 34. The host_aton() function is supposed to be passed a string that is known
1534 to be a valid IP address. However, in the case of IPv6 addresses, it was
1535 not checking this. This is a hostage to fortune. Exim now panics and dies
1536 if the condition is not met. A case was found where this could be provoked
1537 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1538 components; fortuitously, this particular loophole had already been fixed
1539 by change 4.50/55 or 4.44/33 above.
1541 If there are any other similar loopholes, the new check in host_aton()
1542 itself should stop them being exploited. The report I received stated that
1543 data on the command line could provoke the exploit when Exim was running as
1544 exim, but did not say which command line option was involved. All I could
1545 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1546 running as the user.
1549 35. There was a buffer overflow vulnerability in the SPA authentication code
1550 (which came originally from the Samba project). I have added a test to the
1551 spa_base64_to_bits() function which I hope fixes it.
1554 36. The daemon start-up calls getloadavg() while still root for those OS that
1555 need the first call to be done as root, but it missed one case: when
1556 deliver_queue_load_max is set with deliver_drop_privilege. This is
1557 necessary for the benefit of the queue runner, because there is no re-exec
1558 when deliver_drop_privilege is set.
1560 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1561 was in use, was not putting the data itself into the right store pool;
1562 consequently, it could be overwritten for a subsequent message in the same
1563 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1566 38. Sometimes the final signoff response after QUIT could fail to get
1567 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1568 < 0 before doing a fflush(). This bug looks as though it goes back to the
1569 introduction of TLS in release 3.20, but "sometimes" must have been rare
1570 because the tests only now provoked it.
1576 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1577 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1578 Now it returns an integer. A number of calls were still expecting a T/F
1579 return. Fortuitously, in all cases, the tests worked in OK situations,
1580 which is the norm. However, things would have gone wrong on any write
1581 failures on the smtp file descriptor. This function is used when sending
1582 messages over SMTP and also when doing verify callouts.
1584 2. When Exim is called to do synchronous delivery of a locally submitted
1585 message (the -odf or -odi options), it no longer closes stderr before doing
1588 3. Implemented the mua_wrapper option.
1590 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1592 5. Implemented the functions header_remove(), header_testname(),
1593 header_add_at_position(), and receive_remove_recipient(), and exported them
1596 6. If an ACL "warn" statement specified the addition of headers, Exim already
1597 inserted X-ACL-Warn: at the start if there was no header name. However, it
1598 was not making this test for the second and subsequent header lines if
1599 there were newlines in the string. This meant that an invalid header could
1600 be inserted if Exim was badly configured.
1602 7. Allow an ACL "warn" statement to add header lines at the start or after all
1603 the Received: headers, as well as at the end.
1605 8. Added the rcpt_4xx retry error code.
1607 9. Added postmaster_mailfrom=xxx to callout verification option.
1609 10. Added mailfrom=xxxx to the callout verification option, for verify=
1612 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1613 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1615 12. Inserted some casts to stop certain compilers warning when using pointer
1616 differences as field lengths or precisions in printf-type calls (mostly
1617 affecting debugging statements).
1619 13. Added optional readline() support for -be (dynamically loaded).
1621 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1622 same clock tick as a message's arrival, so that its received time was the
1623 same as the "first fail" time on the retry record, and that message
1624 remained on the queue past the ultimate address timeout, every queue runner
1625 would try a delivery (because it was past the ultimate address timeout) but
1626 after another failure, the ultimate address timeout, which should have then
1627 bounced the address, did not kick in. This was a "< instead of <=" error;
1628 in most cases the first failure would have been in the next clock tick
1629 after the received time, and all would be well.
1631 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1632 being recognized when the domain list was tested by the match_domain
1633 condition in an expansion string.
1635 16. Added the ${str2b64: operator.
1637 17. Exim was always calling setrlimit() to set a large limit for the number of
1638 processes, without checking whether the existing limit was already
1639 adequate. (It did check for the limit on file descriptors.) Furthermore,
1640 errors from getrlimit() and setrlimit() were being ignored. Now they are
1641 logged to the main and panic logs, but Exim does carry on, to try to do its
1642 job under whatever limits there are.
1644 18. Imported PCRE 5.0.
1646 19. Trivial typo in log message " temporarily refused connection" (the leading
1649 20. If the log selector return_path_on_delivery was set and an address was
1650 redirected to /dev/null, the delivery process crashed because it assumed
1651 that a return path would always be set for a "successful" delivery. In this
1652 case, the whole delivery is bypassed as an optimization, and therefore no
1655 21. Internal re-arrangement: the function for sending a challenge and reading
1656 a response while authentication was assuming a zero-terminated challenge
1657 string. It's now changed to take a pointer and a length, to allow for
1658 binary data in such strings.
1660 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1662 23. Exim was not respecting finduser_retries when seeking the login of the
1663 uid under which it was called; it was always trying 10 times. (The default
1664 setting of finduser_retries is zero.) Also, it was sleeping after the final
1665 failure, which is pointless.
1667 24. Implemented tls_on_connect_ports.
1669 25. Implemented acl_smtp_predata.
1671 26. If the domain in control=submission is set empty, Exim assumes that the
1672 authenticated id is a complete email address when it generates From: or
1673 Sender: header lines.
1675 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1676 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1677 chown and chgrp in /bin and hostname in /usr/bin.
1679 28. Exim was keeping the "process log" file open after each use, just as it
1680 does for the main log. This opens the possibility of it remaining open for
1681 long periods when the USR1 signal hits a daemon. Occasional processlog
1682 errors were reported, that could have been caused by this. Anyway, it seems
1683 much more sensible not to leave this file open at all, so that is what now
1686 29. The long-running daemon process does not normally write to the log once it
1687 has entered its main loop, and it closes the log before doing so. This is
1688 so that log files can straightforwardly be renamed and moved. However,
1689 there are a couple of unusual error situations where the daemon does write
1690 log entries, and I had neglected to close the log afterwards.
1692 30. The text of an SMTP error response that was received during a remote
1693 delivery was being truncated at 512 bytes. This is too short for some of
1694 the long messages that one sometimes sees. I've increased the limit to
1697 31. It is now possible to make retry rules that apply only when a message has a
1698 specific sender, in particular, an empty sender.
1700 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1701 it possible to be selective about when SMTP synchronization is enforced.
1703 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1705 32. Implemented hosts_connection_nolog.
1707 33. Added an ACL for QUIT.
1709 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1712 35. Added mailbox_size and mailbox_filecount to appendfile.
1714 36. Added control = no_multiline_responses to ACLs.
1716 37. There was a bug in the logic of the code that waits for the clock to tick
1717 in the case where the clock went backwards by a substantial amount such
1718 that the microsecond fraction of "now" was more than the microsecond
1719 fraction of "then" (but the whole seconds number was less).
1721 38. Added support for the libradius Radius client library this is found on
1722 FreeBSD (previously only the radiusclient library was supported).
1728 1. When certain lookups returned multiple values in the form name=value, the
1729 quoting of the values was not always being done properly. Specifically:
1730 (a) If the value started with a double quote, but contained no whitespace,
1732 (b) If the value contained whitespace other than a space character (i.e.
1733 tabs or newlines or carriage returns) it was not quoted.
1734 This fix has been applied to the mysql and pgsql lookups by writing a
1735 separate quoting function and calling it from the lookup code. The fix
1736 should probably also be applied to nisplus, ibase and oracle lookups, but
1737 since I cannot test any of those, I have not disturbed their existing code.
1739 2. A hit in the callout cache for a specific address caused a log line with no
1740 reason for rejecting RCPT. Now it says "Previous (cached) callout
1741 verification failure".
1743 3. There was an off-by-one bug in the queryprogram router. An over-long
1744 return line was truncated at 256 instead of 255 characters, thereby
1745 overflowing its buffer with the terminating zero. As well as fixing this, I
1746 have increased the buffer size to 1024 (and made a note to document this).
1748 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1749 when Exim is waiting for an SMTP response from a remote server, Exim
1750 restarts its select() call on the socket, thereby resetting its timeout.
1751 This is not a problem when such interrupts are rare. Somebody set up a cron
1752 job to run exiwhat every 2 minutes, which is less than the normal select()
1753 timeout (5 or 10 minutes). This meant that the select() timeout never
1754 kicked in because it was always reset. I have fixed this by comparing the
1755 time when an interrupt arrives with the time at the start of the first call
1756 to select(). If more time than the timeout has elapsed, the interrupt is
1757 treated as a timeout.
1759 5. Some internal re-factoring in preparation for the addition of Sieve
1760 extensions (by MH). In particular, the "personal" test is moved to a
1761 separate function, and given an option for scanning Cc: and Bcc: (which is
1762 not set for Exim filters).
1764 6. When Exim created an email address using the login of the caller as the
1765 local part (e.g. when creating a From: or Sender: header line), it was not
1766 quoting the local part when it contained special characters such as @.
1768 7. Installed new OpenBSD configuration files.
1770 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1771 try to make them clearer.
1773 9. Callout options, other than the timeout value, were being ignored when
1774 verifying sender addresses in header lines. For example, when using
1776 verify = header_sender/callout=no_cache
1778 the cache was (incorrectly) being used.
1780 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1781 only the Cygwin environment.
1783 11. When return_path_on_delivery was set as a log selector, if different remote
1784 addresses in the same message used different return paths and parallel
1785 remote delivery occurred, the wrong values would sometimes be logged.
1786 (Whenever a remote delivery process finished, the return path value from
1787 the most recently started remote delivery process was logged.)
1789 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1790 lines when AUTH and/or TLS are in use. This is the "received protocol"
1791 field. Exim used to use "asmtp" for authenticated SMTP, without any
1792 indication (in the protocol name) for TLS use. Now it follows the RFC and
1793 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1794 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1795 names appear in log lines as well as in Received: header lines.
1797 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1798 extensions, and comparison tests, and to fix some bugs.
1800 14. Changes to the "personal" filter test:
1802 (1) The test was buggy in that it was just doing the equivalent of
1803 "contains" tests on header lines. For example, if a user's address was
1804 anne@some.where, the "personal" test would incorrectly be true for
1806 To: susanne@some.where
1808 This test is now done by extracting each address from the header in turn,
1809 and checking the entire address. Other tests that are part of "personal"
1810 are now done using regular expressions (for example, to check local parts
1811 of addresses in From: header lines).
1813 (2) The list of non-personal local parts in From: addresses has been
1814 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1815 taken from the Sieve specification recommendations.
1817 (3) If the message contains any header line starting with "List-" it is
1818 treated as non-personal.
1820 (4) The test for "circular" in the Subject: header line has been removed
1821 because it now seems ill-conceived.
1823 15. Minor typos in src/EDITME comments corrected.
1825 16. Installed latest exipick from John Jetmore.
1827 17. If headers_add on a router specified a text string that was too long for
1828 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1829 of string_sprintf() is now avoided.
1831 18. $message_body_size was not set (it was always zero) when running the DATA
1832 ACL and the local_scan() function.
1834 19. For the "mail" command in an Exim filter, no default was being set for
1835 the once_repeat time, causing a random time value to be used if "once" was
1836 specified. (If the value happened to be <= 0, no repeat happened.) The
1837 default is now 0s, meaning "never repeat". The "vacation" command was OK
1838 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1839 (I found it when inspecting the code).
1841 20. There is now an overall timeout for performing a callout verification. It
1842 defaults to 4 times the callout timeout, which applies to individual SMTP
1843 commands during the callout. The overall timeout applies when there is more
1844 than one host that can be tried. The timeout is checked before trying the
1845 next host. This prevents very long delays if there are a large number of
1846 hosts and all are timing out (e.g. when the network connections are timing
1847 out). The value of the overall timeout can be changed by specifying an
1848 additional sub-option for "callout", called "maxwait". For example:
1850 verify = sender/callout=5s,maxwait=20s
1852 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1853 to the end before writing, but this should make it even safer).
1855 22. Exim was forgetting that it had advertised PIPELINING for the second and
1856 subsequent messages on an SMTP connection. It was also not resetting its
1857 memory on STARTTLS and an internal HELO.
1859 23. When Exim logs an SMTP synchronization error within a session, it now
1860 records whether PIPELINING has been advertised or not.
1862 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1863 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1864 rather than long int.
1866 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1868 26. Added the never_mail option to autoreply.
1874 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1875 crash if the getsockname() call failed; this can happen if a connection is
1876 closed very soon after it is established. The problem was simply in the
1877 order in which certain operations were done, causing Exim to try to write
1878 to the SMTP stream before it had set up the file descriptor. The bug has
1879 been fixed by making things happen in the correct order.
1885 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1886 before the connection was closed, thus losing the rejection response.
1888 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1889 some early Solaris releases, but causes trouble in current releases where
1890 socklen_t is defined.
1892 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1895 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1898 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1899 that releasing the top of it at the end releases what was used for sub-
1900 expansions (unless the block got too big). However, discard this block if
1901 the first thing is a variable or header, so that we can use its block when
1902 it is dynamic (useful for very large $message_headers, for example).
1904 6. Lookups now cache *every* query, not just the most recent. A new, separate
1905 store pool is used for this. It can be recovered when all lookup caches are
1906 flushed. Lookups now release memory at the end of their result strings.
1907 This has involved some general refactoring of the lookup sources.
1909 7. Some code has been added to the store_xxx() functions to reduce the amount
1910 of flapping under certain conditions.
1912 8. log_incoming_interface used to affect only the <= reception log lines. Now
1913 it causes the local interface and port to be added to several more SMTP log
1914 lines, for example "SMTP connection from", and rejection lines.
1916 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1918 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1920 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1921 could be overwritten at the end of the current message (or the start of a
1922 new message if it was set in a HELO ACL). The value is now preserved for
1923 the duration of the SMTP connection.
1925 12. If a transport had a headers_rewrite setting, and a matching header line
1926 contained an unqualified address, that address was qualified, even if it
1927 did not match any rewriting rules. The underlying bug was that the values
1928 of the flags that permit the existence of unqualified sender and recipient
1929 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1930 non-local messages, and by -bnq for local messages) were not being
1931 preserved with the message after it was received.
1933 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1934 "next input=" as part of the text comprising the host identity instead of
1935 the correct text. The code was using the same buffer for two different
1936 strings. However, depending on which order the printing function evaluated
1937 its arguments, the bug did not always show up. Under Linux, for example, my
1938 test suite worked just fine.
1940 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1941 doesn't work with some older versions of Perl. It has been changed to "my",
1942 which in any case is probably the better facility to use.
1944 15. A really picky compiler found some instances of statements for creating
1945 error messages that either had too many or two few arguments for the format
1948 16. The size of the buffer for calls to the DNS resolver has been increased
1949 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1950 for addresses that have a lot of PTR records. This alleviates a problem; it
1951 does not fully solve it.
1953 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1954 buffer now truncates the list and logs the incident, which is the same
1955 action as happens when Exim is looking up a host name and its aliases.
1956 Previously in this situation something unpredictable would happen;
1957 sometimes it was "internal error: store_reset failed".
1959 18. If a server dropped the connection unexpectedly when an Exim client was
1960 using GnuTLS and trying to read a response, the client delivery process
1961 crashed while trying to generate an error log message.
1963 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1964 string, for example:
1966 warn message = H1: something\nH2: something
1968 the text was added as a single header line from Exim's point of view
1969 though it ended up OK in the delivered message. However, searching for the
1970 second and subsequent header lines using $h_h2: did not work. This has been
1971 fixed. Similarly, if a system filter added multiple headers in this way,
1972 the routers could not see them.
1974 20. Expanded the error message when iplsearch is called with an invalid key to
1975 suggest using net-iplsearch in a host list.
1977 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1978 ACLs are no longer actually imposed (and a message to that effect is
1981 22. If a "gecos" field in a passwd entry contained escaped characters, in
1982 particular, if it contained a \" sequence, Exim got it wrong when building
1983 a From: or a Sender: header from that name. A second bug also caused
1984 incorrect handling when an unquoted " was present following a character
1985 that needed quoting.
1987 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1988 was not being matched caselessly.
1990 24. Arranged for all hyphens in the exim.8 source to be escaped with
1993 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1994 properly. Recipient callout cache records were still being keyed to include
1995 the sender, even when use_sender was set false. This led to far more
1996 callouts that were necessary. The sender is no longer included in the key
1997 when use_sender is false.
1999 26. Added "control = submission" modifier to ACLs.
2001 27. Added the ${base62d: operator to decode base 62 numbers.
2003 28. dnsdb lookups can now access SRV records.
2005 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2006 the configuration file.
2008 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2009 (-v) mode. This makes the output for a verbose queue run more intelligible.
2011 31. Added a use_postmaster feature to recipient callouts.
2013 32. Added the $body_zerocount variable, containing the number of binary zero
2014 bytes in the message body.
2016 33. The time of last modification of the "new" subdirectory is now used as the
2017 "mailbox time last read" when there is a quota error for a maildir
2020 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2022 35. Added +ignore_unknown as a special item in host lists.
2024 36. Code for decoding IPv6 addresses in host lists is now included, even if
2025 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2026 address was recognized as an IP address, but was then not correctly decoded
2027 into binary, causing unexpected and incorrect effects when compared with
2034 1. Very minor rewording of debugging text in manualroute to say "list of
2035 hosts" instead of "hostlist".
2037 2. If verify=header_syntax was set, and a header line with an unqualified
2038 address (no domain) and a large number of spaces between the end of the
2039 name and the colon was received, the reception process suffered a buffer
2040 overflow, and (when I tested it) crashed. This was caused by some obsolete
2041 code that should have been removed. The fix is to remove it!
2043 3. When running in the test harness, delay a bit after writing a bounce
2044 message to get a bit more predictability in the log output.
2046 4. Added a call to search_tidyup() just before forking a reception process. In
2047 theory, someone could use a lookup in the expansion of smtp_accept_max_
2048 per_host which, without the tidyup, could leave open a database connection.
2050 5. Added the variables $recipient_data and $sender_data which get set from a
2051 lookup success in an ACL "recipients" or "senders" condition, or a router
2052 "senders" option, similar to $domain_data and $local_part_data.
2054 6. Moved the writing of debug_print from before to after the "senders" test
2057 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2058 problems for message scanning, either using a data ACL, or using
2059 local_scan() because the Received: header was not generated till after they
2060 were called (in order to set the time as the time of reception completion).
2061 I have revised the way this works. The header is now generated after the
2062 body is received, but before the ACL or local_scan() are called. After they
2063 are run, the timestamp in the header is updated.
2069 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2070 before starting a queue runner without re-exec. This happened only when
2071 deliver_drop_privilege was set or when the Exim user was set to root. The
2072 effect of the bug was that timeouts during subsequent deliveries caused
2073 crashes instead of being properly handled. The handler is now left at its
2074 default (and expected) setting.
2076 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2077 message, again when deliver_drop_privilege is set or Exim is run as root.
2078 The bug described in (1) was not present in this case, but the tidying up
2079 of the other signals was missing. I have made the two cases consistent.
2081 3. The ignore_target_hosts setting on a manualroute router was being ignored
2082 for hosts that were looked up using the /MX notation.
2084 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2087 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2088 operated on the sender address. After changing the $sender_address to <>
2089 for the sender address verify, Exim was re-instated it as the original
2090 (before rewriting) address, but remembering that it had rewritten it, so it
2091 wasn't rewriting it again. This bug also had the effect of breaking the
2092 sender address verification caching when the sender address was rewritten.
2094 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2095 This has been changed so that if the ip literal address matches
2096 ignore_target_hosts, the router declines.
2098 7. Added expansion conditions match_domain, match_address, and match_local_
2099 part (NOT match_host).
2101 8. The placeholder for the Received: header didn't have a length field set.
2103 9. Added code to Exim itself and to exim_lock to test for a specific race
2104 condition that could lead to file corruption when using MBX delivery. The
2105 issue is with the lockfile that is created in /tmp. If this file is removed
2106 after a process has opened it but before that process has acquired a lock,
2107 there is the potential for a second process to recreate the file and also
2108 acquire a lock. This could lead to two Exim processes writing to the file
2109 at the same time. The added code performs the same test as UW imapd; it
2110 checks after acquiring the lock that its file descriptor still refers to
2111 the same named file.
2113 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2114 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2117 11. Added the smtp_active_hostname option. If used, this will typically be made
2118 to depend on the incoming interface address. Because $interface_address is
2119 not set up until the daemon has forked a reception process, error responses
2120 that can happen earlier (such as "too many connections") no longer contain
2123 12. If an expansion in a condition on a "warn" statement fails because a lookup
2124 defers, the "warn" statement is abandoned, and the next ACL statement is
2125 processed. Previously this caused the whole ACL to be aborted.
2127 13. Added the iplsearch lookup type.
2129 14. Added ident_timeout as a log selector.
2131 15. Added tls_certificate_verified as a log selector.
2133 16. Added a global option tls_require_ciphers (compare the smtp transport
2134 option of the same name). This controls incoming TLS connections.
2136 17. I finally figured out how to make tls_require_ciphers do a similar thing
2137 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2138 before starting the TLS session.
2140 18. Tabs are now shown as \t in -bP output.
2142 19. If the log selector return_path_on_delivery was set, Exim crashed when
2143 bouncing a message because it had too many Received: header lines.
2145 20. If two routers both had headers_remove settings, and the first one included
2146 a superfluous trailing colon, the final name in the first list and the
2147 first name in the second list were incorrectly joined into one item (with a
2148 colon in the middle).
2154 1. Added -C and -D options to the exinext utility, mainly to make it easier
2155 to include in the automated testing, but these could be helpful when
2156 multiple configurations are in use.
2158 2. The exinext utility was not formatting the output nicely when there was
2159 an alternate port involved in the retry record key, nor when there was a
2160 message id as well (for retries that were specific to a specific message
2161 and a specific host). It was also confused by IPv6 addresses, because of
2162 the additional colons they contain. I have fixed the IPv4 problem, and
2163 patched it up to do a reasonable job for IPv6.
2165 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2166 delivery, the log line now contains "pipelined" if PIPELINING was used.
2168 4. An SMTP transport process used to panic and die if the bind() call to set
2169 an explicit outgoing interface failed. This has been changed; it is now
2170 treated in the same way as a connect() failure.
2172 5. A reference to $sender_host_name in the part of a conditional expansion
2173 that was being skipped was still causing a DNS lookup. This no longer
2176 6. The def: expansion condition was not recognizing references to header lines
2177 that used bh_ and bheader_.
2179 7. Added the _cache feature to named lists.
2181 8. The code for checking quota_filecount in the appendfile transport was
2182 allowing one more file than it should have been.
2184 9. For compatibility with Sendmail, the command line option
2192 and sets the incoming protocol and host name (for trusted callers). The
2193 host name and its colon can be omitted when only the protocol is to be set.
2194 Note the Exim already has two private options, -pd and -ps, that refer to
2195 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2196 "s", but I don't think that's a major issue.
2198 10. A number of refactoring changes to the code, none of which should affect
2201 (a) The number of logging options was getting close to filling up the
2202 32-bit word that was used as a bit map. I have split them into two classes:
2203 those that are passed in the argument to log_write(), and those that are
2204 only ever tested independently outside of that function. These are now in
2205 separate 32-bit words, so there is plenty of room for expansion again.
2206 There is no change in the user interface or the logging behaviour.
2208 (b) When building, for example, log lines, the code previously used a
2209 macro that called string_cat() twice, in order to add two strings. This is
2210 not really sufficiently general. Furthermore, there was one instance where
2211 it was actually wrong because one of the argument was used twice, and in
2212 one call a function was used. (As it happened, calling the function twice
2213 did not affect the overall behaviour.) The macro has been replaced by a
2214 function that can join an arbitrary number of extra strings onto a growing
2217 (c) The code for expansion conditions now uses a table and a binary chop
2218 instead of a serial search (which was left over from when there were very
2219 few conditions). Also, it now recognizes conditions like "pam" even when
2220 the relevant support is not compiled in: a suitably worded error message is
2221 given if an attempt is made to use such a condition.
2223 11. Added ${time_interval:xxxxx}.
2225 12. A bug was causing one of the ddress fields not to be passed back correctly
2226 from remote delivery subprocesses. The field in question was not being
2227 subsequently used, so this caused to problems in practice.
2229 13. Added new log selectors queue_time and deliver_time.
2231 14. Might have fixed a bug in maildirsizefile handling that threw up
2232 "unexpected character" debug warnings, and recalculated the data
2233 unnecessarily. In any case, I expanded the warning message to give more
2236 15. Added the message "Restricted characters in address" to the statements in
2237 the default ACL that block characters like @ and % in local parts.
2239 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2240 Three changes have been made:
2242 (a) There was a serious bug; a negative response to MAIL caused the whole
2243 recipient domain to be cached as invalid, thereby blocking all messages
2244 to all local parts at the same domain, from all senders. This bug has
2245 been fixed. The domain is no longer cached after a negative response to
2246 MAIL if the sender used is not empty.
2248 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2251 (c) A new callout option, "use_sender" has been added for people who want
2252 the modified behaviour.
2258 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2261 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2262 indeed breaks things for older releases.
2264 3. Added additional logging to the case where there is a problem reading data
2265 from a filter that is running in a subprocess using a pipe, in order to
2266 try to track down a specific problem.
2268 4. Testing facility fudge: when running in the test harness and attempting
2269 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2270 getting "No route to host". Convert this to a timeout.
2272 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2275 6. Some OS don't have socklen_t but use size_t instead. This affects the
2276 fifth argument of getsockopt() amongst other things. This is now
2277 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2278 can be set for individual OS. I have set it for SunOS5, OSF1, and
2279 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2280 some earlier ones do not.
2282 7. Change 4.30/15 was not doing the test caselessly.
2284 8. The standard form for an IPv6 address literal was being rejected by address
2285 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2286 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2287 this, as well as the form without the "IPv6" on the front (but only when
2288 address literals are enabled, of course).
2290 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2292 10. Exim crashed if a message with an empty sender address specified by -f
2293 encountered a router with an errors_to setting. This could be provoked only
2294 by a command such as
2298 where an empty string was supplied; "<>" did not hit this bug.
2300 11. Installed PCRE release 4.5.
2302 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2303 remained set. It is now erased.
2305 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2306 times from message ids (which are base 36 rather than the normal 62).
2308 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2309 were being counted as actual protocol errors, and logged if the log
2310 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2311 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2312 and DATA following a set of rejected RCPTs do not count as protocol errors.
2313 In other words, Exim assumes they were pipelined, though this may not
2314 actually be the case. Of course, in all cases the client gets an
2315 appropriate error code.
2317 15. If a lookup fails in an ACL condition, a message about the failure may
2318 be available; it is used if testing the ACL cannot continue, because most
2319 such messages specify what the cause of the deferral is. However, some
2320 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2321 that caused an old message to be retained and used if a later statement
2322 caused a defer, replacing the real cause of the deferral.
2324 16. If an IP address had so many PTR records that the DNS lookup buffer
2325 was not large enough to hold them, Exim could crash while trying to process
2326 the truncated data. It now detects and logs this case.
2328 17. Further to 4.21/58, another change has been made: if (and only if) the
2329 first line of a message (the first header line) ends with CRLF, a bare LF
2330 in a subsequent header line has a space inserted after it, so as not to
2331 terminate the header.
2333 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2334 unnecessarily, used atoi() instead of strtol(), and didn't check the
2335 termination when getting file sizes from file names by regex.
2337 19. Completely re-implemented the support for maildirsize files, in the light
2338 of a number of problems with the previous contributed implementation
2339 (4.30/29). In particular:
2341 . If the quota is zero, the maildirsize file is maintained, but no quota is
2344 . If the maildir directory does not exist, it is created before any attempt
2345 to write a maildirsize file.
2347 . The quota value in the file is just a cache; if the quota is changed in
2348 the transport, the new value overrides.
2350 . A regular expression is available for excluding directories from the
2353 20. The autoreply transport checks the characters in options that define the
2354 message's headers; it allows continued headers, but it was checking with
2355 isspace() after an embedded newline instead of explicitly looking for a
2358 21. If all the "regular" hosts to which an address was routed had passed their
2359 expiry times, and had not reached their retry times, the address was
2360 bounced, even if fallback hosts were defined. Now Exim should go on to try
2363 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2364 equivalent code in the SMTP transport. Some hosts send humungous responses
2365 to HELO/EHLO, more than 1024 it seems.
2367 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2368 gives compiler warnings in some environments. I've now done it "properly",
2371 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2372 (because of problems with the built-in one) was declared to return uschar *
2373 instead of char *, causing compiler failure.
2375 25. Fixed a file descriptor leak when processing alias/forward files.
2377 26. Fixed a minor format string issue in dbfn.c.
2379 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2381 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2382 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2385 29. When a sender address is verified, it is cached, to save repeating the test
2386 when there is more than one recipient in a message. However, when the
2387 verification involves a callout, it is possible for different callout
2388 options to be set for different recipients. It is too complicated to keep
2389 track of this in the cache, so now Exim always runs a verification when a
2390 callout is required, relying on the callout cache for the optimization.
2391 The overhead is duplication of the address routing, but this should not be
2394 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2395 to be verified with callout=postmaster, and the main callout worked but the
2396 postmaster check failed, the verification correctly failed. However, if a
2397 subsequent RCPT command asked for sender verification *without* the
2398 postmaster check, incorrect caching caused this verification also to fail,
2401 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2402 it was not caching the DNS options (qualify_single, search_parents) that
2403 were used when the lookup failed. A subsequent lookup with different
2404 options therefore always gave the same answer, though there were cases
2405 where it should not have. (Example: a "domains = !$mx_any" option on a
2406 dnslookup router: the "domains" option is always processed without any
2407 widening, but the router might have qualify_single set.) Now Exim uses the
2408 cached value only when the same options are set.
2410 32. Added John Jetmore's "exipick" utility to the distribution.
2412 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2413 than a timeout (e.g. a certificate is required, and is not provided), an
2414 Exim server now closes the connection immediately. Previously it waited for
2415 the client to close - but if the client is SSL, it seems that they each
2416 wait for each other, leading to a delay before one of them times out.
2418 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2419 maintained 0.8.x compatibility because I don't think many are using it, and
2420 it is clearly obsolete.
2422 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2425 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2426 client certificate was expired. A simple patch fixes this, though I don't
2427 understand the full logic of why the verify callback is called multiple
2430 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2431 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2432 which causes problems with some clients (such as the Certicom SSL Plus
2433 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2434 disables the coutermeasure allowing Eudora to connect."
2436 38. Exim was not checking that a write() to a log file succeeded. This could
2437 lead to Bad Things if a log got too big, in particular if it hit a file
2438 size limit. Exim now panics and dies if it cannot write to a log file, just
2439 as it does if it cannot open a log file.
2441 39. Modified OS/Makefile-Linux so that it now contains
2443 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2445 The two -D definitions ensure that Exim is compiled with large file
2446 support, which makes it possible to handle log files that are bigger than
2449 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2450 instance) a domain was checked against a named list that involved a lookup,
2451 causing $domain_data to be set, then another domain was checked against the
2452 same list, then the first domain was re-checked, the value of $domain_data
2453 after the final check could be wrong. In particular, if the second check
2454 failed, it could be set empty. This bug probably also applied to
2457 41. The strip_trailing_dot option was not being applied to the address given
2458 with the -f command-line option.
2460 42. The code for reading a message's header from the spool was incrementing
2461 $received_count, but never initializing it. This meant that the value was
2462 incorrect (doubled) while delivering a message in the same process in which
2463 it was received. In the most common configuration of Exim, this never
2464 happens - a fresh exec is done - but it can happen when
2465 deliver_drop_privilege is set.
2467 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2468 it now includes up to 150 characters of the unexpected data in the log
2471 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2472 and building data strings. The size of both of these buffers was 10 000
2473 bytes - far larger than anybody would *ever* want, thought I. Needless to
2474 say, somebody hit the limit. I have increased the maximum line length to
2475 20 000 and the maximum data length of concatenated lines to 100 000. I have
2476 also fixed two bugs, because there was no checking on these buffers. Tsk,
2477 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2478 buffer is too small.
2480 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2481 lsearch lookups. Now it does.
2483 46. When parsing a route_list item in a manualroute router, a fixed-length
2484 buffer was used for the list of hosts. I made this 1024 bytes long,
2485 thinking that nobody would ever have a list of hosts that long. Wrong.
2486 Somebody had a whole pile of complicated expansion conditions, and the
2487 string was silently truncated, leading to an expansion error. It turns out
2488 that it is easier to change to an unlimited length (owing to other changes
2489 that have happened since this code was originally written) than to build
2490 structure for giving a limitation error. The length of the item that
2491 expands into the list of hosts is now unlimited.
2493 47. The lsearch lookup could not handle data where the length of text line was
2494 more than 4095 characters. Such lines were truncated, leading to shortened
2495 data being returned. It should now handle lines of any length.
2497 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2498 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2501 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2502 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2503 no longer happens in comment lines. A list of replacements is now placed
2504 at the head of all of the source files, except those whose only change is
2505 to replace PERL_COMMAND in the very first #! line.
2507 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2508 messages on the queue, with a bottom-up merge sort, using code contributed
2509 by Michael Haardt. This should make operations like -bp somewhat faster on
2510 large queues. It won't affect queue runners, except when queue_run_in_order
2513 51. Installed eximstats 1.31 in the distribution.
2515 52. Added support for SRV lookups to the dnslookup router.
2517 53. If an ACL referred to $message_body or $message_body_end, the value was not
2518 reset for any messages that followed in the same SMTP session.
2520 54. The store-handling optimization for building very long strings was not
2521 differentiating between the different store pools. I don't think this
2522 actually made any difference in practice, but I've tidied it.
2524 55. While running the routers to verify a sender address, $sender_address
2525 was still set to the sender address. This is wrong, because when routing to
2526 send a bounce to the sender, it would be empty. Therefore, I have changed
2527 it so that, while verifying a sender address, $sender_address is set to <>.
2528 (There is no change to what happens when verifying a recipient address.)
2530 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2531 target A or AAAA records (if not already returned) without resetting the
2532 qualify_single or search_parents options of the DNS resolver. These are
2533 inappropriate in this case because the targets of MX and SRV records must
2534 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2535 target that, when qualified, matched something in the local domain. These
2536 two options are now turned off when doing these lookups.
2538 57. It seems that at least some releases of Reiserfs (which does not have the
2539 concept of a fixed number of inodes) returns zero and not -1 for the
2540 number of available inodes. This interacted badly with check_spool_inodes,
2541 which assumed that -1 was the "no such thing" setting. What I have done is
2542 to check that the total number of inodes is greater than zero before doing
2543 the test of how many are available.
2545 58. When a "warn" ACL statement has a log_message modifier, the message is
2546 remembered, and not repeated. This is to avoid a lot of repetition when a
2547 message has many recipients that cause the same warning to be written.
2548 Howewer, Exim was preserving the list of already written lines for an
2549 entire SMTP session, which doesn't seem right. The memory is now reset if a
2550 new message is started.
2552 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2553 debugging output unless log_rewrite was also set.
2555 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2556 of (int)(handle) when we know that handle contains (void *)(-1).
2558 61. The Exim daemon panic-logs an error return when it closes the incoming
2559 connection. However "connection reset by peer" seems to be common, and
2560 isn't really an error worthy of noting specially, so that particular error
2563 62. When Exim is trying to find all the local interfaces, it used to panic and
2564 die if the ioctl to get the interface flags failed. However, it seems that
2565 on at least one OS (Solaris 9) it is possible to have an interface that is
2566 included in the list of interfaces, but for which you get a failure error
2567 for this call. This happens when the interface is not "plumbed" into a
2568 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2569 failure of the "get flags" call assumes that the interface is down.
2571 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2572 makes life easier for people who are doing arithmetic on fields extracted
2573 from dates, where you often get leading zeros that should not be
2574 interpreted as octal.
2576 64. Added qualify_domain to the redirect router, to override the global
2579 65. If a pathologically long header line contained very many addresses (the
2580 report of this problem mentioned 10 000) and each of them was rewritten,
2581 Exim could use up a very large amount of memory. (It kept on making new
2582 copies of the header line as it rewrote, and never released the old ones.)
2583 At the expense of a bit more processing, the header rewriting function has
2584 been changed so that it no longer eats memory in this way.
2586 66. The generation of the Received: header has been moved from the time that a
2587 message starts to be received, to the time that it finishes. The timestamp
2588 in the Received: header should now be very close to that of the <= log
2589 line. There are two side-effects of this change:
2591 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2592 logged header lines no longer include the local Received: line, because
2593 it has not yet been created. The same applies to a copy of the message
2594 that is returned to a non-SMTP sender when a message is rejected.
2596 (b) When a filter file is tested using -bf, no additional Received: header
2597 is added to the test message. After some thought, I decided that this
2600 This change does not affect the value of $received_for. It is still set
2601 after address rewriting, but before local_scan() is called.
2603 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2605 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2606 gave an unhelpful panic error message, and a defer error. I have managed to
2607 change this behaviour so that it now rejects any supplied certificate,
2608 which seems right, as the list of acceptable certificates is empty.
2610 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2611 gave an unhelpful defer error. I have not managed to make this reject any
2612 supplied certificates, but the error message it gives is "no certificate
2613 supplied", which is not helpful.
2615 70. exigrep's output now also includes lines that are not associated with any
2616 message, but which match the given pattern. Implemented by a patch from
2617 Martin Sluka, which also tidied up the Perl a bit.
2619 71. Recipient callout verification, like sender verification, was using <> in
2620 the MAIL FROM command. This isn't really the right thing, since the actual
2621 sender may affect whether the remote host accepts the recipient or not. I
2622 have changed it to use the actual sender in the callout; this means that
2623 the cache record is now keyed on a recipient/sender pair, not just the
2624 recipient address. There doesn't seem to be a real danger of callout loops,
2625 since a callout by the remote host to check the sender would use <>.
2626 [SEE ABOVE: changed after hitting problems.]
2628 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2629 temporary errors. However, in the case of such a code being given after
2630 the end of a data transmission (i.e. after ".") Exim was failing to write
2631 a retry record for the message. (Yes, there was some broken host that was
2632 actually sending 8xx at this point.)
2634 73. An unknown lookup type in a host list could cause Exim to panic-die when
2635 the list was checked. (An example that provoked this was putting <; in the
2636 middle of a list instead of at the start.) If this happened during a DATA
2637 ACL check, a -D file could be left lying around. This kind of configuration
2638 error no longer causes Exim to die; instead it causes a defer errror. The
2639 incident is still logged to the main and panic logs.
2641 74. Buglet left over from Exim 3 conversion. The message "too many messages
2642 in one connection" was written to the rejectlog but not the mainlog, except
2643 when address rewriting (yes!) was being logged.
2645 75. Added write_rejectlog option.
2647 76. When a system filter was run not as root (that is, when system_filter_user
2648 was set), the values of the $n variables were not being returned to the
2649 main process; thus, they were not subsequently available in the $sn
2652 77. Added +return_path_on_delivery log selector.
2654 78. A connection timeout was being treated differently from recipients deferred
2655 when testing hosts_max_try with a message that was older than the host's
2656 retry timeout. (The host should not be counted, thus allowing all hosts to
2657 be tried at least once before bouncing.) This may have been the cause of an
2658 occasionally reported bug whereby a message would remain on the queue
2659 longer than the retry timeout, but would be bounced if a delivery was
2660 forced. I say "may" because I never totally pinned down the problem;
2661 setting up timeout/retry tests is difficult. See also the next item.
2663 79. The ultimate address timeout was not being applied to errors that involved
2664 a combination of host plus message (for example, a timeout on a MAIL
2665 command). When an address resolved to a number of possible hosts, and they
2666 were not all tried for each delivery (e.g. because of hosts_max_try), a
2667 message could remain on the queue longer than the retry timeout.
2669 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2672 81. Fixed an obscure SMTP outgoing bug which required at least the following
2673 conditions: (a) there was another message waiting for the same server;
2674 (b) the server returned 5xx to all RCPT commands in the first message so
2675 that the message was not completed; (c) the server dropped the connection
2676 or gave a negative response to the RSET that Exim sends to abort the
2677 transaction. The observed case was a dropped connection after DATA that had
2678 been sent in pipelining mode. That is, the server had advertised PIPELINING
2679 but was not implementing it correctly. The effect of the bug was incorrect
2680 behaviour, such as trying another host, and this could lead to a crash.
2686 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2687 and daemon.c were passed as pointers to ints; they should have been
2688 pointers to socklen_t variables (which are typically unsigned ints).
2690 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2693 3. Fixed a really odd bug that affected only the testing scheme; patching a
2694 certain fixed string in the binary changed the value of another string that
2695 happened to be identical to the end of the original first string.
2697 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2698 name", it returns that address as the IP address. On some operating
2699 systems (e.g. Solaris), it also passes back the IP address string as the
2700 "host name". However, on others (e.g. Linux), it passes back an empty
2701 string. Exim wasn't checking for this, and was changing the host name to an
2702 empty string, assuming it had been canonicized.
2704 5. Although rare, it is permitted to have more than one PTR record for a given
2705 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2706 all the names associated with an address, because they do in Solaris.
2707 However, it seems that they do not in Linux for data that comes from the
2708 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2709 I found this out when I moved to a new Linux workstation and tried to run
2710 the Exim test suite.
2712 To get round this problem I have changed the code so that it now does its
2713 own call to the DNS to look up PTR records when searching for a host name.
2714 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2715 addresses that are only in /etc/hosts are still found.
2717 This behaviour is, however, controlled by an option called host_lookup_
2718 order, which defaults to "bydns:byaddr". If people want to use the other
2719 order, or indeed, just use one or the other means of lookup, they can
2720 specify it in this variable.
2722 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2723 some operating systems, this comes back from gethostbyaddr() as an empty
2724 string, and this is what Exim used to test for. However, it seems that in
2725 other systems, "." is yielded. Exim now tests for this case too.
2727 7. The values of check_spool_space and check_log_space are now held internally
2728 as a number of kilobytes instead of an absolute number of bytes. If a
2729 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2730 kilobyte. This means that much larger values can be stored.
2732 8. Exim monitor: an attempt to get the action menu when not actually pointing
2733 at a message produces an empty menu entitled "No message selected". This
2734 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2735 no entries in it ("Shell widget menu has zero width and/or height"). So I
2736 have added a single, blank menu entry in this case.
2738 9. Added ${quote_local_part.
2740 10. MIME decoding is now applied to the contents of Subject: header lines when
2743 11. Now that a reference to $sender_host_address automatically causes a reverse
2744 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2745 host lookup before query-style lookups in lists that might use this
2746 variable. This has therefore been abolished, and the "net-" prefix is no
2747 longer necessary for query-style lookups.
2749 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2750 have been a typo for LFLAGS, so it has been changed.
2752 13. The install script calls Exim with "-C /dev/null" in order to find the
2753 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2754 to be output. Howeve, since Exim outputs its version number before the
2755 error, it didn't break the script. It just looked ugly. I fixed this by
2756 always allowing "-C /dev/null" if the caller is root.
2758 14. Ignore overlarge ACL variable number when reading spool file - insurance
2759 against a later release with more variables having written the file.
2761 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2762 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2763 this, as well as the form without the "IPv6" on the front.
2765 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2766 OS/Makefile-Darwin file.
2768 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2769 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2771 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2772 variable. However, some OS use ints for inodes. I've added cast to long int
2773 to get rid of the compiler warning.
2775 19. I had forgotten to lock out "/../" in configuration file names when
2776 ALT_CONFIG_PREFIX was set.
2778 20. Routers used for verification do not need to specify transports. However,
2779 if such a router generated a host list, and callout was configured, Exim
2780 crashed, because it could not find a port number from the (non-existent)
2781 transport. It now assumes port 25 in this circumstance.
2783 21. Added the -t option to exigrep.
2785 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2786 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2787 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2788 them is set, LOOKUP_LSEARCH is forced.
2790 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2792 24. Added sender and host information to the "rejected by local_scan()" log
2793 line; previously there was no indication of these.
2795 25. Added .include_if_exists.
2797 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2798 Linux. It turns out that not all file systems support this. Apparently some
2799 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2800 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2801 not supported on the file descriptor, is now ignored when Exim is trying to
2802 sync a directory. This applies only to Linux.
2804 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2806 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2807 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2810 29. Added support for maildirsize files from supplied patch (modified a bit).
2812 30. The use of :fail: followed by an empty string could lead Exim to respond to
2813 sender verification failures with (e.g.):
2815 550 Verification failed for <xxx>
2816 550 Sender verify failed
2818 where the first response line was missing the '-' that indicates it is not
2819 the final line of the response.
2821 31. The loop for finding the name of the user that called Exim had a hardwired
2822 limit of 10; it now uses the value of finduser_retries, which is used for
2823 all other user lookups.
2825 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2828 33. Exim was neglecting to zero errno before one call of strtol() when
2829 expanding a string and expecting an integer value. On some systems this
2830 resulted in spurious "integer overflow" errors. Also, it was casting the
2831 result into an int without checking.
2833 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2834 did not work. The code looks as if it has *never* worked, though it appears
2835 to have been documented since at least releast 1.62. I have made it work.
2837 35. The "timeout_DNS" error in retry rules, also documented since at least
2838 1.62, also never worked. As it isn't clear exactly what this means, and
2839 clearly it isn't a major issue, I have abolished the feature by treating it
2840 as "timeout", and writing a warning to the main and panic logs.
2842 36. The display of retry rules for -brt wasn't always showing the error code
2845 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2846 timeout_connect_A, timeout_connect_MX.
2848 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2849 to the empty sender.
2851 39. The daemon was not analysing the content of -oX till after it had closed
2852 stderr and disconnected from the controlling terminal. This meant that any
2853 syntax errors were only noted on the panic log, and the return code from
2854 the command was 0. By re-arranging the code a little, I've made the
2855 decoding happen first, so such errors now appear on stderr, and the return
2856 code is 1. However, the actual setting up of the sockets still happens in
2857 the disconnected process, so errors there are still only recorded on the
2860 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2861 connections (as happens on some IP stacks) was logged at start up time as
2862 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2863 it from "IPv6 and IPv4", which means that two separate sockets are being
2866 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2867 says whether AF_INET or AF_INET6 was passed as an argument.
2869 42. Exiwhat output was messed up when time zones were included in log
2872 43. Exiwhat now gives more information about the daemon's listening ports,
2873 and whether -tls-on-connect was used.
2875 44. The "port" option of the smtp transport is now expanded.
2877 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2878 silently ignored. Now an error message is written to the main and panic
2881 46. There's a new ACL modifier called "logwrite" which writes to a log file
2882 as soon as it is encountered.
2884 47. Added $local_user_uid and $local_user_gid at routing time.
2886 48. Exim crashed when trying to verify a sender address that was being
2889 49. Exim was recognizing only a space character after ".include". It now also
2890 recognizes a tab character.
2892 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2893 extracting the relevant information from the specification. The man page no
2894 longer contains scrambled data for the -d option, and I've added a section
2895 at the front about calling Exim under different names.
2897 51. Added "extra_headers" argument to the "mail" command in filter files.
2899 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2902 53. Installed eximstats 1.29.
2904 54. Added transport_filter_timeout as a generic transport option.
2906 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2907 Cc: header lines. This was required by RFC 822, but it not required by RFC
2910 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2911 incoming messages that did not have them. Now it does so only if the
2912 message originates locally, that is, if there is no associated remote host
2913 address. When Resent- header lines are present, this applies to the Resent-
2914 lines rather than the non-Resent- lines.
2916 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2917 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2919 58. Messages for configuration errors now include the name of the main
2920 configuration file - useful now that there may be more than one file in a
2921 list (.included file names were always shown).
2923 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2924 for those rare installations that do not start the daemon as root or run it
2925 setuid root. I've cut out the call to initgroups() if the daemon is not
2928 60. The Exim user and group can now be bound into the binary as text strings
2929 that are looked up at the start of Exim's processing.
2931 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2933 62. Added $mailstore_basename variable.
2935 63. Installed patch to sieve.c from Michael Haardt.
2937 64. When Exim failed to open the panic log after failing to open the main log,
2938 the original message it was trying to log was written to stderr and debug
2939 output, but if they were not available (the usual case in production), it
2940 was lost. Now it is written to syslog before the two lines that record the
2941 failures to open the logs.
2943 65. Users' Exim filters run in subprocesses under the user's uid. It is
2944 possible for a "deliver" command or an alias in a "personal" command to
2945 provoke an address rewrite. If logging of address rewriting is configured,
2946 this fails because the process is not running as root or exim. There may be
2947 a better way of dealing with this, but for the moment (because 4.30 needs
2948 to be released), I have disabled address rewrite logging when running a
2949 filter in a non-root, non-exim process.
2955 1. The buildconfig auxiliary program wasn't quoting the value set for
2956 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2957 not defined. This bug was masked in 4.22 by the effect that was fixed in
2960 2. Some messages that were rejected after a message id was allocated were
2961 shown as "incomplete" by exigrep. It no longer does this for messages that
2962 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2964 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2965 have allow_domain_literals set, the ID did not get logged in the <= line.
2966 Domain literals are now always recognized in Message-ID: header lines.
2968 4. The first argument for a ${extract expansion item is the key name or field
2969 number. Leading and trailing spaces in this item were not being ignored,
2970 causing some misleading effects.
2972 5. When deliver_drop_privilege was set, single queue runner processes started
2973 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2974 same command in the process it spins off) were not dropping privilege.
2976 6. When the daemon running as "exim" started a queue runner, it always
2977 re-executed Exim in the spun-off process. This is a waste of effort when
2978 deliver_drop_privilege is set. The new process now just calls the
2979 queue-runner function directly.
2985 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2988 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2989 ignored. Though the use of -oP was forcing the writing of a pid file, it
2990 was always written to the default place.
2992 3. If the message "no IP address found for host xxxx" is generated during
2993 incoming verification, it is now followed by identification of the incoming
2994 connection (so you can more easily find what provoked it).
2996 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2998 5. Added some features to "harden" Exim a bit more against certain attacks:
3000 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3001 be put in Local/Makefile. This is like the never_users runtime option,
3002 but it cannot be overridden. The default setting is "root".
3004 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3005 prefix string with which any file named in a -C command line option
3008 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3009 is retained for -C and -D only if the caller of Exim is root. Without
3010 it, the exim user may also use -C and -D and retain privilege.
3012 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3013 command line option is disabled.
3015 6. Macro names set by the -D option must start with an upper case letter, just
3016 like macro names defined in the configuration file.
3018 7. Added "dereference=" facility to LDAP.
3020 8. Two instances of the typo "uknown" in the source files are fixed.
3022 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3023 the Configure-Makefile script screwed up while processing it.
3025 10. Incorporated PCRE 4.4.
3027 11. The SMTP synchronization check was not operating right at the start of an
3028 SMTP session. For example, it could not catch a HELO sent before the client
3029 waited for the greeting. There is now a check for outstanding input at the
3030 point when the greeting is written. Because of the duplex, asynchronous
3031 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3032 way, but not yet received, when the check is performed.
3034 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3035 on TCP/IP sockets, because this apparently causes some broken clients to
3038 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3039 unchanged) from the Cygwin maintainer.
3041 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3042 was supported instead of testing for mbx. Effectively a typo.
3044 15. The spa authenticator server code was not checking that the input it
3045 received was valid base64.
3047 16. The debug output line for the "set" modifier in ACLs was not showing the
3048 name of the variable that was being set.
3050 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3052 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3053 lookup on its own. Something else was needed to trigger the lookup. For
3054 example, a match in host_lookup or the need for a host name in a host list.
3055 Now, if $sender_host_name is referenced and the host name has not yet been
3056 looked up, a lookup is performed. If the lookup fails, the variable remains
3057 empty, and $host_lookup_failed is set to "1".
3059 19. Added "eqi" as a case-independent comparison operator.
3061 20. The saslauthd authentication condition could segfault if neither service
3062 nor realm was specified.
3064 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3065 error message that was logged was misleading, and incoming SMTP
3066 connections were dropped. The message is now more accurate, and temporary
3067 errors are given to SMTP connections.
3069 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3070 (except RSET) with a 421 error, until QUIT is received. However, it was
3071 failing to send a response to QUIT.
3073 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3074 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3075 ACL is now run after the helo_try_verify_hosts code.
3077 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3078 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3079 case-independent, so other case variants are also recognized). Apparently
3080 some systems use these upper case variants.
3082 25. If more than two messages were waiting for the same host, and a transport
3083 filter was specified for the transport, Exim sent two messages over the
3084 same TCP/IP connection, and then failed with "socket operation on non-
3085 socket" when it tried to send the third.
3087 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3089 27. The extern definition of crypt16() in expand.c was not being excluded when
3090 the OS had its own crypt16() function.
3092 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3093 as a preferred synonym for return_size_limit, both as an option and as an
3096 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3098 30. Changed the default configuration ACL to relax the local part checking rule
3099 for addresses that are not in any local domains. For these addresses,
3100 slashes and pipe symbols are allowed within local parts, but the sequence
3101 /../ is explicitly forbidden.
3103 31. SPA server authentication was not clearing the challenge buffer before
3106 32. log_message in a "warn" ACL statement was writing to the reject log as
3107 well as to the main log, which contradicts the documentation and doesn't
3108 seem right (because no rejection is happening). So I have stopped it.
3110 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3111 However, I am unable to do any testing of this.
3113 34. Fixed an infelicity in the appendfile transport. When checking directories
3114 for a mailbox, to see if any needed to be created, it was accidentally
3115 using path names with one or more superfluous leading slashes; tracing
3116 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3118 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3119 contents are added to the log line that is written for every discarded
3120 recipient. (Previously a log_message setting was ignored.)
3122 36. The ${quote: operator now quotes the string if it is empty.
3124 37. The install script runs exim in order to find its version number. If for
3125 some reason other than non-existence or emptiness, which it checks, it
3126 could not run './exim', it was installing it with an empty version number,
3127 i.e. as "exim-". This error state is now caught, and the installation is
3130 38. An argument was missing from the function that creates an error message
3131 when Exim fails to connect to the socket for saslauthd authentication.
3132 This could cause Exim to crash, or give a corrupted message.
3134 39. Added isip, isip4, and isip6 to ${if conditions.
3136 40. The ACL variables $acl_xx are now saved with the message, and can be
3137 accessed later in routers, transports, and filters.
3139 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3140 strings in the file are not string-expanded.
3142 42. If a MAIL command specified a SIZE value that was too large to fit into an
3143 int variable, the check against message_size_limit failed. Such values are
3144 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3145 day this will have to be increased, but I don't think I want to be around
3146 when emails are that large.
3153 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3154 iconv() is not standard in FreeBSD.
3156 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3157 IPv6 enabled. The observed symptom was a segmentation fault on return from
3158 the function os_common_find_running_interfaces() in src/os.c.
3160 3. In the check_special_case() function in daemon.c I had used "errno" as an
3161 argument name, which causes warnings on some systems. This was basically a
3162 typo, since it was named "eno" in the comments!
3164 4. The code that waits for the clock to tick (at a resolution of some fraction
3165 of a second) so as to ensure message-id uniqueness was always waiting for
3166 at least one whole tick, when it could have waited for less. [This is
3167 almost certainly not relevant at current processor speeds, where it is
3168 unlikely to ever wait at all. But we try to future-proof.]
3170 5. The function that sleeps for a time interval that includes fractions of a
3171 second contained a race. It did not block SIGALRM between setting the
3172 timer, and suspending (a couple of lines later). If the interval was short
3173 and the sigsuspend() was delayed until after it had expired, the suspension
3174 never ended. On busy systems this could lead to processes getting stuck for
3177 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3178 process, before it forks any delivery processes. The open lookup caching
3179 mechanism meant that the open file or database connection was passed into
3180 the delivery process. The problem was that delivery processes always tidy
3181 up cached lookup data. This could cause a problem for the next delivery
3182 process started by the queue runner, because the external queue runner
3183 process does not know about the closure. So the next delivery process
3184 still has data in the lookup cache. In the case of a file lookup, there was
3185 no problem because closing a file descriptor in a subprocess doesn't affect
3186 the parent. However, if the lookup was caching a connection to a database,
3187 the connection was closed, and the second delivery process was likely to
3188 see errors such as "PGSQL: query failed: server closed the connection
3189 unexpectedly". The problem has been fixed by closing all cached lookups
3190 in a queue runner before running a delivery process.
3192 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3193 seem to have the "const" qualifier which it has on other OS. I've
3196 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3197 *of the same type* (client or server) with the same public name that an
3198 error should be diagnosed.
3200 9. When Exim looked up a host name for an IP address, but failed to find the
3201 original IP address when looking up the host name (a safety check), it
3202 output the message "<ip address> does not match any IP for NULL", which was
3203 confusing, to say the least. The bug was that the host name should have
3204 appeared instead of "NULL".
3206 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3207 user that is built into the binary, and the -C or -D options is used, root
3208 privilege is dropped before the configuration file is read. In addition,
3209 logging is switched to stderr instead of the normal log files. If the
3210 configuration then re-defines the Exim user, the unprivileged environment
3211 is probably not what is expected, so Exim logs a panic warning message (but
3214 However, if deliver_drop_privilege is set, the unprivileged state may well
3215 be exactly what is intended, so the warning has been cut out in that case,
3216 and Exim is allowed to try to write to its normal log files.
3222 1. smtp_return_error_details was not giving details for temporary sender
3223 or receiver verification errors.
3225 2. Diagnose a configuration error if two authenticators have the same public
3228 3. Exim used not to create the message log file for a message until the first
3229 delivery attempt. This could be confusing when incoming messages were held
3230 for policy or load reasons. The message log file is now created at the time
3231 the message is received, and an initial "Received" line is written to it.
3233 4. The automatically generated man page for command line options had a minor
3234 bug that caused no ill effects; however, a more serious problem was that
3235 the procedure for building the man page automatically didn't always
3236 operate. Consequently, release 4.20 contains an out-of-date version. This
3237 shouldn't happen again.
3239 5. When building Exim with embedded Perl support, the script that builds the
3240 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3241 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3243 6. The freeze_tell option was not being used for messages that were frozen on
3244 arrival, either by an ACL or by local_scan().
3246 7. Added the smtp_incomplete_transaction log selector.
3248 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3249 was accepting AUTH without a new EHLO.
3251 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3252 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3255 10. Logging of TCP/IP connections (when configured) now happens in the main
3256 daemon process instead of the child process, so that the TCP/IP connection
3257 count is more accurate (but it can never be perfect).
3259 11. The use of "drop" in a nested ACL was not being handled correctly in the
3260 outer ACL. Now, if condition failure induced by the nested "drop" causes
3261 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3262 or "require"), the connection is dropped.
3264 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3265 that yield "discard" can now be used with an "accept" or a "discard" verb,
3266 but an error is generated for any others (because I can't see a useful way
3267 to define what should happen).
3269 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3270 are now processed in the same way as lines in the Exim configuration file.
3271 In particular, continuation lines are supported.
3273 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3275 15. Added -ti meaning -t -i.
3277 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3278 domains, and warn by logging if others are found.
3280 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3281 returned from GIFCONF when Exim is trying to find the list of interfaces on
3282 a host. The code in os.c has been modified to copy each ifreq to an aligned
3283 structure in all cases.
3285 Also, in some cases, the returned ifreq's were being copied to a 'struct
3286 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3287 means the last couple of bytes of an IPv6 address could be chopped if the
3288 ifreq contained only a normal sockaddr (14 bytes storage).
3290 18. Named domain lists were not supported in the hosts_treat_as_local option.
3291 An entry such as +xxxx was not recognized, and was treated as a literal
3294 19. Ensure that header lines added by a DATA ACL are included in the reject log
3295 if the ACL subsequently rejects the message.
3297 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3298 MD5 (which is deprecated).
3300 21. When testing a filter file using -bf, Exim was writing a message when it
3301 took the sender from a "From " line in the message, but it was not doing so
3302 when it took $return_path from a Return-Path: header line. It now does.
3304 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3305 with a valid header line field name (a series of printing characters
3306 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3308 23. Changed "disc" in the source to "disk" to conform to the documentation and
3309 the book and for uniformity.
3311 24. Ignore Sendmail's -Ooption=value command line item.
3313 25. When execve() failed while trying to run a command in a pipe transport,
3314 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3315 could be confused with a return value of 69 from the command itself. This
3316 has been changed to 127, the value the shell returns if it is asked to run
3317 a non-existent command. The wording for the related log line suggests a
3318 non-existent command as the problem.
3320 26. If received_header_text expands to an empty string, do not add a Received:
3321 header line to the message. (Well, it adds a token one on the spool, but
3322 marks it "old" so that it doesn't get used or transmitted.)
3324 27. Installed eximstats 1.28 (addition of -nt option).
3326 28. There was no check for failure on the call to getsockname() in the daemon
3327 code. This can fail if there is a shortage of resources on the system, with
3328 ENOMEM, for example. A temporary error is now given on failure.
3330 29. Contrary to the C standard, it seems that in some environments, the
3331 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3332 program. Exim now does this explicitly; it affects the formatting of
3333 timestamps using strftime().
3335 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3336 complaints. I've now initialized all the variables, to avoid this.
3338 32. Header lines added by a system filter were not being "seen" during
3339 transport-time rewrites.
3341 33. The info_callback() function passed to OpenSSL is set up with type void
3342 (*)(SSL *, int, int), as described somewhere. However, when calling the
3343 function (actually a macro) that sets it up, the type void(*)() is
3344 expected. I've put in a cast to prevent warnings from picky compilers.
3346 34. If a DNS black list lookup found a CNAME record, but there were no A
3347 records associated with the domain it pointed at, Exim crashed.
3349 35. If a DNS black list lookup returned more than one A record, Exim ignored
3350 all but the first. It now scans all returned addresses if a particular IP
3351 value is being sought. In this situation, the contents of the
3352 $dnslist_value variable are a list of all the addresses, separated by a
3355 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3356 to accept a host name and all its aliases if the forward lookup for any of
3357 them yielded the IP address of the incoming connection. Now it accepts only
3358 those names whose forward lookup yields the correct IP address. Any other
3359 names are discarded. This closes a loophole whereby a rogue DNS
3360 administrator could create reverse DNS records to break through a
3361 wildcarded host restriction in an ACL.
3363 37. If a user filter or a system filter that ran in a subprocess used any of
3364 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3365 the wrong values were passed to the pipe command ($thisaddress had the
3366 value of $0, $0 had the value of $1, etc). This bug was introduced by
3367 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3369 38. Improved the line breaking for long SMTP error messages from ACLs.
3370 Previously, if there was no break point between 40 and 75 characters, Exim
3371 left the rest of the message alone. Two changes have been made: (a) I've
3372 reduced the minimum length to 35 characters; (b) if it can't find a break
3373 point between 35 and 75 characters, it looks ahead and uses the first one
3374 that it finds. This may give the occasional overlong line, but at least the
3375 remaining text gets split now.
3377 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3378 file descriptors might be low, and that setting 1000 would always raise it.
3379 It turns out that in some environments, the limit is already over 1000 and
3380 that lowering it causes trouble. So now Exim takes care not to decrease it.
3382 40. When delivering a message, the value of $return_path is set to $sender_
3383 address at the start of routing (routers may change the value). By an
3384 oversight, this default was not being set up when an address was tested by
3385 -bt or -bv, which affected the outcome if any router or filter referred to
3388 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3389 log only when "message" or "log_message" are set. However, if one of the
3390 conditions was an address verification, or a call to a nested ACL, the
3391 messages generated by the underlying test were being passed through. This
3392 no longer happens. The underlying message is available in $acl_verify_
3393 message for both "message" and "log_message" expansions, so it can be
3394 passed through if needed.
3396 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3397 new expansion $bh_ to give the encoded byte string without charset
3398 translation. Translation happens only if iconv() is available; HAVE_ICONV
3399 indicates this at build time. HEADERS_CHARSET gives the charset to
3400 translate to; headers_charset can change it in the configuration, and
3401 "headers charset" can change it in an individual filter file.
3403 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3404 that creates RFC 2047 encoded "words" labels them as that charset instead
3405 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3406 expansion operator; (ii) when Exim creates a From: line for a local
3407 message; (iii) when a header line is rewritten to include a "phrase" part.
3409 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3410 buggy, causing it to skip the first lines of messages whose message ID
3411 ended in 'D'. This would not have bitten before Exim release 4.14, because
3412 message IDs were unlikely to end in 'D' before then. The effect was to have
3413 incorrect size information for certain domains.
3415 45. #include "config.h" was missing at the start of the crypt16.c module. This
3416 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3419 46. If there was a timeout during a "random" callout check, Exim treated it as
3420 a failure of the random address, and carried on sending RSET and the real
3421 address. If the delay was just some slowness somewhere, the response to the
3422 original RCPT would be taken as a response to RSET and so on, causing
3423 mayhem of various kinds.
3425 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3426 when I implemented it. It didn't allow for the fact that some option values
3427 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3428 the right test for positive values.
3430 48. Domain names in DNS records are case-independent. Exim always looks them up
3431 in lower case. Some resolvers return domain names in exactly the case they
3432 appear in the zone file, that is, they may contain uppercase letters. Not
3433 all resolvers do this - some return always lower case. Exim was treating a
3434 change of case by a resolver as a change of domain, similar to a widening
3435 of a domain abbreviation. This triggered its re-routing code and so it was
3436 trying to route what was effectively the same domain again. This normally
3437 caused routing to fail (because the router wouldn't handle the domain
3438 twice). Now Exim checks for this case specially, and just changes the
3439 casing of the domain that it ultimately uses when it transmits the message
3442 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3445 50. If a filter generated a file delivery with a non-absolute name (possible if
3446 no home directory exists for the router), the forbid_file option was not
3449 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3450 the existing equality matching.
3452 52. Exim was using ints instead of ino_t variables in some places where it was
3453 dealing with inode numbers.
3455 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3456 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3457 called TMPDIR, and if it finds it is different, it changes its value.
3459 54. The smtp_printf() function is now made available to local_scan() so
3460 additional output lines can be written before returning. There is also an
3461 smtp_fflush() function to enable the detection of a dropped connection.
3462 The variables smtp_input and smtp_batched_input are exported to
3465 55. Changed the default runtime configuration: the message "Unknown user"
3466 has been removed from the ACL, and instead placed on the localuser router,
3467 using the cannot_route_message feature. This means that any verification
3468 failures that generate their own messages won't get overridden. Similarly,
3469 the "Unrouteable address" message that was in the ACL for unverifiable
3470 relay addresses has also been removed.
3472 56. Added hosts_avoid_esmtp to the smtp transport.
3474 57. The exicyclog script was not checking for the esoteric option
3475 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3476 will work only if exicyclog is run under the appropriate euid.
3478 58. Following a discussion on the list, the rules by which Exim recognises line
3479 endings on incoming messages have been changed. The -dropcr and drop_cr
3480 options are now no-ops, retained only for backwards compatibility. The
3481 following line terminators are recognized: LF CRLF CR. However, special
3482 processing applies to CR:
3484 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3485 nor a local message in the state where . is a terminator.
3487 (ii) If a bare CR is encountered in a header line, an extra space is added
3488 after the line terminator so as not to end the header. The reasoning
3489 behind this is that bare CRs in header lines are most likely either
3490 to be mistakes, or people trying to play silly games.
3492 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3493 was being incorrectly given as 18 bytes larger than it should have been.
3494 This is a VOB (very old bug).
3496 60. This may never have affected anything current, but just in case it has:
3497 When the local host is found other than at the start of a list of hosts,
3498 the local host, those with the same MX, and any that follow, are discarded.
3499 When the list in question was part of a longer list of hosts, the following
3500 hosts (not currently being processed) were also being discarded. This no
3501 longer happens. I'm not sure if this situation could ever has previously
3504 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3507 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3508 header. This is something that is recommended in a new Internet Draft, and
3509 is something that is documented as being done by Sendmail. There are two
3510 possible values. For messages generated by the autoreply transport, Exim
3513 Auto-Submitted: auto-replied
3515 whereas for all other generated messages (e.g. bounces) it adds
3517 Auto-Submitted: auto-generated
3519 63. The "personal" condition in filters now includes a test for the
3520 Auto-Submitted: header. If it contains the string "auto-" the message it
3521 not considered personal.
3523 64. Added rcpt_include_affixes as a generic transport option.
3525 65. Added queue_only_override (default true).
3527 66. Added the syslog_duplication option.
3529 67. If what should have been the first header line of a message consisted of
3530 a space followed by a colon, Exim was mis-interpreting it as a header line.
3531 It isn't of course - it is syntactically invalid and should therefore be
3532 treated as the start of the message body. The misbehaviour could have
3533 caused a number of strange effects, including loss of data in subsequent
3534 header lines, and spool format errors.
3536 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3537 client host had authenticated. This control can now be exercised by an ACL
3538 for more flexibility.
3540 69. By default, callouts do not happen when testing with -bh. There is now a
3541 variant, -bhc, which does actually run the callout code, including
3542 consulting and updating the callout cache.
3544 70. Added support for saslauthd authentication, courtesy of Alexander
3547 71. If statvfs() failed on the spool or log directories while checking their
3548 size for availability, Exim confusingly gave the error "space shortage".
3549 Furthermore, in debugging mode it crashed with a floating point exception.
3550 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3551 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3552 serious problem, Exim now writes to the main and panic logs when this
3553 happens, with details of the failure. It then refuses to accept the
3554 incoming message, giving the message "spool directory problem" or "log
3555 directory problem" with a 421 code for SMTP messages.
3557 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3558 0, 1, and 2 exist, because some OS complain for execs without them (see
3559 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3560 descriptors. However, the code omitted to check that the open succeeded,
3561 causing mysterious errors if for some reason the permissions on /dev/null
3562 got screwed. Now Exim writes a message to the main and panic logs, and
3563 bombs out if it can't open /dev/null.
3565 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3566 interact so that it is all more flexible. It is supposed to remain
3567 backwards compatible. Also added extra_local_interfaces.
3569 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3570 to bomb out with an assertion failure - to the client this appears as a
3571 connection drop. This problem occurs in the part of the code that was taken
3572 from the Samba project. Fortunately, the assertion is in a very simple
3573 function, so I have fixed this by reproducing the function inline in the
3574 one place where it is called, and arranging for authentication to fail
3575 instead of killing the process with assert().
3577 75. The SPA client code was not working when the server requested OEM rather
3578 than Unicode encoding.
3580 76. Added code to make require_files with a specific uid setting more usable in
3581 the case where statting the file as root fails - usually a non-root-mounted
3582 NFS file system. When this happens and the failure is EACCES, Exim now
3583 forks a subprocess and does the per-uid checking as the relevant uid.
3585 77. Added process_log_path.
3587 78. If log_file_path was not explicitly set, a setting of check_log_space or
3588 check_log_inodes was ignored.
3590 79. If a space check for the spool or log partitions fails, the incident is now
3591 logged. Of course, in the latter case the data may get lost...
3593 80. Added the %p formatting code to string_format() so that it can be used to
3594 print addresses in debug_print(). Adjusted all the address printing in the
3595 debugging in store.c to use %p rather than %d.
3597 81. There was a concern that a line of code in smtp_in.c could overflow a
3598 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3599 initially expressed, the concern was not well-founded, because trailing
3600 spaces are removed early. However, if the trailing spaces were followed by
3601 a NULL, they did not get removed, so the overflow was possible. Two fixes
3604 (a) I re-wrote the offending code in a cleaner fashion.
3605 (b) If an incoming SMTP command contains a NULL character, it is rejected
3608 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3609 runs initgroups(), so that if the Exim user is in any additional groups,
3610 they will be used during message reception.
3616 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob