1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.267 2005/11/21 12:04:23 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
18 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
37 PH/03 For some reason, the default setting of PATH when running a command from
38 a pipe transport was just "/usr/bin". I have changed it to
41 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
42 anything to be listed in the output from -bV.
44 PH/05 When a filter generated an autoreply, the entire To: header line was
45 quoted in the delivery log line, like this:
47 => >A.N.Other <ano@some.domain> <original@ddress> ...
49 This has been changed so that it extracts the operative address. There
50 may be more than one such address. If so, they are comma-separated, like
53 => >ano@some.domain,ona@other.domain <original@ddress> ...
55 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
56 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
57 not being looked up in the rDNS to get a host name, Exim was showing the
58 IP address twice in Received: lines, even though the IP addresses were
59 identical. For example:
61 Received: from [1.2.3.4] (helo=[1.2.3.4])
63 However, if the real host name was known, it was omitting the HELO data
64 if it matched the actual IP address. This has been tidied up so that it
65 doesn't show the same IP address twice.
67 PH/07 When both +timestamp and +memory debugging was on, the value given by
68 $tod_xxx expansions could be wrong, because the tod_stamp() function was
69 called by the debug printing, thereby overwriting the timestamp buffer.
70 Debugging no longer uses the tod_stamp() function when +timestamp is set.
72 PH/08 When the original message was included in an autoreply transport, it
73 always said "this is a copy of the message, including all the headers",
74 even if body_only or headers_only was set. It now gives an appropriate
77 PH/09 Applied a patch from the Sieve maintainer which:
80 o adds the (disabled) notify extension core
81 o adds some debug output for the result of if/elsif tests
82 o points to the current vacation draft in the documentation
83 and documents the missing references header update
87 o fixes a bug in processing the envelope test (when testing
88 multiple envelope elements, the last element determinted the
91 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
92 Electronic Mail") by including:
94 Auto-submitted: auto-generated
96 in the messages that it generates (bounce messages and others, such as
97 warnings). In the case of bounce messages for non-SMTP mesages, there was
98 also a typo: it was using "Auto_submitted" (underscore instead of
99 hyphen). Since every message generated by Exim is necessarily in response
100 to another message, thes have all been changed to:
102 Auto-Submitted: auto-replied
104 in accordance with these statements in the RFC:
106 The auto-replied keyword:
108 - SHOULD be used on messages sent in direct response to another
109 message by an automatic process,
111 - MUST NOT be used on manually-generated messages,
113 - MAY be used on Delivery Status Notifications (DSNs) and Message
114 Disposition Notifications (MDNs),
116 - MUST NOT be used on messages generated by automatic or periodic
117 processes, except for messages which are automatic responses to
120 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
121 to the default Received: header definition.
123 PH/12 Added log selector acl_warn_skipped (default on).
125 PH/13 After a successful wildlsearch lookup, discard the values of numeric
126 variables because (a) they are in the wrong storage pool and (b) even if
127 they were copied, it wouldn't work properly because of the caching.
129 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
130 checking when decoding. Apparently there are clients that generate
131 overlong encoded strings. Why am I not surprised?
133 PH/15 If the first argument of "${if match_address" was not empty, but did not
134 contain an "@" character, Exim crashed. Now it writes a panic log message
135 and treats the condition as false.
137 PH/16 In autoreply, treat an empty string for "once" the same as unset.
139 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
140 extension "envelope-auth". The code is finished and in agreement with
141 other implementations, but there is no documentation so far and in fact,
142 nobody wrote the draft yet. This extension is currently #undef'ed, thus
143 not changing the active code.
145 Print executed "if" and "elsif" statements when debugging is used. This
146 helps a great deal to understand what a filter does.
148 Document more things not specified clearly in RFC3028. I had all this
149 sorted out, when out of a sudden new issues came to my mind. Oops."
151 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
154 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
155 canonical form (as documented). However, after a host name lookup from
156 the IP address, check_host() was doing a simple string comparison with
157 addresses acquired from the DNS when checking that the found name did
158 have the original IP as one of its addresses. Since any found IPv6
159 addresses are likely to be in abbreviated form, the comparison could
160 fail. Luckily, there already exists a function for doing the comparison
161 by converting both addresses to binary, so now that is used instead of
168 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
169 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
172 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
173 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
175 PH/03 Typo: missing ".o" in src/pcre/Makefile.
177 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
178 header line, restrict the check to what is listed in RFCs 2369 and 2929.
179 Also, for "Auto-Submitted", treat anything other than "no" as
180 non-personal, in accordance with RFC 3834. (Previously it treated
181 anything starting "auto-" as non-personal.)
183 TF/01 The control=submission/name=... option had a problem with syntax
184 errors if the name included a slash character. The /name= option
185 now slurps the rest of the string, so it can include any characters
186 but it must come last in the list of options (after /sender_retain
189 PH/05 Some modifications to the interface to the fake nameserver for the new
192 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
193 into the default Received: header string.
200 TK/01 Added the "success_on_redirect" address verification option. See
201 NewStuff for rationale and an example.
203 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
205 PH/02 Patch to exigrep to allow it to work on syslog lines.
207 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
208 fread() to skip over the body file's header line, because in Cygwin the
209 header line is locked and is inaccessible.
211 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
212 co-exist for some time) to make it clear that it is the Exim ID that is
213 referenced, not the Message-ID: header line.
215 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
216 string_format() function, because snprintf() does not exist on all
219 PH/06 The use of forbid_filter_existstest now also locks out the use of the
220 ${stat: expansion item.
222 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
223 protocol synchronization error", to keep the pedants happy.
225 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
226 well as for IRIX systems, when gcc is being used. See the host.c source
229 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
231 PH/10 Named domain lists were not working if used in a queue_smtp_domains
234 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
235 transport and to the smtp transport in LMTP mode.
237 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
239 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
240 run a filter in a subprocess. This could lead to confusion in subsequent
241 lookups in the parent process. There should also be a search_tidyup() at
242 the end of the subprocess.
244 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
245 only if the host matched helo_try_verify_hosts, which caused the
246 verification to occur when the EHLO/HELO command was issued. The ACL just
247 tested the remembered result. Now, if a previous verification attempt has
248 not happened, "verify = helo" does it there and then.
250 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
252 TK/03 Fix log output including CR from clamd.
254 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
255 exist provoked a memory error which could cause a segfault.
257 PH/15 Installed PCRE 6.2
259 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
261 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
262 of the problem. Specifically, suggested +O2 rather than +O1 for the
265 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
267 PH/20 If a delivery was routed to a non-standard port by means of an SRV
268 record, the port was not correctly logged when the outgoing_port log
269 selector was set (it logged the transort's default port).
271 PH/21 Added support for host-specific ports to manualroute, queryprogram,
272 fallback_hosts, and "hosts" in the smtp transport.
274 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
275 host errors such as "Connection refused".
277 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
278 authentication with radiusclient 0.4.9:
280 - Error returned from rc_read_config was caught wrongly
281 - Username/password not passed on to radius server due to wrong length.
283 The presumption is that some radiusclient API changes for 4.51/PH/17
284 were not taken care of correctly. The code is still untested by me (my
285 Linux distribution still has 0.3.2 of radiusclient), but it was
286 contributed by a Radius user.
288 PH/24 When doing a callout, the value of $domain wasn't set correctly when
289 expanding the "port" option of the smtp transport.
291 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
292 while reading a MIME header. Thanks to Tom Hughes for a patch.
294 PH/24 Include config.h inside local_scan.h so that configuration settings are
297 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
298 that in an ACL for RCPT (for example), you can examine exactly what was
301 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
302 commands, but it was not correctly comparing the address with the actual
303 client host address. Thus, it would show the EHLO address in Received:
304 header lines when this was not necessary.
306 PH/27 Added the % operator to ${eval:}.
308 PH/28 Exim tries to create and chdir to its spool directory when it starts;
309 it should be ignoring failures (because with -C, for example, it has lost
310 privilege). It wasn't ignoring creation failures other than "already
313 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
316 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
317 because an input line was too long, either on its own, or by virtue of
318 too many continuations, the temporary file was not being removed, and the
319 return code was incorrect.
321 PH/31 Missing "BOOL" in function definition in filtertest.c.
323 PH/32 Applied Sieve patches from the maintainer.
325 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
327 PH/33 Added "verify = not_blind".
329 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
330 Local/Makefile (with some defaults set). These are used in built scripts
331 such as exicyclog, but they have never been used in the exim_install
332 script (though there are many overriding facilities there). I have
333 arranged that the exim_install script now takes note of these two
336 PH/35 Installed configuration files for Dragonfly.
338 PH/36 When a locally submitted message by a trusted user did not contain a
339 From: header, and the sender address was obtained from -f or from an SMTP
340 MAIL command, and the trusted user did not use -F to supply a sender
341 name, $originator_name was incorrectly used when constructing a From:
342 header. Furthermore, $originator_name was used for submission mode
343 messages from external hosts without From: headers in a similar way,
344 which is clearly wrong.
346 PH/37 Added control=suppress_local_fixups.
348 PH/38 When log_selector = +received_sender was set, and the addition of the
349 sender made the log line's construction buffer exactly full, or one byte
350 less than full, an overflow happened when the terminating "\n" was
353 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
354 when the result of a list match is failure because a DNS lookup failed.
356 PH/40 RM_COMMAND is now used in the building process.
358 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
359 the "build-* directories that it finds.
361 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
362 address in a domain literal was a prefix of an interface address.
364 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
365 when verifying a sender address, unless rewrite_headers is false.
367 PH/44 Wrote a long comment about why errors_to addresses are verified as
368 recipients, not senders.
370 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
371 the ratelimit ACL was added.
373 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
375 PH/46 Added extra information about PostgreSQL errors to the error string.
377 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
378 avoiding the need to install special zones in a real server. This is
379 backwards compatible; if it can't find the fake resolver, it drops back.
380 Thus, both old and new test suites can be run.
382 TF/02 Added util/ratelimit.pl
384 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
385 clock is set back in time.
387 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
388 Candler <B.Candler@pobox.com>.
390 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
391 OK for addresses that are the result of redirections.
393 PH/48 A number of further additions for the benefit of the new test suite,
394 including a fake gethostbyname() that interfaces to the fake DNS resolver
397 TF/06 The fix for widen_domains has also been applied to qualify_single and
398 search_parents which are the other dnslookup options that can cause
401 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
404 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
406 TF/07 Exim produced the error message "an SRV record indicated no SMTP
407 service" if it encountered an MX record with an empty target hostname.
408 The message is now "an MX or SRV record indicated no SMTP service".
410 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
411 if the DNS of the sending site is misconfigured. This is quite a
412 common situation. This change restores the behaviour of treating a
413 helo verification defer as a failure.
415 PH/51 If self=fail was set on a router, the bounce message did not include the
416 actual error message.
422 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
424 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
425 command itself ended in error, the underlying message about the transport
426 filter timeout was being overwritten with the pipe command error. Now the
427 underlying error message should be appended to the second error message.
429 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
431 PH/02 Reduce the amount of output that "make" produces by default. Full output
432 can still be requested.
434 PH/03 The warning log line about a condition test deferring for a "warn" verb
435 was being output only once per connection, rather than after each
436 occurrence (because it was using the same function as for successful
437 "warn" verbs). This seems wrong, so I have changed it.
439 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
440 it should not have, which might have caused a crash in the right
441 circumstances, but probably never did.
443 PH/04 Installed a modified version of Tony Finch's patch to make submission
444 mode fix the return path as well as the Sender: header line, and to
445 add a /name= option so that you can make the user's friendly name appear
448 TF/03 Added the control = fakedefer ACL modifier.
450 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
451 Mark Lowes for thorough testing.
453 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
455 TK/03 Merged latest SRS patch from Miles Wilton.
457 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
458 with the definition in sysexits.h (which is #included earlier).
459 Fortunately, Exim does not actually use EX_OK. The code used to try to
460 preserve the sysexits.h value, by assumimg that macro definitions were
461 scanned for macro replacements. I have been disabused of this notion,
462 so now the code just undefines EX_OK before #including unistd.h.
464 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
465 in the smtp transport. When a block could not be written in a single
466 write() function, the timeout was being re-applied to each part-write.
467 This seems wrong - if the receiver was accepting one byte at a time it
468 would take for ever. The timeout is now adjusted when this happens. It
469 doesn't have to be particularly precise.
471 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
472 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
474 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
475 without a domain if the check to <postmaster@domain> fails.
477 SC/01 Eximstats: added -xls and the ability to specify output files
478 (patch written by Frank Heydlauf).
480 SC/02 Eximstats: use FileHandles for outputing results.
482 SC/03 Eximstats: allow any combination of xls, txt, and html output.
484 SC/04 Eximstats: fixed display of large numbers with -nvr option
486 SC/05 Eximstats: fixed merging of reports with empty tables.
488 SC/06 Eximstats: added the -include_original_destination flag
490 SC/07 Eximstats: removed tabs and trailing whitespace.
492 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
494 TK/06 MBOX spool code: Add real "From " MBOX separator line
495 so the .eml file is really in mbox format (even though
496 most programs do not really care). Patch from Alex Miller.
498 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
499 The latter is generated from $received_to and is only set if the
500 message has one envelope recipient. SA can use these headers,
501 obviously out-of-the-box. Patch from Alex Miller.
503 PH/08 The ${def test on a variable was returning false if the variable's
504 value was "0", contrary to what the specification has always said!
505 The result should be true unless the variable is empty.
507 PH/09 The syntax error of a character other than { following "${if
508 def:variable_name" (after optional whitespace) was not being diagnosed.
509 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
510 accidental colon was present, for example, could give incorrect results.
512 PH/10 Tidied the code in a number of places where the st_size field of a stat()
513 result is used (not including appendfile, where other changes are about
516 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
517 This involved changing a lot of size variables from int to off_t. It
518 should work with maildirs and everything.
520 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
521 spamd dying while we are connected to it.
523 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
526 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
527 function for BDB 4.3.
529 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
531 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
532 the sources for PCRE have been re-arranged and more files are now
535 PH/15 The code I had for printing potentially long long variables in PH/11
536 above was not the best (it lost precision). The length of off_t variables
537 is now inspected at build time, and an appropriate printing format (%ld
538 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
539 to be "long long int" or "long int". This is needed for the internal
540 formatting function string_vformat().
542 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
543 the configuration file to be ":syslog", then the script "guesses" where
544 the logs files are, rather than using the compiled in default. In our
545 case the guess is not the same as the compiled default, so the script
546 suddenly stopped working when I started to use syslog. The patch checks
547 to see if log_file_path is "". If so, it attempts to read it from exim
548 with no configuration file to get the compiled in version, before it
549 falls back to the previous guessing code."
551 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
552 implementing BATV in an Exim configuration. See NewStuff for the gory
555 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
556 Makefile that are specific to HP-UX.
558 PH/18 If the "use_postmaster" option was set for a recipient callout together
559 with the "random" option, the postmaster address was used as the MAIL
560 FROM address for the random test, but not for the subsequent recipient
561 test. It is now used for both.
563 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
564 patch removes a few documentation additions to RFC 3028, because the
565 latest draft now contains them. It adds the new en;ascii-case comparator
566 and a new error check for 8bit text in MIME parts. Comparator and
567 require names are now matched exactly. I enabled the subaddress
568 extension, but it is not well tested yet (read: it works for me)."
570 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
571 rework some of the code of TK/09 above to avoid the hardwired use of
572 "%lld" and "long long". Replaced the call to snprintf() with a call to
575 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
576 records point to non-existent hosts", "retry timeout exceeded", and
577 "retry time not reached for any host after a long failure period".
579 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
580 experimental DomainKeys support:
582 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
583 (2) On an error such as an illegally used "control", the wrong name for
584 the control was given.
586 These problems did NOT occur unless DomainKeys support was compiled.
588 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
590 PH/24 Added ${if match_ip condition.
592 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
593 databases so that it will be absolutely obvious if a crash occurs in the
594 DB library. This is a regular occurrence (often caused by mis-matched
597 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
598 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
599 picked up on a user's system that detects such things. There doesn't seem
600 to be a gcc warning option for this - only an attribute that has to be
601 put on the function's prototype. It seems that in Fedora Core 4 they have
602 set this on a number of new functions. No doubt there will be more in due
605 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
606 specify a transport. However, if an address that was verified by such a
607 router was the subject of a callout, Exim crashed because it tried to
608 read the rcpt_include_affixes from the non-existent transport. Now it
609 just assumes that the setting of that option is false. This bug was
610 introduced by 4.51/PH/31.
612 PH/28 Changed -d+all to exclude +memory, because that information is very
613 rarely of interest, but it makes the output a lot bigger. People tend to
614 do -d+all out of habit.
616 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
617 code in os-type was giving problems when libc.so lives in lib64, like on
620 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
621 aren't the modern standard, and it seems that some systems' include files
622 don't always have them. Exim was already checking for some of the newer
623 ones like T_AAAA, and defining it itself. I've added checks for all the
624 record types that Exim uses.
626 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
627 not automatically generating a new one, as it is supposed to. This
628 prevented TLS from working. If the file did exist, but contained invalid
629 data, a new version was generated, as expected. It was only the case of a
630 non-existent file that was broken.
632 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
633 with a change in libdomainkeys > 0.64.
635 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
636 from DNS. If the selector record carries the flag, it now has
637 precedence over the domain-wide flag.
639 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
641 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
642 the use of a port name that isn't defined in /etc/services) occurred, the
643 message was deferred as in a normal delivery, and thus remained on the
644 spool, instead of being failed because of the mua_wrapper setting. This
645 is now fixed, and I tidied up some of the mua_wrapper messages at the
648 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
649 the messages in a hash of arrays rather than using individual hashes.
650 This is a bit cleaner and results in dramatic memory savings, albeit
651 at a slight CPU cost.
653 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
654 as requested by Marc Sherman.
656 SC/10 Eximstats: added histograms for user specified patterns as requested
659 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
661 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
662 fopen() in the content-scanning modules that did not already have it.
668 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
669 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
671 TK/02 Fix ACL "control" statement not being available in MIME ACL.
673 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
675 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
676 to test Sieve filters that use "vacation".
678 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
679 that changes the way the GnuTLS parameters are stored in the cache file.
680 The new format can be generated externally. For backward compatibility,
681 if the data in the cache doesn't make sense, Exim assumes it has read an
682 old-format file, and it generates new data and writes a new file. This
683 means that you can't go back to an older release without removing the
686 PH/03 A redirect router that has both "unseen" and "one_time" set does not
687 work if there are any delivery delays because "one_time" forces the
688 parent to be marked "delivered", so its unseen clone is never tried
689 again. For this reason, Exim now forbids the simultaneous setting of
692 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
693 redirected to themselves ("homonym" addresses). Read the long ChangeLog
694 entry if you want to know the details. The fix, however, neglected to
695 consider the case when local delivery batching is involved. The test for
696 "previously delivered" was not happening when checking to see if an
697 address could be batched with a previous (undelivered) one; under
698 certain circumstances this could lead to multiple deliveries to the same
701 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
702 in its include files, and this causes problems building Exim.
704 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
705 header_syntax) but Exim was just ignoring anything given after a slash.
706 In particular, this caused confusion with an attempt to use "verify =
707 reverse_host_lookup/defer_ok". An error is now given when options are
708 supplied for verify items that do not have them. (Maybe reverse_host_
709 lookup should have a defer_ok option, but that's a different point.)
711 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
712 defined by RFC 821) to 2048, because there were problems with some AUTH
713 commands, and RFC 1869 says the size should be increased for extended
714 SMTP commands that take arguments.
716 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
719 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
720 "unknown" error; now it says that the functionality isn't in the binary.
722 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
723 an address' error message when a string expansion fails (syntax or
724 whatever). Otherwise the password may appear in the log. Following change
725 PH/42 below, there is no longer a chance of it appearing in a bounce
728 PH/11 Installed exipick version 20050225.0 from John Jetmore.
730 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
731 of its addresses was ever tried. (Bugzilla bug #2.)
733 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
734 the result incorrectly in the debug output. (It correctly added a newline
735 to what was transported.)
737 TF/01 Added $received_time.
739 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
740 commented out examples of how to interface to a virus scanner and to
741 SpamAssassin. Also added commented examples of av_scanner and
742 spamd_address settings.
744 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
745 and controls are allowed in which ACLs. There were a couple of minor
746 errors. Some of the entries in the conditions table (which is a table of
747 where they are NOT allowed) were getting very unwieldy; rewrote them as a
748 negation of where the condition IS allowed.
750 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
752 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
753 header file does not have a version number, so I've had to invent a new
754 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
755 API. The code is untested by me (my Linux distribution still has 0.3.2 of
756 radiusclient), but it was contributed by a Radius user.
758 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
759 files or directories, for OpenSSL.
761 PH/19 When an Exim process that is running as root has to create an Exim log
762 file, it does so in a subprocess that runs as exim:exim so as to get the
763 ownership right at creation (otherwise, other Exim processes might see
764 the file with the wrong ownership). There was no test for failure of this
765 fork() call, which would lead to the process getting stuck as it waited
766 for a non-existent subprocess. Forks do occasionally fail when resources
767 run out. I reviewed all the other calls to fork(); they all seem to check
770 PH/20 When checking for unexpected SMTP input at connect time (before writing
771 the banner), Exim was not dealing correctly with a non-positive return
772 from the read() function. If the client had disconnected by this time,
773 the result was a log entry for a synchronization error with an empty
774 string after "input=" when read() returned zero. If read() returned -1
775 (an event I could not check), uninitialized data bytes were printed.
776 There were reports of junk text (parts of files, etc) appearing after
779 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
781 PH/22 Added support for macro redefinition, and (re)definition in between
782 driver and ACL definitions.
784 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
785 forgetting to use the resulting value; it was using the unexpanded value.
787 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
788 hadn't been configured. The fix is from Juergen Kreileder, who
789 understands it better than I do:
791 "Here's what I see happening with three configured cyrus_sasl
792 authenticators configured (plain, login, cram-md5):
794 On startup auth_cyrus_sasl_init() gets called for each of these.
795 This means three calls to sasl_listmech() without a specified mech_list.
796 => SASL tests which mechs of all available mechs actually work
797 => three warnings about OTP not working
798 => the returned list contains: plain, login, cram-md5, digest-md5, ...
800 With the patch, sasl_listmech() also gets called three times. But now
801 SASL's mech_list option is set to the server_mech specified in the the
802 authenticator. Or in other words, the answer from sasl_listmech()
803 gets limited to just the mech you're testing for (which is different
805 => the return list contains just 'plain' or 'login', 'cram-md5' or
806 nothing depending on the value of ob->server_mech.
808 I've just tested the patch: Authentication still works fine,
809 unavailable mechs specified in the exim configuration are still
810 caught, and the auth.log warnings about OTP are gone."
812 PH/25 When debugging is enabled, the contents of the command line are added
813 to the debugging output, even when log_selector=+arguments is not
816 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
817 answer is "GNU", and only if the return is "GNU/something" is the answer
820 PH/27 $acl_verify_message is now set immediately after the failure of a
821 verification in an ACL, and so is available in subsequent modifiers. In
822 particular, the message can be preserved by coding like this:
824 warn !verify = sender
825 set acl_m0 = $acl_verify_message
827 Previously, $acl_verify_message was set only while expanding "message"
828 and "log_message" when a very denied access.
830 PH/28 Modified OS/os.c-Linux with
832 -#ifndef OS_LOAD_AVERAGE
833 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
835 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
836 nomenclature these days.)
838 PH/29 Installed patch from the Sieve maintainer that adds the options
839 sieve_useraddress and sieve_subaddress to the redirect router.
841 PH/30 In these circumstances:
842 . Two addresses routed to the same list of hosts;
843 . First host does not offer TLS;
844 . First host accepts first address;
845 . First host gives temporary error to second address;
846 . Second host offers TLS and a TLS session is established;
847 . Second host accepts second address.
848 Exim incorrectly logged both deliveries with the TLS parameters (cipher
849 and peerdn, if requested) that were in fact used only for the second
852 PH/31 When doing a callout as part of verifying an address, Exim was not paying
853 attention to any local part prefix or suffix that was matched by the
854 router that accepted the address. It now behaves in the same way as it
855 does for delivery: the affixes are removed from the local part unless
856 rcpt_include_affixes is set on the transport.
858 PH/32 Add the sender address, as F=<...>, to the log line when logging a
859 timeout during the DATA phase of an incoming message.
861 PH/33 Sieve envelope tests were broken for match types other than :is. I have
862 applied a patch sanctioned by the Sieve maintainer.
864 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
865 the uid or gid is negative. A case of a negative gid caused this to be
866 noticed. The fix allows for either to be negative.
868 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
869 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
872 PH/36 The change PH/12 above was broken. Fixed it.
874 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
875 the grounds that routing the same address twice would always produce the
876 same answer. This might have been true once, but it is certainly no
877 longer true now. Routing a child address may depend on the previous
878 routing that produced that child. Some complicated redirection strategies
879 went wrong when messages had multiple recipients, and made Exim's
880 behaviour dependent on the order in which the addresses were given.
882 I have moved the duplicate checking until after the routing is complete.
883 Exim scans the addresses that are assigned to local and remote
884 transports, and removes any duplicates. This means that more work will be
885 done, as duplicates will always all be routed, but duplicates are
886 presumably rare, so I don't expect this is of any significance.
888 For deliveries to pipes, files, and autoreplies, the duplicate checking
889 still happens during the routing process, since they are not going to be
892 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
893 It corrects a timeout issue with spamd. This is Ian's comment: "The
894 background is that sometimes spamd either never reads data from a
895 connection it has accepted, or it never writes response data. The exiscan
896 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
897 blindly assumes that writes won't block so it may never time out."
899 PH/39 Allow G after quota size as well as K and M.
901 PH/40 The value set for $authenticated_id in an authenticator may not contain
902 binary zeroes or newlines because the value is written to log lines and
903 to spool files. There was no check on this. Now the value is run through
904 the string_printing() function so that such characters are converted to
905 printable escape sequences.
907 PH/41 $message_linecount is a new variable that contains the total number of
908 lines in the message. Compare $body_linecount, which is the count for the
911 PH/42 Exim no longer gives details of delivery errors for specific addresses in
912 bounce and delay warning messages, except in certain special cases, which
915 (a) An SMTP error message from a remote host;
916 (b) A message specified in a :fail: redirection;
917 (c) A message specified in a "fail" command in a system filter;
918 (d) A message specified in a FAIL return from the queryprogram router;
919 (e) A message specified by the cannot_route_message router option.
921 In these cases only, Exim does include the error details in bounce and
922 warning messages. There are also a few cases where bland messages such
923 as "unrouteable address" or "local delivery error" are given.
925 PH/43 $value is now also set for the "else" part of a ${run expansion.
927 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
928 being worked on, but at least Exim now implements the latest version to
931 PH/45 In a pipe transport, although a timeout while waiting for the pipe
932 process to complete was treated as a delivery failure, a timeout while
933 writing the message to the pipe was logged, but erroneously treated as a
934 successful delivery. Such timeouts include transport filter timeouts. For
935 consistency with the overall process timeout, these timeouts are now
936 treated as errors, giving rise to delivery failures by default. However,
937 there is now a new Boolean option for the pipe transport called
938 timeout_defer, which, if set TRUE, converts the failures into defers for
939 both kinds of timeout. A transport filter timeout is now identified in
942 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
943 systems where "make" and "gmake" are different, calling "gmake" at top
944 level broke things. I've arranged for the value of $(MAKE) to be passed
945 from the Makefile to this script so that it can call the same version of
949 A note about Exim versions 4.44 and 4.50
950 ----------------------------------------
952 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
953 changes of various kinds. As a consequence, a big documentation update was
954 needed. This delayed the release for rather longer than seemed good, especially
955 in the light of a couple of (minor) security issues. Therefore, the changes
956 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
957 release. So 4.44 and 4.50 are in effect two different branches that both start
960 I have left the 4.50 change log unchanged; it contains all the changes since
961 4.43. The change log for 4.44 is below; many of its items are identical to
962 those for 4.50. This seems to be the most sensible way to preserve the
963 historical information.
969 1. Minor wording change to the doc/README.SIEVE file.
971 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
972 computation of the current number of files was incorrect.
974 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
975 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
978 4. Give more explanation in the error message when the command for a transport
979 filter fails to execute.
981 5. There are several places where Exim runs a non-Exim command in a
982 subprocess. The SIGUSR1 signal should be disabled for these processes. This
983 was being done only for the command run by the queryprogram router. It is
984 now done for all such subprocesses. The other cases are: ${run, transport
985 filters, and the commands run by the lmtp and pipe transports.
987 6. Added CONFIGURE_GROUP build-time option.
989 7. Some older OS have a limit of 256 on the maximum number of file
990 descriptors. Exim was using setrlimit() to set 1000 as a large value
991 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
992 systems. I've change it so that if it can't get 1000, it tries for 256.
994 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
995 was an oversight, and furthermore, ever since the addition of extra
996 controls (e.g. 4.43/32), the checks on when to allow different forms of
997 "control" were broken. There should now be diagnostics for all cases when a
998 control that does not make sense is encountered.
1000 9. Added the /retain_sender option to "control=submission".
1002 10. $recipients is now available in the predata ACL (oversight).
1004 11. Tidy the search cache before the fork to do a delivery from a message
1005 received from the command line. Otherwise the child will trigger a lookup
1006 failure and thereby defer the delivery if it tries to use (for example) a
1007 cached ldap connection that the parent has called unbind on.
1009 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1010 of $address_data from the recipient verification was clobbered by the
1011 sender verification.
1013 13. The value of address_data from a sender verification is now available in
1014 $sender_address_data in subsequent conditions in the ACL statement.
1016 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1018 15. Added a new option "connect=<time>" to callout options, to set a different
1021 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1022 was its contents. (It was OK if the option was not defined at all.)
1024 17. A "Completed" log line is now written for messages that are removed from
1025 the spool by the -Mrm option.
1027 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1028 information about exactly what failed.
1030 19. Added -dd to debug only the daemon process.
1032 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1033 handles timeouts, both on the server side and network timeouts. Renamed the
1034 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1036 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1037 It is now set to "smtps".
1039 22. $host_address is now set to the target address during the checking of
1040 ignore_target_hosts.
1042 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1043 being passed; this would have caused $sender_host_name to have been used if
1044 matching the list had actually called for a host name (not very likely,
1045 since this list is usually IP addresses). A host name is now passed as
1048 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1049 code that specifies a non-restarting handler (typically sigaction() in
1050 modern systems) in an attempt to fix a rare and obscure crash bug.
1052 25. Narrowed the window for a race in the daemon that could cause it to ignore
1053 SIGCHLD signals. This is not a major problem, because they are used only to
1054 wake it up if nothing else does.
1056 26. A malformed maildirsize file could cause Exim to calculate negative values
1057 for the mailbox size or file count. Odd effects could occur as a result.
1058 The maildirsize information is now recalculated if the size or filecount
1061 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1062 support for a long time. Removed HAVE_SYS_VFS_H.
1064 28. Installed the latest version of exipick from John Jetmore.
1066 29. In an address list, if the pattern was not a regular expression, an empty
1067 subject address (from a bounce message) matched only if the pattern was an
1068 empty string. Non-empty patterns were not even tested. This was the wrong
1069 because it is perfectly reasonable to use an empty address as part of a
1070 database query. An empty address is now tested by patterns that are
1071 lookups. However, all the other forms of pattern expect the subject to
1072 contain a local part and a domain, and therefore, for them, an empty
1073 address still always fails if the pattern is not itself empty.
1075 30. Exim went into a mad DNS loop when attempting to do a callout where the
1076 host was specified on an smtp transport, and looking it up yielded more
1077 than one IP address.
1079 31. Re-factored the code for checking spool and log partition space into a
1080 function that finds that data and another that does the check. The former
1081 is then used to implement four new variables: $spool_space, $log_space,
1082 $spool_inodes, and $log_inodes.
1084 32. The RFC2047 encoding function was originally intended for short strings
1085 such as real names; it was not keeping to the 75-character limit for
1086 encoded words that the RFC imposes. It now respects the limit, and
1087 generates multiple encoded words if necessary. To be on the safe side, I
1088 have increased the buffer size for the ${rfc2047: expansion operator from
1091 33. It is now permitted to omit both strings after an "if" condition; if the
1092 condition is true, the result is "true". As before, when the second string
1093 is omitted, a false condition yields an empty string. This makes it less
1094 cumbersome to write custom ACL and router conditions.
1096 34. Failure to deliver a bounce message always caused it to be frozen, even if
1097 there was an errors_to setting on the router. The errors_to setting is now
1100 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1101 canonical form (fully expanded) before being placed in
1102 $sender_host_address.
1104 36. The table in the code that translates DNS record types into text (T_A to
1105 "A" for instance) was missing entries for NS and CNAME. It is just possible
1106 that this could have caused confusion if both these types were looked up
1107 for the same domain, because the text type is used as part of Exim's
1108 per-process caching. But the chance of anyone hitting this buglet seems
1111 37. The dnsdb lookup has been extended in a number of ways.
1113 (1) There is a new type, "zns", which walks up the domain tree until it
1114 finds some nameserver records. It should be used with care.
1116 (2) There is a new type, "mxh", which is like "mx" except that it returns
1117 just the host names, not the priorities.
1119 (3) It is now possible to give a list of domains (or IP addresses) to be
1120 looked up. The behaviour when one of the lookups defers can be
1121 controlled by a keyword.
1123 (4) It is now possible to specify the separator character for use when
1124 multiple records are returned.
1126 38. The dnslists ACL condition has been extended: it is now possible to supply
1127 a list of IP addresses and/or domains to be looked up in a particular DNS
1130 39. Added log_selector=+queue_time_overall.
1132 40. When running the queue in the test harness, wait just a tad after forking a
1133 delivery process, to get repeatability of debugging output.
1135 41. Include certificate and key file names in error message when GnuTLS fails
1136 to set them up, because the GnuTLS error message doesn't include the name
1137 of the failing file when there is a problem reading it.
1139 42. Allow both -bf and -bF in the same test run.
1141 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1143 44. The "Exiscan patch" is now merged into the mainline Exim source.
1145 45. Sometimes the final signoff response after QUIT could fail to get
1146 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1147 < 0 before doing a fflush(). This bug looks as though it goes back to the
1148 introduction of TLS in release 3.20, but "sometimes" must have been rare
1149 because the tests only now provoked it.
1151 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1152 (this can affect the format of dates).
1154 47. exim_tidydb, when checking for the continued existence of a message for
1155 which it has found a message-specific retry record, was not finding
1156 messages that were in split spool directories. Consequently, it was
1157 deleting retry records that should have stayed in existence.
1159 48. Steve fixed some bugs in eximstats.
1161 49. The SPA authentication driver was not abandoning authentication and moving
1162 on to the next authenticator when an expansion was forced to fail,
1163 contradicting the general specification for all authenticators. Instead it
1164 was generating a temporary error. It now behaves as specified.
1166 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1167 (the order specifies the preference for clients). The order is now AES256,
1168 AES128, 3DES, ARCFOUR128.
1170 51. Small patch to Sieve code - explicitly set From: when generating an
1173 52. Exim crashed if a remote delivery caused a very long error message to be
1174 recorded - for instance if somebody sent an entire SpamAssassin report back
1175 as a large number of 550 error lines. This bug was coincidentally fixed by
1176 increasing the size of one of Exim's internal buffers (big_buffer) that
1177 happened as part of the Exiscan merge. However, to be on the safe side, I
1178 have made the code more robust (and fixed the comments that describe what
1181 53. Now that there can be additional text after "Completed" in log lines (if
1182 the queue_time_overall log selector is set), a one-byte patch to exigrep
1183 was needed to allow it to recognize "Completed" as not the last thing in
1186 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1187 patch that reportedly fixes this has been added. I am not expert enough to
1188 create a test for it. This is what the patch creator wrote:
1190 "I found a little strange behaviour of ldap code when working with
1191 Windows 2003 AD Domain, where users was placed in more than one
1192 Organization Units. When I tried to give exim partial DN, the exit code
1193 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1194 But simultaneously result of request was absolutely normal ldap result,
1195 so I produce this patch..."
1197 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1198 REFERENCE, so I have modified the code to exclude the patch when that macro
1201 55. Some experimental protocols are using DNS PTR records for new purposes. The
1202 keys for these records are domain names, not reversed IP addresses. The
1203 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1204 leaves it alone. Component reversal etc. now happens only for IP addresses.
1207 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1209 57. Double the size of the debug message buffer (to 2048) so that more of very
1210 long debug lines gets shown.
1212 58. The exicyclog utility now does better if the number of log files to keep
1213 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1215 59. Two changes related to the smtp_active_hostname option:
1217 (1) $smtp_active_hostname is now available as a variable.
1218 (2) The default for smtp_banner uses $smtp_active_hostname instead
1219 of $primary_hostname.
1221 60. The host_aton() function is supposed to be passed a string that is known
1222 to be a valid IP address. However, in the case of IPv6 addresses, it was
1223 not checking this. This is a hostage to fortune. Exim now panics and dies
1224 if the condition is not met. A case was found where this could be provoked
1225 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1226 components; fortuitously, this particular loophole had already been fixed
1227 by change 4.50/55 above.
1229 If there are any other similar loopholes, the new check in host_aton()
1230 itself should stop them being exploited. The report I received stated that
1231 data on the command line could provoke the exploit when Exim was running as
1232 exim, but did not say which command line option was involved. All I could
1233 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1234 running as the user.
1237 61. There was a buffer overflow vulnerability in the SPA authentication code
1238 (which came originally from the Samba project). I have added a test to the
1239 spa_base64_to_bits() function which I hope fixes it.
1242 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1243 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1245 63. The daemon start-up calls getloadavg() while still root for those OS that
1246 need the first call to be done as root, but it missed one case: when
1247 deliver_queue_load_max is set with deliver_drop_privilege. This is
1248 necessary for the benefit of the queue runner, because there is no re-exec
1249 when deliver_drop_privilege is set.
1251 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1252 This has been fixed.
1254 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1255 was in use, was not putting the data itself into the right store pool;
1256 consequently, it could be overwritten for a subsequent message in the same
1257 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1260 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1262 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1263 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1264 the function were treating the return as a boolean value, which happened to
1265 work because 0=false and not-0=true, but is not correct code.
1267 68. The host_aton() function was not handling scoped IPv6 addresses (those
1268 with, for example, "%eth0" on the end) correctly.
1270 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1271 negated items (that is, ~something) in unsigned ints. Some compilers
1272 apparently mutter when there is no cast.
1274 70. If an address verification called from an ACL failed, and did not produce a
1275 user-specific message (i.e. there was only a "system" message), nothing was
1276 put in $acl_verify_message. In this situation, it now puts the system
1279 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1280 session; change 4.31/43 added the unwanted input to the log line - except
1281 that it did not do this in the start of session case. It now does.
1283 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1284 This is wrong and can cause the other end to generate a synchronization
1285 error if it is another Exim or anything else that does the synchronization
1286 check. A QUIT command is no longer sent after a timeout.
1288 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1289 during host lookups.
1291 74. The defer_ok option of callout verification was not working if it was used
1292 when verifying addresses in header lines, that is, for this case:
1294 verify = header_sender/callout=defer_ok
1296 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1297 those file descriptors could be used for SMTP connections. If anything
1298 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1299 could be sent to the SMTP client, causing chaos. The daemon now opens
1300 stdin, stdout, and stderr to /dev/null when it puts itself into the
1303 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1304 log by default. The user can override this with suitable Perl magic.
1306 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1307 the log message when discard triggers, was not working for the DATA ACL or
1308 for the non-SMTP ACL.
1310 78. Error message wording change in sieve.c.
1312 79. If smtp_accept_max_per_host was set, the number of connections could be
1313 restricted to fewer than expected, because the daemon was trying to set up
1314 a new connection before checking whether the processes handling previous
1315 connections had finished. The check for completed processes is now done
1316 earlier. On busy systems, this bug wouldn't be noticed because something
1317 else would have woken the daemon, and it would have reaped the completed
1320 80. If a message was submitted locally by a user whose login name contained one
1321 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1322 It caused a spool format error. I have fixed the spool reading code. A
1323 related problem was that the "from" clause in the Received: line became
1324 illegal because of the space(s). It is now covered by ${quote_local_part.
1326 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1329 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1331 83. Patch from Sieve maintainer to fix unterminated string problem in
1332 "vacation" handling.
1334 84. Some minor changes to the Linux configuration files to help with other
1335 OS variants using glibc.
1337 85. One more patch for Sieve to update vacation handling to latest spec.
1340 ----------------------------------------------------
1341 See the note above about the 4.44 and 4.50 releases.
1342 ----------------------------------------------------
1348 1. Change 4.43/35 introduced a bug that caused file counts to be
1349 incorrectly computed when quota_filecount was set in an appendfile
1352 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1353 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1356 3. Give more explanation in the error message when the command for a transport
1357 filter fails to execute.
1359 4. There are several places where Exim runs a non-Exim command in a
1360 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1361 was being done only for the command run by the queryprogram router. It is
1362 now done for all such subprocesses. The other cases are: ${run, transport
1363 filters, and the commands run by the lmtp and pipe transports.
1365 5. Some older OS have a limit of 256 on the maximum number of file
1366 descriptors. Exim was using setrlimit() to set 1000 as a large value
1367 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1368 systems. I've change it so that if it can't get 1000, it tries for 256.
1370 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1371 was an oversight, and furthermore, ever since the addition of extra
1372 controls (e.g. 4.43/32), the checks on when to allow different forms of
1373 "control" were broken. There should now be diagnostics for all cases when a
1374 control that does not make sense is encountered.
1376 7. $recipients is now available in the predata ACL (oversight).
1378 8. Tidy the search cache before the fork to do a delivery from a message
1379 received from the command line. Otherwise the child will trigger a lookup
1380 failure and thereby defer the delivery if it tries to use (for example) a
1381 cached ldap connection that the parent has called unbind on.
1383 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1384 of $address_data from the recipient verification was clobbered by the
1385 sender verification.
1387 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1388 was its contents. (It was OK if the option was not defined at all.)
1390 11. A "Completed" log line is now written for messages that are removed from
1391 the spool by the -Mrm option.
1393 12. $host_address is now set to the target address during the checking of
1394 ignore_target_hosts.
1396 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1397 being passed; this would have caused $sender_host_name to have been used if
1398 matching the list had actually called for a host name (not very likely,
1399 since this list is usually IP addresses). A host name is now passed as
1402 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1403 code that specifies a non-restarting handler (typically sigaction() in
1404 modern systems) in an attempt to fix a rare and obscure crash bug.
1406 15. Narrowed the window for a race in the daemon that could cause it to ignore
1407 SIGCHLD signals. This is not a major problem, because they are used only to
1408 wake it up if nothing else does.
1410 16. A malformed maildirsize file could cause Exim to calculate negative values
1411 for the mailbox size or file count. Odd effects could occur as a result.
1412 The maildirsize information is now recalculated if the size or filecount
1415 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1416 support for a long time. Removed HAVE_SYS_VFS_H.
1418 18. Updated exipick to current release from John Jetmore.
1420 19. Allow an empty sender to be matched against a lookup in an address list.
1421 Previously the only cases considered were a regular expression, or an
1424 20. Exim went into a mad DNS lookup loop when doing a callout where the
1425 host was specified on the transport, if the DNS lookup yielded more than
1428 21. The RFC2047 encoding function was originally intended for short strings
1429 such as real names; it was not keeping to the 75-character limit for
1430 encoded words that the RFC imposes. It now respects the limit, and
1431 generates multiple encoded words if necessary. To be on the safe side, I
1432 have increased the buffer size for the ${rfc2047: expansion operator from
1435 22. Failure to deliver a bounce message always caused it to be frozen, even if
1436 there was an errors_to setting on the router. The errors_to setting is now
1439 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1440 canonical form (fully expanded) before being placed in
1441 $sender_host_address.
1443 24. Updated eximstats to version 1.33
1445 25. Include certificate and key file names in error message when GnuTLS fails
1446 to set them up, because the GnuTLS error message doesn't include the name
1447 of the failing file when there is a problem reading it.
1449 26. Expand error message when OpenSSL has problems setting up cert/key files.
1452 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1453 (this can affect the format of dates).
1455 28. exim_tidydb, when checking for the continued existence of a message for
1456 which it has found a message-specific retry record, was not finding
1457 messages that were in split spool directories. Consequently, it was
1458 deleting retry records that should have stayed in existence.
1460 29. eximstats updated to version 1.35
1461 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1462 1.35 - bugfix such that pie charts by volume are generated correctly
1464 30. The SPA authentication driver was not abandoning authentication and moving
1465 on to the next authenticator when an expansion was forced to fail,
1466 contradicting the general specification for all authenticators. Instead it
1467 was generating a temporary error. It now behaves as specified.
1469 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1470 (the order specifies the preference for clients). The order is now AES256,
1471 AES128, 3DES, ARCFOUR128.
1473 31. Small patch to Sieve code - explicitly set From: when generating an
1476 32. Exim crashed if a remote delivery caused a very long error message to be
1477 recorded - for instance if somebody sent an entire SpamAssassin report back
1478 as a large number of 550 error lines. This bug was coincidentally fixed by
1479 increasing the size of one of Exim's internal buffers (big_buffer) that
1480 happened as part of the Exiscan merge. However, to be on the safe side, I
1481 have made the code more robust (and fixed the comments that describe what
1484 33. Some experimental protocols are using DNS PTR records for new purposes. The
1485 keys for these records are domain names, not reversed IP addresses. The
1486 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1487 leaves it alone. Component reversal etc. now happens only for IP addresses.
1490 34. The host_aton() function is supposed to be passed a string that is known
1491 to be a valid IP address. However, in the case of IPv6 addresses, it was
1492 not checking this. This is a hostage to fortune. Exim now panics and dies
1493 if the condition is not met. A case was found where this could be provoked
1494 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1495 components; fortuitously, this particular loophole had already been fixed
1496 by change 4.50/55 or 4.44/33 above.
1498 If there are any other similar loopholes, the new check in host_aton()
1499 itself should stop them being exploited. The report I received stated that
1500 data on the command line could provoke the exploit when Exim was running as
1501 exim, but did not say which command line option was involved. All I could
1502 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1503 running as the user.
1506 35. There was a buffer overflow vulnerability in the SPA authentication code
1507 (which came originally from the Samba project). I have added a test to the
1508 spa_base64_to_bits() function which I hope fixes it.
1511 36. The daemon start-up calls getloadavg() while still root for those OS that
1512 need the first call to be done as root, but it missed one case: when
1513 deliver_queue_load_max is set with deliver_drop_privilege. This is
1514 necessary for the benefit of the queue runner, because there is no re-exec
1515 when deliver_drop_privilege is set.
1517 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1518 was in use, was not putting the data itself into the right store pool;
1519 consequently, it could be overwritten for a subsequent message in the same
1520 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1523 38. Sometimes the final signoff response after QUIT could fail to get
1524 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1525 < 0 before doing a fflush(). This bug looks as though it goes back to the
1526 introduction of TLS in release 3.20, but "sometimes" must have been rare
1527 because the tests only now provoked it.
1533 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1534 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1535 Now it returns an integer. A number of calls were still expecting a T/F
1536 return. Fortuitously, in all cases, the tests worked in OK situations,
1537 which is the norm. However, things would have gone wrong on any write
1538 failures on the smtp file descriptor. This function is used when sending
1539 messages over SMTP and also when doing verify callouts.
1541 2. When Exim is called to do synchronous delivery of a locally submitted
1542 message (the -odf or -odi options), it no longer closes stderr before doing
1545 3. Implemented the mua_wrapper option.
1547 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1549 5. Implemented the functions header_remove(), header_testname(),
1550 header_add_at_position(), and receive_remove_recipient(), and exported them
1553 6. If an ACL "warn" statement specified the addition of headers, Exim already
1554 inserted X-ACL-Warn: at the start if there was no header name. However, it
1555 was not making this test for the second and subsequent header lines if
1556 there were newlines in the string. This meant that an invalid header could
1557 be inserted if Exim was badly configured.
1559 7. Allow an ACL "warn" statement to add header lines at the start or after all
1560 the Received: headers, as well as at the end.
1562 8. Added the rcpt_4xx retry error code.
1564 9. Added postmaster_mailfrom=xxx to callout verification option.
1566 10. Added mailfrom=xxxx to the callout verification option, for verify=
1569 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1570 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1572 12. Inserted some casts to stop certain compilers warning when using pointer
1573 differences as field lengths or precisions in printf-type calls (mostly
1574 affecting debugging statements).
1576 13. Added optional readline() support for -be (dynamically loaded).
1578 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1579 same clock tick as a message's arrival, so that its received time was the
1580 same as the "first fail" time on the retry record, and that message
1581 remained on the queue past the ultimate address timeout, every queue runner
1582 would try a delivery (because it was past the ultimate address timeout) but
1583 after another failure, the ultimate address timeout, which should have then
1584 bounced the address, did not kick in. This was a "< instead of <=" error;
1585 in most cases the first failure would have been in the next clock tick
1586 after the received time, and all would be well.
1588 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1589 being recognized when the domain list was tested by the match_domain
1590 condition in an expansion string.
1592 16. Added the ${str2b64: operator.
1594 17. Exim was always calling setrlimit() to set a large limit for the number of
1595 processes, without checking whether the existing limit was already
1596 adequate. (It did check for the limit on file descriptors.) Furthermore,
1597 errors from getrlimit() and setrlimit() were being ignored. Now they are
1598 logged to the main and panic logs, but Exim does carry on, to try to do its
1599 job under whatever limits there are.
1601 18. Imported PCRE 5.0.
1603 19. Trivial typo in log message " temporarily refused connection" (the leading
1606 20. If the log selector return_path_on_delivery was set and an address was
1607 redirected to /dev/null, the delivery process crashed because it assumed
1608 that a return path would always be set for a "successful" delivery. In this
1609 case, the whole delivery is bypassed as an optimization, and therefore no
1612 21. Internal re-arrangement: the function for sending a challenge and reading
1613 a response while authentication was assuming a zero-terminated challenge
1614 string. It's now changed to take a pointer and a length, to allow for
1615 binary data in such strings.
1617 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1619 23. Exim was not respecting finduser_retries when seeking the login of the
1620 uid under which it was called; it was always trying 10 times. (The default
1621 setting of finduser_retries is zero.) Also, it was sleeping after the final
1622 failure, which is pointless.
1624 24. Implemented tls_on_connect_ports.
1626 25. Implemented acl_smtp_predata.
1628 26. If the domain in control=submission is set empty, Exim assumes that the
1629 authenticated id is a complete email address when it generates From: or
1630 Sender: header lines.
1632 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1633 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1634 chown and chgrp in /bin and hostname in /usr/bin.
1636 28. Exim was keeping the "process log" file open after each use, just as it
1637 does for the main log. This opens the possibility of it remaining open for
1638 long periods when the USR1 signal hits a daemon. Occasional processlog
1639 errors were reported, that could have been caused by this. Anyway, it seems
1640 much more sensible not to leave this file open at all, so that is what now
1643 29. The long-running daemon process does not normally write to the log once it
1644 has entered its main loop, and it closes the log before doing so. This is
1645 so that log files can straightforwardly be renamed and moved. However,
1646 there are a couple of unusual error situations where the daemon does write
1647 log entries, and I had neglected to close the log afterwards.
1649 30. The text of an SMTP error response that was received during a remote
1650 delivery was being truncated at 512 bytes. This is too short for some of
1651 the long messages that one sometimes sees. I've increased the limit to
1654 31. It is now possible to make retry rules that apply only when a message has a
1655 specific sender, in particular, an empty sender.
1657 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1658 it possible to be selective about when SMTP synchronization is enforced.
1660 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1662 32. Implemented hosts_connection_nolog.
1664 33. Added an ACL for QUIT.
1666 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1669 35. Added mailbox_size and mailbox_filecount to appendfile.
1671 36. Added control = no_multiline_responses to ACLs.
1673 37. There was a bug in the logic of the code that waits for the clock to tick
1674 in the case where the clock went backwards by a substantial amount such
1675 that the microsecond fraction of "now" was more than the microsecond
1676 fraction of "then" (but the whole seconds number was less).
1678 38. Added support for the libradius Radius client library this is found on
1679 FreeBSD (previously only the radiusclient library was supported).
1685 1. When certain lookups returned multiple values in the form name=value, the
1686 quoting of the values was not always being done properly. Specifically:
1687 (a) If the value started with a double quote, but contained no whitespace,
1689 (b) If the value contained whitespace other than a space character (i.e.
1690 tabs or newlines or carriage returns) it was not quoted.
1691 This fix has been applied to the mysql and pgsql lookups by writing a
1692 separate quoting function and calling it from the lookup code. The fix
1693 should probably also be applied to nisplus, ibase and oracle lookups, but
1694 since I cannot test any of those, I have not disturbed their existing code.
1696 2. A hit in the callout cache for a specific address caused a log line with no
1697 reason for rejecting RCPT. Now it says "Previous (cached) callout
1698 verification failure".
1700 3. There was an off-by-one bug in the queryprogram router. An over-long
1701 return line was truncated at 256 instead of 255 characters, thereby
1702 overflowing its buffer with the terminating zero. As well as fixing this, I
1703 have increased the buffer size to 1024 (and made a note to document this).
1705 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1706 when Exim is waiting for an SMTP response from a remote server, Exim
1707 restarts its select() call on the socket, thereby resetting its timeout.
1708 This is not a problem when such interrupts are rare. Somebody set up a cron
1709 job to run exiwhat every 2 minutes, which is less than the normal select()
1710 timeout (5 or 10 minutes). This meant that the select() timeout never
1711 kicked in because it was always reset. I have fixed this by comparing the
1712 time when an interrupt arrives with the time at the start of the first call
1713 to select(). If more time than the timeout has elapsed, the interrupt is
1714 treated as a timeout.
1716 5. Some internal re-factoring in preparation for the addition of Sieve
1717 extensions (by MH). In particular, the "personal" test is moved to a
1718 separate function, and given an option for scanning Cc: and Bcc: (which is
1719 not set for Exim filters).
1721 6. When Exim created an email address using the login of the caller as the
1722 local part (e.g. when creating a From: or Sender: header line), it was not
1723 quoting the local part when it contained special characters such as @.
1725 7. Installed new OpenBSD configuration files.
1727 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1728 try to make them clearer.
1730 9. Callout options, other than the timeout value, were being ignored when
1731 verifying sender addresses in header lines. For example, when using
1733 verify = header_sender/callout=no_cache
1735 the cache was (incorrectly) being used.
1737 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1738 only the Cygwin environment.
1740 11. When return_path_on_delivery was set as a log selector, if different remote
1741 addresses in the same message used different return paths and parallel
1742 remote delivery occurred, the wrong values would sometimes be logged.
1743 (Whenever a remote delivery process finished, the return path value from
1744 the most recently started remote delivery process was logged.)
1746 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1747 lines when AUTH and/or TLS are in use. This is the "received protocol"
1748 field. Exim used to use "asmtp" for authenticated SMTP, without any
1749 indication (in the protocol name) for TLS use. Now it follows the RFC and
1750 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1751 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1752 names appear in log lines as well as in Received: header lines.
1754 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1755 extensions, and comparison tests, and to fix some bugs.
1757 14. Changes to the "personal" filter test:
1759 (1) The test was buggy in that it was just doing the equivalent of
1760 "contains" tests on header lines. For example, if a user's address was
1761 anne@some.where, the "personal" test would incorrectly be true for
1763 To: susanne@some.where
1765 This test is now done by extracting each address from the header in turn,
1766 and checking the entire address. Other tests that are part of "personal"
1767 are now done using regular expressions (for example, to check local parts
1768 of addresses in From: header lines).
1770 (2) The list of non-personal local parts in From: addresses has been
1771 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1772 taken from the Sieve specification recommendations.
1774 (3) If the message contains any header line starting with "List-" it is
1775 treated as non-personal.
1777 (4) The test for "circular" in the Subject: header line has been removed
1778 because it now seems ill-conceived.
1780 15. Minor typos in src/EDITME comments corrected.
1782 16. Installed latest exipick from John Jetmore.
1784 17. If headers_add on a router specified a text string that was too long for
1785 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1786 of string_sprintf() is now avoided.
1788 18. $message_body_size was not set (it was always zero) when running the DATA
1789 ACL and the local_scan() function.
1791 19. For the "mail" command in an Exim filter, no default was being set for
1792 the once_repeat time, causing a random time value to be used if "once" was
1793 specified. (If the value happened to be <= 0, no repeat happened.) The
1794 default is now 0s, meaning "never repeat". The "vacation" command was OK
1795 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1796 (I found it when inspecting the code).
1798 20. There is now an overall timeout for performing a callout verification. It
1799 defaults to 4 times the callout timeout, which applies to individual SMTP
1800 commands during the callout. The overall timeout applies when there is more
1801 than one host that can be tried. The timeout is checked before trying the
1802 next host. This prevents very long delays if there are a large number of
1803 hosts and all are timing out (e.g. when the network connections are timing
1804 out). The value of the overall timeout can be changed by specifying an
1805 additional sub-option for "callout", called "maxwait". For example:
1807 verify = sender/callout=5s,maxwait=20s
1809 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1810 to the end before writing, but this should make it even safer).
1812 22. Exim was forgetting that it had advertised PIPELINING for the second and
1813 subsequent messages on an SMTP connection. It was also not resetting its
1814 memory on STARTTLS and an internal HELO.
1816 23. When Exim logs an SMTP synchronization error within a session, it now
1817 records whether PIPELINING has been advertised or not.
1819 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1820 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1821 rather than long int.
1823 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1825 26. Added the never_mail option to autoreply.
1831 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1832 crash if the getsockname() call failed; this can happen if a connection is
1833 closed very soon after it is established. The problem was simply in the
1834 order in which certain operations were done, causing Exim to try to write
1835 to the SMTP stream before it had set up the file descriptor. The bug has
1836 been fixed by making things happen in the correct order.
1842 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1843 before the connection was closed, thus losing the rejection response.
1845 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1846 some early Solaris releases, but causes trouble in current releases where
1847 socklen_t is defined.
1849 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1852 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1855 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1856 that releasing the top of it at the end releases what was used for sub-
1857 expansions (unless the block got too big). However, discard this block if
1858 the first thing is a variable or header, so that we can use its block when
1859 it is dynamic (useful for very large $message_headers, for example).
1861 6. Lookups now cache *every* query, not just the most recent. A new, separate
1862 store pool is used for this. It can be recovered when all lookup caches are
1863 flushed. Lookups now release memory at the end of their result strings.
1864 This has involved some general refactoring of the lookup sources.
1866 7. Some code has been added to the store_xxx() functions to reduce the amount
1867 of flapping under certain conditions.
1869 8. log_incoming_interface used to affect only the <= reception log lines. Now
1870 it causes the local interface and port to be added to several more SMTP log
1871 lines, for example "SMTP connection from", and rejection lines.
1873 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1875 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1877 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1878 could be overwritten at the end of the current message (or the start of a
1879 new message if it was set in a HELO ACL). The value is now preserved for
1880 the duration of the SMTP connection.
1882 12. If a transport had a headers_rewrite setting, and a matching header line
1883 contained an unqualified address, that address was qualified, even if it
1884 did not match any rewriting rules. The underlying bug was that the values
1885 of the flags that permit the existence of unqualified sender and recipient
1886 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1887 non-local messages, and by -bnq for local messages) were not being
1888 preserved with the message after it was received.
1890 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1891 "next input=" as part of the text comprising the host identity instead of
1892 the correct text. The code was using the same buffer for two different
1893 strings. However, depending on which order the printing function evaluated
1894 its arguments, the bug did not always show up. Under Linux, for example, my
1895 test suite worked just fine.
1897 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1898 doesn't work with some older versions of Perl. It has been changed to "my",
1899 which in any case is probably the better facility to use.
1901 15. A really picky compiler found some instances of statements for creating
1902 error messages that either had too many or two few arguments for the format
1905 16. The size of the buffer for calls to the DNS resolver has been increased
1906 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1907 for addresses that have a lot of PTR records. This alleviates a problem; it
1908 does not fully solve it.
1910 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1911 buffer now truncates the list and logs the incident, which is the same
1912 action as happens when Exim is looking up a host name and its aliases.
1913 Previously in this situation something unpredictable would happen;
1914 sometimes it was "internal error: store_reset failed".
1916 18. If a server dropped the connection unexpectedly when an Exim client was
1917 using GnuTLS and trying to read a response, the client delivery process
1918 crashed while trying to generate an error log message.
1920 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1921 string, for example:
1923 warn message = H1: something\nH2: something
1925 the text was added as a single header line from Exim's point of view
1926 though it ended up OK in the delivered message. However, searching for the
1927 second and subsequent header lines using $h_h2: did not work. This has been
1928 fixed. Similarly, if a system filter added multiple headers in this way,
1929 the routers could not see them.
1931 20. Expanded the error message when iplsearch is called with an invalid key to
1932 suggest using net-iplsearch in a host list.
1934 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1935 ACLs are no longer actually imposed (and a message to that effect is
1938 22. If a "gecos" field in a passwd entry contained escaped characters, in
1939 particular, if it contained a \" sequence, Exim got it wrong when building
1940 a From: or a Sender: header from that name. A second bug also caused
1941 incorrect handling when an unquoted " was present following a character
1942 that needed quoting.
1944 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1945 was not being matched caselessly.
1947 24. Arranged for all hyphens in the exim.8 source to be escaped with
1950 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1951 properly. Recipient callout cache records were still being keyed to include
1952 the sender, even when use_sender was set false. This led to far more
1953 callouts that were necessary. The sender is no longer included in the key
1954 when use_sender is false.
1956 26. Added "control = submission" modifier to ACLs.
1958 27. Added the ${base62d: operator to decode base 62 numbers.
1960 28. dnsdb lookups can now access SRV records.
1962 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1963 the configuration file.
1965 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1966 (-v) mode. This makes the output for a verbose queue run more intelligible.
1968 31. Added a use_postmaster feature to recipient callouts.
1970 32. Added the $body_zerocount variable, containing the number of binary zero
1971 bytes in the message body.
1973 33. The time of last modification of the "new" subdirectory is now used as the
1974 "mailbox time last read" when there is a quota error for a maildir
1977 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1979 35. Added +ignore_unknown as a special item in host lists.
1981 36. Code for decoding IPv6 addresses in host lists is now included, even if
1982 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1983 address was recognized as an IP address, but was then not correctly decoded
1984 into binary, causing unexpected and incorrect effects when compared with
1991 1. Very minor rewording of debugging text in manualroute to say "list of
1992 hosts" instead of "hostlist".
1994 2. If verify=header_syntax was set, and a header line with an unqualified
1995 address (no domain) and a large number of spaces between the end of the
1996 name and the colon was received, the reception process suffered a buffer
1997 overflow, and (when I tested it) crashed. This was caused by some obsolete
1998 code that should have been removed. The fix is to remove it!
2000 3. When running in the test harness, delay a bit after writing a bounce
2001 message to get a bit more predictability in the log output.
2003 4. Added a call to search_tidyup() just before forking a reception process. In
2004 theory, someone could use a lookup in the expansion of smtp_accept_max_
2005 per_host which, without the tidyup, could leave open a database connection.
2007 5. Added the variables $recipient_data and $sender_data which get set from a
2008 lookup success in an ACL "recipients" or "senders" condition, or a router
2009 "senders" option, similar to $domain_data and $local_part_data.
2011 6. Moved the writing of debug_print from before to after the "senders" test
2014 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2015 problems for message scanning, either using a data ACL, or using
2016 local_scan() because the Received: header was not generated till after they
2017 were called (in order to set the time as the time of reception completion).
2018 I have revised the way this works. The header is now generated after the
2019 body is received, but before the ACL or local_scan() are called. After they
2020 are run, the timestamp in the header is updated.
2026 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2027 before starting a queue runner without re-exec. This happened only when
2028 deliver_drop_privilege was set or when the Exim user was set to root. The
2029 effect of the bug was that timeouts during subsequent deliveries caused
2030 crashes instead of being properly handled. The handler is now left at its
2031 default (and expected) setting.
2033 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2034 message, again when deliver_drop_privilege is set or Exim is run as root.
2035 The bug described in (1) was not present in this case, but the tidying up
2036 of the other signals was missing. I have made the two cases consistent.
2038 3. The ignore_target_hosts setting on a manualroute router was being ignored
2039 for hosts that were looked up using the /MX notation.
2041 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2044 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2045 operated on the sender address. After changing the $sender_address to <>
2046 for the sender address verify, Exim was re-instated it as the original
2047 (before rewriting) address, but remembering that it had rewritten it, so it
2048 wasn't rewriting it again. This bug also had the effect of breaking the
2049 sender address verification caching when the sender address was rewritten.
2051 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2052 This has been changed so that if the ip literal address matches
2053 ignore_target_hosts, the router declines.
2055 7. Added expansion conditions match_domain, match_address, and match_local_
2056 part (NOT match_host).
2058 8. The placeholder for the Received: header didn't have a length field set.
2060 9. Added code to Exim itself and to exim_lock to test for a specific race
2061 condition that could lead to file corruption when using MBX delivery. The
2062 issue is with the lockfile that is created in /tmp. If this file is removed
2063 after a process has opened it but before that process has acquired a lock,
2064 there is the potential for a second process to recreate the file and also
2065 acquire a lock. This could lead to two Exim processes writing to the file
2066 at the same time. The added code performs the same test as UW imapd; it
2067 checks after acquiring the lock that its file descriptor still refers to
2068 the same named file.
2070 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2071 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2074 11. Added the smtp_active_hostname option. If used, this will typically be made
2075 to depend on the incoming interface address. Because $interface_address is
2076 not set up until the daemon has forked a reception process, error responses
2077 that can happen earlier (such as "too many connections") no longer contain
2080 12. If an expansion in a condition on a "warn" statement fails because a lookup
2081 defers, the "warn" statement is abandoned, and the next ACL statement is
2082 processed. Previously this caused the whole ACL to be aborted.
2084 13. Added the iplsearch lookup type.
2086 14. Added ident_timeout as a log selector.
2088 15. Added tls_certificate_verified as a log selector.
2090 16. Added a global option tls_require_ciphers (compare the smtp transport
2091 option of the same name). This controls incoming TLS connections.
2093 17. I finally figured out how to make tls_require_ciphers do a similar thing
2094 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2095 before starting the TLS session.
2097 18. Tabs are now shown as \t in -bP output.
2099 19. If the log selector return_path_on_delivery was set, Exim crashed when
2100 bouncing a message because it had too many Received: header lines.
2102 20. If two routers both had headers_remove settings, and the first one included
2103 a superfluous trailing colon, the final name in the first list and the
2104 first name in the second list were incorrectly joined into one item (with a
2105 colon in the middle).
2111 1. Added -C and -D options to the exinext utility, mainly to make it easier
2112 to include in the automated testing, but these could be helpful when
2113 multiple configurations are in use.
2115 2. The exinext utility was not formatting the output nicely when there was
2116 an alternate port involved in the retry record key, nor when there was a
2117 message id as well (for retries that were specific to a specific message
2118 and a specific host). It was also confused by IPv6 addresses, because of
2119 the additional colons they contain. I have fixed the IPv4 problem, and
2120 patched it up to do a reasonable job for IPv6.
2122 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2123 delivery, the log line now contains "pipelined" if PIPELINING was used.
2125 4. An SMTP transport process used to panic and die if the bind() call to set
2126 an explicit outgoing interface failed. This has been changed; it is now
2127 treated in the same way as a connect() failure.
2129 5. A reference to $sender_host_name in the part of a conditional expansion
2130 that was being skipped was still causing a DNS lookup. This no longer
2133 6. The def: expansion condition was not recognizing references to header lines
2134 that used bh_ and bheader_.
2136 7. Added the _cache feature to named lists.
2138 8. The code for checking quota_filecount in the appendfile transport was
2139 allowing one more file than it should have been.
2141 9. For compatibility with Sendmail, the command line option
2149 and sets the incoming protocol and host name (for trusted callers). The
2150 host name and its colon can be omitted when only the protocol is to be set.
2151 Note the Exim already has two private options, -pd and -ps, that refer to
2152 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2153 "s", but I don't think that's a major issue.
2155 10. A number of refactoring changes to the code, none of which should affect
2158 (a) The number of logging options was getting close to filling up the
2159 32-bit word that was used as a bit map. I have split them into two classes:
2160 those that are passed in the argument to log_write(), and those that are
2161 only ever tested independently outside of that function. These are now in
2162 separate 32-bit words, so there is plenty of room for expansion again.
2163 There is no change in the user interface or the logging behaviour.
2165 (b) When building, for example, log lines, the code previously used a
2166 macro that called string_cat() twice, in order to add two strings. This is
2167 not really sufficiently general. Furthermore, there was one instance where
2168 it was actually wrong because one of the argument was used twice, and in
2169 one call a function was used. (As it happened, calling the function twice
2170 did not affect the overall behaviour.) The macro has been replaced by a
2171 function that can join an arbitrary number of extra strings onto a growing
2174 (c) The code for expansion conditions now uses a table and a binary chop
2175 instead of a serial search (which was left over from when there were very
2176 few conditions). Also, it now recognizes conditions like "pam" even when
2177 the relevant support is not compiled in: a suitably worded error message is
2178 given if an attempt is made to use such a condition.
2180 11. Added ${time_interval:xxxxx}.
2182 12. A bug was causing one of the ddress fields not to be passed back correctly
2183 from remote delivery subprocesses. The field in question was not being
2184 subsequently used, so this caused to problems in practice.
2186 13. Added new log selectors queue_time and deliver_time.
2188 14. Might have fixed a bug in maildirsizefile handling that threw up
2189 "unexpected character" debug warnings, and recalculated the data
2190 unnecessarily. In any case, I expanded the warning message to give more
2193 15. Added the message "Restricted characters in address" to the statements in
2194 the default ACL that block characters like @ and % in local parts.
2196 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2197 Three changes have been made:
2199 (a) There was a serious bug; a negative response to MAIL caused the whole
2200 recipient domain to be cached as invalid, thereby blocking all messages
2201 to all local parts at the same domain, from all senders. This bug has
2202 been fixed. The domain is no longer cached after a negative response to
2203 MAIL if the sender used is not empty.
2205 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2208 (c) A new callout option, "use_sender" has been added for people who want
2209 the modified behaviour.
2215 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2218 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2219 indeed breaks things for older releases.
2221 3. Added additional logging to the case where there is a problem reading data
2222 from a filter that is running in a subprocess using a pipe, in order to
2223 try to track down a specific problem.
2225 4. Testing facility fudge: when running in the test harness and attempting
2226 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2227 getting "No route to host". Convert this to a timeout.
2229 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2232 6. Some OS don't have socklen_t but use size_t instead. This affects the
2233 fifth argument of getsockopt() amongst other things. This is now
2234 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2235 can be set for individual OS. I have set it for SunOS5, OSF1, and
2236 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2237 some earlier ones do not.
2239 7. Change 4.30/15 was not doing the test caselessly.
2241 8. The standard form for an IPv6 address literal was being rejected by address
2242 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2243 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2244 this, as well as the form without the "IPv6" on the front (but only when
2245 address literals are enabled, of course).
2247 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2249 10. Exim crashed if a message with an empty sender address specified by -f
2250 encountered a router with an errors_to setting. This could be provoked only
2251 by a command such as
2255 where an empty string was supplied; "<>" did not hit this bug.
2257 11. Installed PCRE release 4.5.
2259 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2260 remained set. It is now erased.
2262 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2263 times from message ids (which are base 36 rather than the normal 62).
2265 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2266 were being counted as actual protocol errors, and logged if the log
2267 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2268 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2269 and DATA following a set of rejected RCPTs do not count as protocol errors.
2270 In other words, Exim assumes they were pipelined, though this may not
2271 actually be the case. Of course, in all cases the client gets an
2272 appropriate error code.
2274 15. If a lookup fails in an ACL condition, a message about the failure may
2275 be available; it is used if testing the ACL cannot continue, because most
2276 such messages specify what the cause of the deferral is. However, some
2277 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2278 that caused an old message to be retained and used if a later statement
2279 caused a defer, replacing the real cause of the deferral.
2281 16. If an IP address had so many PTR records that the DNS lookup buffer
2282 was not large enough to hold them, Exim could crash while trying to process
2283 the truncated data. It now detects and logs this case.
2285 17. Further to 4.21/58, another change has been made: if (and only if) the
2286 first line of a message (the first header line) ends with CRLF, a bare LF
2287 in a subsequent header line has a space inserted after it, so as not to
2288 terminate the header.
2290 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2291 unnecessarily, used atoi() instead of strtol(), and didn't check the
2292 termination when getting file sizes from file names by regex.
2294 19. Completely re-implemented the support for maildirsize files, in the light
2295 of a number of problems with the previous contributed implementation
2296 (4.30/29). In particular:
2298 . If the quota is zero, the maildirsize file is maintained, but no quota is
2301 . If the maildir directory does not exist, it is created before any attempt
2302 to write a maildirsize file.
2304 . The quota value in the file is just a cache; if the quota is changed in
2305 the transport, the new value overrides.
2307 . A regular expression is available for excluding directories from the
2310 20. The autoreply transport checks the characters in options that define the
2311 message's headers; it allows continued headers, but it was checking with
2312 isspace() after an embedded newline instead of explicitly looking for a
2315 21. If all the "regular" hosts to which an address was routed had passed their
2316 expiry times, and had not reached their retry times, the address was
2317 bounced, even if fallback hosts were defined. Now Exim should go on to try
2320 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2321 equivalent code in the SMTP transport. Some hosts send humungous responses
2322 to HELO/EHLO, more than 1024 it seems.
2324 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2325 gives compiler warnings in some environments. I've now done it "properly",
2328 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2329 (because of problems with the built-in one) was declared to return uschar *
2330 instead of char *, causing compiler failure.
2332 25. Fixed a file descriptor leak when processing alias/forward files.
2334 26. Fixed a minor format string issue in dbfn.c.
2336 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2338 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2339 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2342 29. When a sender address is verified, it is cached, to save repeating the test
2343 when there is more than one recipient in a message. However, when the
2344 verification involves a callout, it is possible for different callout
2345 options to be set for different recipients. It is too complicated to keep
2346 track of this in the cache, so now Exim always runs a verification when a
2347 callout is required, relying on the callout cache for the optimization.
2348 The overhead is duplication of the address routing, but this should not be
2351 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2352 to be verified with callout=postmaster, and the main callout worked but the
2353 postmaster check failed, the verification correctly failed. However, if a
2354 subsequent RCPT command asked for sender verification *without* the
2355 postmaster check, incorrect caching caused this verification also to fail,
2358 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2359 it was not caching the DNS options (qualify_single, search_parents) that
2360 were used when the lookup failed. A subsequent lookup with different
2361 options therefore always gave the same answer, though there were cases
2362 where it should not have. (Example: a "domains = !$mx_any" option on a
2363 dnslookup router: the "domains" option is always processed without any
2364 widening, but the router might have qualify_single set.) Now Exim uses the
2365 cached value only when the same options are set.
2367 32. Added John Jetmore's "exipick" utility to the distribution.
2369 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2370 than a timeout (e.g. a certificate is required, and is not provided), an
2371 Exim server now closes the connection immediately. Previously it waited for
2372 the client to close - but if the client is SSL, it seems that they each
2373 wait for each other, leading to a delay before one of them times out.
2375 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2376 maintained 0.8.x compatibility because I don't think many are using it, and
2377 it is clearly obsolete.
2379 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2382 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2383 client certificate was expired. A simple patch fixes this, though I don't
2384 understand the full logic of why the verify callback is called multiple
2387 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2388 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2389 which causes problems with some clients (such as the Certicom SSL Plus
2390 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2391 disables the coutermeasure allowing Eudora to connect."
2393 38. Exim was not checking that a write() to a log file succeeded. This could
2394 lead to Bad Things if a log got too big, in particular if it hit a file
2395 size limit. Exim now panics and dies if it cannot write to a log file, just
2396 as it does if it cannot open a log file.
2398 39. Modified OS/Makefile-Linux so that it now contains
2400 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2402 The two -D definitions ensure that Exim is compiled with large file
2403 support, which makes it possible to handle log files that are bigger than
2406 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2407 instance) a domain was checked against a named list that involved a lookup,
2408 causing $domain_data to be set, then another domain was checked against the
2409 same list, then the first domain was re-checked, the value of $domain_data
2410 after the final check could be wrong. In particular, if the second check
2411 failed, it could be set empty. This bug probably also applied to
2414 41. The strip_trailing_dot option was not being applied to the address given
2415 with the -f command-line option.
2417 42. The code for reading a message's header from the spool was incrementing
2418 $received_count, but never initializing it. This meant that the value was
2419 incorrect (doubled) while delivering a message in the same process in which
2420 it was received. In the most common configuration of Exim, this never
2421 happens - a fresh exec is done - but it can happen when
2422 deliver_drop_privilege is set.
2424 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2425 it now includes up to 150 characters of the unexpected data in the log
2428 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2429 and building data strings. The size of both of these buffers was 10 000
2430 bytes - far larger than anybody would *ever* want, thought I. Needless to
2431 say, somebody hit the limit. I have increased the maximum line length to
2432 20 000 and the maximum data length of concatenated lines to 100 000. I have
2433 also fixed two bugs, because there was no checking on these buffers. Tsk,
2434 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2435 buffer is too small.
2437 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2438 lsearch lookups. Now it does.
2440 46. When parsing a route_list item in a manualroute router, a fixed-length
2441 buffer was used for the list of hosts. I made this 1024 bytes long,
2442 thinking that nobody would ever have a list of hosts that long. Wrong.
2443 Somebody had a whole pile of complicated expansion conditions, and the
2444 string was silently truncated, leading to an expansion error. It turns out
2445 that it is easier to change to an unlimited length (owing to other changes
2446 that have happened since this code was originally written) than to build
2447 structure for giving a limitation error. The length of the item that
2448 expands into the list of hosts is now unlimited.
2450 47. The lsearch lookup could not handle data where the length of text line was
2451 more than 4095 characters. Such lines were truncated, leading to shortened
2452 data being returned. It should now handle lines of any length.
2454 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2455 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2458 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2459 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2460 no longer happens in comment lines. A list of replacements is now placed
2461 at the head of all of the source files, except those whose only change is
2462 to replace PERL_COMMAND in the very first #! line.
2464 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2465 messages on the queue, with a bottom-up merge sort, using code contributed
2466 by Michael Haardt. This should make operations like -bp somewhat faster on
2467 large queues. It won't affect queue runners, except when queue_run_in_order
2470 51. Installed eximstats 1.31 in the distribution.
2472 52. Added support for SRV lookups to the dnslookup router.
2474 53. If an ACL referred to $message_body or $message_body_end, the value was not
2475 reset for any messages that followed in the same SMTP session.
2477 54. The store-handling optimization for building very long strings was not
2478 differentiating between the different store pools. I don't think this
2479 actually made any difference in practice, but I've tidied it.
2481 55. While running the routers to verify a sender address, $sender_address
2482 was still set to the sender address. This is wrong, because when routing to
2483 send a bounce to the sender, it would be empty. Therefore, I have changed
2484 it so that, while verifying a sender address, $sender_address is set to <>.
2485 (There is no change to what happens when verifying a recipient address.)
2487 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2488 target A or AAAA records (if not already returned) without resetting the
2489 qualify_single or search_parents options of the DNS resolver. These are
2490 inappropriate in this case because the targets of MX and SRV records must
2491 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2492 target that, when qualified, matched something in the local domain. These
2493 two options are now turned off when doing these lookups.
2495 57. It seems that at least some releases of Reiserfs (which does not have the
2496 concept of a fixed number of inodes) returns zero and not -1 for the
2497 number of available inodes. This interacted badly with check_spool_inodes,
2498 which assumed that -1 was the "no such thing" setting. What I have done is
2499 to check that the total number of inodes is greater than zero before doing
2500 the test of how many are available.
2502 58. When a "warn" ACL statement has a log_message modifier, the message is
2503 remembered, and not repeated. This is to avoid a lot of repetition when a
2504 message has many recipients that cause the same warning to be written.
2505 Howewer, Exim was preserving the list of already written lines for an
2506 entire SMTP session, which doesn't seem right. The memory is now reset if a
2507 new message is started.
2509 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2510 debugging output unless log_rewrite was also set.
2512 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2513 of (int)(handle) when we know that handle contains (void *)(-1).
2515 61. The Exim daemon panic-logs an error return when it closes the incoming
2516 connection. However "connection reset by peer" seems to be common, and
2517 isn't really an error worthy of noting specially, so that particular error
2520 62. When Exim is trying to find all the local interfaces, it used to panic and
2521 die if the ioctl to get the interface flags failed. However, it seems that
2522 on at least one OS (Solaris 9) it is possible to have an interface that is
2523 included in the list of interfaces, but for which you get a failure error
2524 for this call. This happens when the interface is not "plumbed" into a
2525 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2526 failure of the "get flags" call assumes that the interface is down.
2528 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2529 makes life easier for people who are doing arithmetic on fields extracted
2530 from dates, where you often get leading zeros that should not be
2531 interpreted as octal.
2533 64. Added qualify_domain to the redirect router, to override the global
2536 65. If a pathologically long header line contained very many addresses (the
2537 report of this problem mentioned 10 000) and each of them was rewritten,
2538 Exim could use up a very large amount of memory. (It kept on making new
2539 copies of the header line as it rewrote, and never released the old ones.)
2540 At the expense of a bit more processing, the header rewriting function has
2541 been changed so that it no longer eats memory in this way.
2543 66. The generation of the Received: header has been moved from the time that a
2544 message starts to be received, to the time that it finishes. The timestamp
2545 in the Received: header should now be very close to that of the <= log
2546 line. There are two side-effects of this change:
2548 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2549 logged header lines no longer include the local Received: line, because
2550 it has not yet been created. The same applies to a copy of the message
2551 that is returned to a non-SMTP sender when a message is rejected.
2553 (b) When a filter file is tested using -bf, no additional Received: header
2554 is added to the test message. After some thought, I decided that this
2557 This change does not affect the value of $received_for. It is still set
2558 after address rewriting, but before local_scan() is called.
2560 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2562 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2563 gave an unhelpful panic error message, and a defer error. I have managed to
2564 change this behaviour so that it now rejects any supplied certificate,
2565 which seems right, as the list of acceptable certificates is empty.
2567 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2568 gave an unhelpful defer error. I have not managed to make this reject any
2569 supplied certificates, but the error message it gives is "no certificate
2570 supplied", which is not helpful.
2572 70. exigrep's output now also includes lines that are not associated with any
2573 message, but which match the given pattern. Implemented by a patch from
2574 Martin Sluka, which also tidied up the Perl a bit.
2576 71. Recipient callout verification, like sender verification, was using <> in
2577 the MAIL FROM command. This isn't really the right thing, since the actual
2578 sender may affect whether the remote host accepts the recipient or not. I
2579 have changed it to use the actual sender in the callout; this means that
2580 the cache record is now keyed on a recipient/sender pair, not just the
2581 recipient address. There doesn't seem to be a real danger of callout loops,
2582 since a callout by the remote host to check the sender would use <>.
2583 [SEE ABOVE: changed after hitting problems.]
2585 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2586 temporary errors. However, in the case of such a code being given after
2587 the end of a data transmission (i.e. after ".") Exim was failing to write
2588 a retry record for the message. (Yes, there was some broken host that was
2589 actually sending 8xx at this point.)
2591 73. An unknown lookup type in a host list could cause Exim to panic-die when
2592 the list was checked. (An example that provoked this was putting <; in the
2593 middle of a list instead of at the start.) If this happened during a DATA
2594 ACL check, a -D file could be left lying around. This kind of configuration
2595 error no longer causes Exim to die; instead it causes a defer errror. The
2596 incident is still logged to the main and panic logs.
2598 74. Buglet left over from Exim 3 conversion. The message "too many messages
2599 in one connection" was written to the rejectlog but not the mainlog, except
2600 when address rewriting (yes!) was being logged.
2602 75. Added write_rejectlog option.
2604 76. When a system filter was run not as root (that is, when system_filter_user
2605 was set), the values of the $n variables were not being returned to the
2606 main process; thus, they were not subsequently available in the $sn
2609 77. Added +return_path_on_delivery log selector.
2611 78. A connection timeout was being treated differently from recipients deferred
2612 when testing hosts_max_try with a message that was older than the host's
2613 retry timeout. (The host should not be counted, thus allowing all hosts to
2614 be tried at least once before bouncing.) This may have been the cause of an
2615 occasionally reported bug whereby a message would remain on the queue
2616 longer than the retry timeout, but would be bounced if a delivery was
2617 forced. I say "may" because I never totally pinned down the problem;
2618 setting up timeout/retry tests is difficult. See also the next item.
2620 79. The ultimate address timeout was not being applied to errors that involved
2621 a combination of host plus message (for example, a timeout on a MAIL
2622 command). When an address resolved to a number of possible hosts, and they
2623 were not all tried for each delivery (e.g. because of hosts_max_try), a
2624 message could remain on the queue longer than the retry timeout.
2626 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2629 81. Fixed an obscure SMTP outgoing bug which required at least the following
2630 conditions: (a) there was another message waiting for the same server;
2631 (b) the server returned 5xx to all RCPT commands in the first message so
2632 that the message was not completed; (c) the server dropped the connection
2633 or gave a negative response to the RSET that Exim sends to abort the
2634 transaction. The observed case was a dropped connection after DATA that had
2635 been sent in pipelining mode. That is, the server had advertised PIPELINING
2636 but was not implementing it correctly. The effect of the bug was incorrect
2637 behaviour, such as trying another host, and this could lead to a crash.
2643 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2644 and daemon.c were passed as pointers to ints; they should have been
2645 pointers to socklen_t variables (which are typically unsigned ints).
2647 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2650 3. Fixed a really odd bug that affected only the testing scheme; patching a
2651 certain fixed string in the binary changed the value of another string that
2652 happened to be identical to the end of the original first string.
2654 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2655 name", it returns that address as the IP address. On some operating
2656 systems (e.g. Solaris), it also passes back the IP address string as the
2657 "host name". However, on others (e.g. Linux), it passes back an empty
2658 string. Exim wasn't checking for this, and was changing the host name to an
2659 empty string, assuming it had been canonicized.
2661 5. Although rare, it is permitted to have more than one PTR record for a given
2662 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2663 all the names associated with an address, because they do in Solaris.
2664 However, it seems that they do not in Linux for data that comes from the
2665 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2666 I found this out when I moved to a new Linux workstation and tried to run
2667 the Exim test suite.
2669 To get round this problem I have changed the code so that it now does its
2670 own call to the DNS to look up PTR records when searching for a host name.
2671 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2672 addresses that are only in /etc/hosts are still found.
2674 This behaviour is, however, controlled by an option called host_lookup_
2675 order, which defaults to "bydns:byaddr". If people want to use the other
2676 order, or indeed, just use one or the other means of lookup, they can
2677 specify it in this variable.
2679 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2680 some operating systems, this comes back from gethostbyaddr() as an empty
2681 string, and this is what Exim used to test for. However, it seems that in
2682 other systems, "." is yielded. Exim now tests for this case too.
2684 7. The values of check_spool_space and check_log_space are now held internally
2685 as a number of kilobytes instead of an absolute number of bytes. If a
2686 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2687 kilobyte. This means that much larger values can be stored.
2689 8. Exim monitor: an attempt to get the action menu when not actually pointing
2690 at a message produces an empty menu entitled "No message selected". This
2691 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2692 no entries in it ("Shell widget menu has zero width and/or height"). So I
2693 have added a single, blank menu entry in this case.
2695 9. Added ${quote_local_part.
2697 10. MIME decoding is now applied to the contents of Subject: header lines when
2700 11. Now that a reference to $sender_host_address automatically causes a reverse
2701 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2702 host lookup before query-style lookups in lists that might use this
2703 variable. This has therefore been abolished, and the "net-" prefix is no
2704 longer necessary for query-style lookups.
2706 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2707 have been a typo for LFLAGS, so it has been changed.
2709 13. The install script calls Exim with "-C /dev/null" in order to find the
2710 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2711 to be output. Howeve, since Exim outputs its version number before the
2712 error, it didn't break the script. It just looked ugly. I fixed this by
2713 always allowing "-C /dev/null" if the caller is root.
2715 14. Ignore overlarge ACL variable number when reading spool file - insurance
2716 against a later release with more variables having written the file.
2718 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2719 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2720 this, as well as the form without the "IPv6" on the front.
2722 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2723 OS/Makefile-Darwin file.
2725 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2726 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2728 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2729 variable. However, some OS use ints for inodes. I've added cast to long int
2730 to get rid of the compiler warning.
2732 19. I had forgotten to lock out "/../" in configuration file names when
2733 ALT_CONFIG_PREFIX was set.
2735 20. Routers used for verification do not need to specify transports. However,
2736 if such a router generated a host list, and callout was configured, Exim
2737 crashed, because it could not find a port number from the (non-existent)
2738 transport. It now assumes port 25 in this circumstance.
2740 21. Added the -t option to exigrep.
2742 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2743 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2744 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2745 them is set, LOOKUP_LSEARCH is forced.
2747 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2749 24. Added sender and host information to the "rejected by local_scan()" log
2750 line; previously there was no indication of these.
2752 25. Added .include_if_exists.
2754 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2755 Linux. It turns out that not all file systems support this. Apparently some
2756 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2757 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2758 not supported on the file descriptor, is now ignored when Exim is trying to
2759 sync a directory. This applies only to Linux.
2761 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2763 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2764 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2767 29. Added support for maildirsize files from supplied patch (modified a bit).
2769 30. The use of :fail: followed by an empty string could lead Exim to respond to
2770 sender verification failures with (e.g.):
2772 550 Verification failed for <xxx>
2773 550 Sender verify failed
2775 where the first response line was missing the '-' that indicates it is not
2776 the final line of the response.
2778 31. The loop for finding the name of the user that called Exim had a hardwired
2779 limit of 10; it now uses the value of finduser_retries, which is used for
2780 all other user lookups.
2782 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2785 33. Exim was neglecting to zero errno before one call of strtol() when
2786 expanding a string and expecting an integer value. On some systems this
2787 resulted in spurious "integer overflow" errors. Also, it was casting the
2788 result into an int without checking.
2790 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2791 did not work. The code looks as if it has *never* worked, though it appears
2792 to have been documented since at least releast 1.62. I have made it work.
2794 35. The "timeout_DNS" error in retry rules, also documented since at least
2795 1.62, also never worked. As it isn't clear exactly what this means, and
2796 clearly it isn't a major issue, I have abolished the feature by treating it
2797 as "timeout", and writing a warning to the main and panic logs.
2799 36. The display of retry rules for -brt wasn't always showing the error code
2802 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2803 timeout_connect_A, timeout_connect_MX.
2805 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2806 to the empty sender.
2808 39. The daemon was not analysing the content of -oX till after it had closed
2809 stderr and disconnected from the controlling terminal. This meant that any
2810 syntax errors were only noted on the panic log, and the return code from
2811 the command was 0. By re-arranging the code a little, I've made the
2812 decoding happen first, so such errors now appear on stderr, and the return
2813 code is 1. However, the actual setting up of the sockets still happens in
2814 the disconnected process, so errors there are still only recorded on the
2817 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2818 connections (as happens on some IP stacks) was logged at start up time as
2819 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2820 it from "IPv6 and IPv4", which means that two separate sockets are being
2823 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2824 says whether AF_INET or AF_INET6 was passed as an argument.
2826 42. Exiwhat output was messed up when time zones were included in log
2829 43. Exiwhat now gives more information about the daemon's listening ports,
2830 and whether -tls-on-connect was used.
2832 44. The "port" option of the smtp transport is now expanded.
2834 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2835 silently ignored. Now an error message is written to the main and panic
2838 46. There's a new ACL modifier called "logwrite" which writes to a log file
2839 as soon as it is encountered.
2841 47. Added $local_user_uid and $local_user_gid at routing time.
2843 48. Exim crashed when trying to verify a sender address that was being
2846 49. Exim was recognizing only a space character after ".include". It now also
2847 recognizes a tab character.
2849 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2850 extracting the relevant information from the specification. The man page no
2851 longer contains scrambled data for the -d option, and I've added a section
2852 at the front about calling Exim under different names.
2854 51. Added "extra_headers" argument to the "mail" command in filter files.
2856 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2859 53. Installed eximstats 1.29.
2861 54. Added transport_filter_timeout as a generic transport option.
2863 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2864 Cc: header lines. This was required by RFC 822, but it not required by RFC
2867 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2868 incoming messages that did not have them. Now it does so only if the
2869 message originates locally, that is, if there is no associated remote host
2870 address. When Resent- header lines are present, this applies to the Resent-
2871 lines rather than the non-Resent- lines.
2873 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2874 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2876 58. Messages for configuration errors now include the name of the main
2877 configuration file - useful now that there may be more than one file in a
2878 list (.included file names were always shown).
2880 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2881 for those rare installations that do not start the daemon as root or run it
2882 setuid root. I've cut out the call to initgroups() if the daemon is not
2885 60. The Exim user and group can now be bound into the binary as text strings
2886 that are looked up at the start of Exim's processing.
2888 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2890 62. Added $mailstore_basename variable.
2892 63. Installed patch to sieve.c from Michael Haardt.
2894 64. When Exim failed to open the panic log after failing to open the main log,
2895 the original message it was trying to log was written to stderr and debug
2896 output, but if they were not available (the usual case in production), it
2897 was lost. Now it is written to syslog before the two lines that record the
2898 failures to open the logs.
2900 65. Users' Exim filters run in subprocesses under the user's uid. It is
2901 possible for a "deliver" command or an alias in a "personal" command to
2902 provoke an address rewrite. If logging of address rewriting is configured,
2903 this fails because the process is not running as root or exim. There may be
2904 a better way of dealing with this, but for the moment (because 4.30 needs
2905 to be released), I have disabled address rewrite logging when running a
2906 filter in a non-root, non-exim process.
2912 1. The buildconfig auxiliary program wasn't quoting the value set for
2913 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2914 not defined. This bug was masked in 4.22 by the effect that was fixed in
2917 2. Some messages that were rejected after a message id was allocated were
2918 shown as "incomplete" by exigrep. It no longer does this for messages that
2919 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2921 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2922 have allow_domain_literals set, the ID did not get logged in the <= line.
2923 Domain literals are now always recognized in Message-ID: header lines.
2925 4. The first argument for a ${extract expansion item is the key name or field
2926 number. Leading and trailing spaces in this item were not being ignored,
2927 causing some misleading effects.
2929 5. When deliver_drop_privilege was set, single queue runner processes started
2930 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2931 same command in the process it spins off) were not dropping privilege.
2933 6. When the daemon running as "exim" started a queue runner, it always
2934 re-executed Exim in the spun-off process. This is a waste of effort when
2935 deliver_drop_privilege is set. The new process now just calls the
2936 queue-runner function directly.
2942 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2945 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2946 ignored. Though the use of -oP was forcing the writing of a pid file, it
2947 was always written to the default place.
2949 3. If the message "no IP address found for host xxxx" is generated during
2950 incoming verification, it is now followed by identification of the incoming
2951 connection (so you can more easily find what provoked it).
2953 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2955 5. Added some features to "harden" Exim a bit more against certain attacks:
2957 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2958 be put in Local/Makefile. This is like the never_users runtime option,
2959 but it cannot be overridden. The default setting is "root".
2961 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2962 prefix string with which any file named in a -C command line option
2965 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2966 is retained for -C and -D only if the caller of Exim is root. Without
2967 it, the exim user may also use -C and -D and retain privilege.
2969 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2970 command line option is disabled.
2972 6. Macro names set by the -D option must start with an upper case letter, just
2973 like macro names defined in the configuration file.
2975 7. Added "dereference=" facility to LDAP.
2977 8. Two instances of the typo "uknown" in the source files are fixed.
2979 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2980 the Configure-Makefile script screwed up while processing it.
2982 10. Incorporated PCRE 4.4.
2984 11. The SMTP synchronization check was not operating right at the start of an
2985 SMTP session. For example, it could not catch a HELO sent before the client
2986 waited for the greeting. There is now a check for outstanding input at the
2987 point when the greeting is written. Because of the duplex, asynchronous
2988 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2989 way, but not yet received, when the check is performed.
2991 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2992 on TCP/IP sockets, because this apparently causes some broken clients to
2995 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2996 unchanged) from the Cygwin maintainer.
2998 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2999 was supported instead of testing for mbx. Effectively a typo.
3001 15. The spa authenticator server code was not checking that the input it
3002 received was valid base64.
3004 16. The debug output line for the "set" modifier in ACLs was not showing the
3005 name of the variable that was being set.
3007 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3009 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3010 lookup on its own. Something else was needed to trigger the lookup. For
3011 example, a match in host_lookup or the need for a host name in a host list.
3012 Now, if $sender_host_name is referenced and the host name has not yet been
3013 looked up, a lookup is performed. If the lookup fails, the variable remains
3014 empty, and $host_lookup_failed is set to "1".
3016 19. Added "eqi" as a case-independent comparison operator.
3018 20. The saslauthd authentication condition could segfault if neither service
3019 nor realm was specified.
3021 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3022 error message that was logged was misleading, and incoming SMTP
3023 connections were dropped. The message is now more accurate, and temporary
3024 errors are given to SMTP connections.
3026 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3027 (except RSET) with a 421 error, until QUIT is received. However, it was
3028 failing to send a response to QUIT.
3030 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3031 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3032 ACL is now run after the helo_try_verify_hosts code.
3034 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3035 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3036 case-independent, so other case variants are also recognized). Apparently
3037 some systems use these upper case variants.
3039 25. If more than two messages were waiting for the same host, and a transport
3040 filter was specified for the transport, Exim sent two messages over the
3041 same TCP/IP connection, and then failed with "socket operation on non-
3042 socket" when it tried to send the third.
3044 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3046 27. The extern definition of crypt16() in expand.c was not being excluded when
3047 the OS had its own crypt16() function.
3049 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3050 as a preferred synonym for return_size_limit, both as an option and as an
3053 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3055 30. Changed the default configuration ACL to relax the local part checking rule
3056 for addresses that are not in any local domains. For these addresses,
3057 slashes and pipe symbols are allowed within local parts, but the sequence
3058 /../ is explicitly forbidden.
3060 31. SPA server authentication was not clearing the challenge buffer before
3063 32. log_message in a "warn" ACL statement was writing to the reject log as
3064 well as to the main log, which contradicts the documentation and doesn't
3065 seem right (because no rejection is happening). So I have stopped it.
3067 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3068 However, I am unable to do any testing of this.
3070 34. Fixed an infelicity in the appendfile transport. When checking directories
3071 for a mailbox, to see if any needed to be created, it was accidentally
3072 using path names with one or more superfluous leading slashes; tracing
3073 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3075 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3076 contents are added to the log line that is written for every discarded
3077 recipient. (Previously a log_message setting was ignored.)
3079 36. The ${quote: operator now quotes the string if it is empty.
3081 37. The install script runs exim in order to find its version number. If for
3082 some reason other than non-existence or emptiness, which it checks, it
3083 could not run './exim', it was installing it with an empty version number,
3084 i.e. as "exim-". This error state is now caught, and the installation is
3087 38. An argument was missing from the function that creates an error message
3088 when Exim fails to connect to the socket for saslauthd authentication.
3089 This could cause Exim to crash, or give a corrupted message.
3091 39. Added isip, isip4, and isip6 to ${if conditions.
3093 40. The ACL variables $acl_xx are now saved with the message, and can be
3094 accessed later in routers, transports, and filters.
3096 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3097 strings in the file are not string-expanded.
3099 42. If a MAIL command specified a SIZE value that was too large to fit into an
3100 int variable, the check against message_size_limit failed. Such values are
3101 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3102 day this will have to be increased, but I don't think I want to be around
3103 when emails are that large.
3110 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3111 iconv() is not standard in FreeBSD.
3113 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3114 IPv6 enabled. The observed symptom was a segmentation fault on return from
3115 the function os_common_find_running_interfaces() in src/os.c.
3117 3. In the check_special_case() function in daemon.c I had used "errno" as an
3118 argument name, which causes warnings on some systems. This was basically a
3119 typo, since it was named "eno" in the comments!
3121 4. The code that waits for the clock to tick (at a resolution of some fraction
3122 of a second) so as to ensure message-id uniqueness was always waiting for
3123 at least one whole tick, when it could have waited for less. [This is
3124 almost certainly not relevant at current processor speeds, where it is
3125 unlikely to ever wait at all. But we try to future-proof.]
3127 5. The function that sleeps for a time interval that includes fractions of a
3128 second contained a race. It did not block SIGALRM between setting the
3129 timer, and suspending (a couple of lines later). If the interval was short
3130 and the sigsuspend() was delayed until after it had expired, the suspension
3131 never ended. On busy systems this could lead to processes getting stuck for
3134 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3135 process, before it forks any delivery processes. The open lookup caching
3136 mechanism meant that the open file or database connection was passed into
3137 the delivery process. The problem was that delivery processes always tidy
3138 up cached lookup data. This could cause a problem for the next delivery
3139 process started by the queue runner, because the external queue runner
3140 process does not know about the closure. So the next delivery process
3141 still has data in the lookup cache. In the case of a file lookup, there was
3142 no problem because closing a file descriptor in a subprocess doesn't affect
3143 the parent. However, if the lookup was caching a connection to a database,
3144 the connection was closed, and the second delivery process was likely to
3145 see errors such as "PGSQL: query failed: server closed the connection
3146 unexpectedly". The problem has been fixed by closing all cached lookups
3147 in a queue runner before running a delivery process.
3149 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3150 seem to have the "const" qualifier which it has on other OS. I've
3153 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3154 *of the same type* (client or server) with the same public name that an
3155 error should be diagnosed.
3157 9. When Exim looked up a host name for an IP address, but failed to find the
3158 original IP address when looking up the host name (a safety check), it
3159 output the message "<ip address> does not match any IP for NULL", which was
3160 confusing, to say the least. The bug was that the host name should have
3161 appeared instead of "NULL".
3163 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3164 user that is built into the binary, and the -C or -D options is used, root
3165 privilege is dropped before the configuration file is read. In addition,
3166 logging is switched to stderr instead of the normal log files. If the
3167 configuration then re-defines the Exim user, the unprivileged environment
3168 is probably not what is expected, so Exim logs a panic warning message (but
3171 However, if deliver_drop_privilege is set, the unprivileged state may well
3172 be exactly what is intended, so the warning has been cut out in that case,
3173 and Exim is allowed to try to write to its normal log files.
3179 1. smtp_return_error_details was not giving details for temporary sender
3180 or receiver verification errors.
3182 2. Diagnose a configuration error if two authenticators have the same public
3185 3. Exim used not to create the message log file for a message until the first
3186 delivery attempt. This could be confusing when incoming messages were held
3187 for policy or load reasons. The message log file is now created at the time
3188 the message is received, and an initial "Received" line is written to it.
3190 4. The automatically generated man page for command line options had a minor
3191 bug that caused no ill effects; however, a more serious problem was that
3192 the procedure for building the man page automatically didn't always
3193 operate. Consequently, release 4.20 contains an out-of-date version. This
3194 shouldn't happen again.
3196 5. When building Exim with embedded Perl support, the script that builds the
3197 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3198 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3200 6. The freeze_tell option was not being used for messages that were frozen on
3201 arrival, either by an ACL or by local_scan().
3203 7. Added the smtp_incomplete_transaction log selector.
3205 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3206 was accepting AUTH without a new EHLO.
3208 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3209 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3212 10. Logging of TCP/IP connections (when configured) now happens in the main
3213 daemon process instead of the child process, so that the TCP/IP connection
3214 count is more accurate (but it can never be perfect).
3216 11. The use of "drop" in a nested ACL was not being handled correctly in the
3217 outer ACL. Now, if condition failure induced by the nested "drop" causes
3218 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3219 or "require"), the connection is dropped.
3221 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3222 that yield "discard" can now be used with an "accept" or a "discard" verb,
3223 but an error is generated for any others (because I can't see a useful way
3224 to define what should happen).
3226 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3227 are now processed in the same way as lines in the Exim configuration file.
3228 In particular, continuation lines are supported.
3230 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3232 15. Added -ti meaning -t -i.
3234 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3235 domains, and warn by logging if others are found.
3237 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3238 returned from GIFCONF when Exim is trying to find the list of interfaces on
3239 a host. The code in os.c has been modified to copy each ifreq to an aligned
3240 structure in all cases.
3242 Also, in some cases, the returned ifreq's were being copied to a 'struct
3243 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3244 means the last couple of bytes of an IPv6 address could be chopped if the
3245 ifreq contained only a normal sockaddr (14 bytes storage).
3247 18. Named domain lists were not supported in the hosts_treat_as_local option.
3248 An entry such as +xxxx was not recognized, and was treated as a literal
3251 19. Ensure that header lines added by a DATA ACL are included in the reject log
3252 if the ACL subsequently rejects the message.
3254 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3255 MD5 (which is deprecated).
3257 21. When testing a filter file using -bf, Exim was writing a message when it
3258 took the sender from a "From " line in the message, but it was not doing so
3259 when it took $return_path from a Return-Path: header line. It now does.
3261 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3262 with a valid header line field name (a series of printing characters
3263 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3265 23. Changed "disc" in the source to "disk" to conform to the documentation and
3266 the book and for uniformity.
3268 24. Ignore Sendmail's -Ooption=value command line item.
3270 25. When execve() failed while trying to run a command in a pipe transport,
3271 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3272 could be confused with a return value of 69 from the command itself. This
3273 has been changed to 127, the value the shell returns if it is asked to run
3274 a non-existent command. The wording for the related log line suggests a
3275 non-existent command as the problem.
3277 26. If received_header_text expands to an empty string, do not add a Received:
3278 header line to the message. (Well, it adds a token one on the spool, but
3279 marks it "old" so that it doesn't get used or transmitted.)
3281 27. Installed eximstats 1.28 (addition of -nt option).
3283 28. There was no check for failure on the call to getsockname() in the daemon
3284 code. This can fail if there is a shortage of resources on the system, with
3285 ENOMEM, for example. A temporary error is now given on failure.
3287 29. Contrary to the C standard, it seems that in some environments, the
3288 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3289 program. Exim now does this explicitly; it affects the formatting of
3290 timestamps using strftime().
3292 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3293 complaints. I've now initialized all the variables, to avoid this.
3295 32. Header lines added by a system filter were not being "seen" during
3296 transport-time rewrites.
3298 33. The info_callback() function passed to OpenSSL is set up with type void
3299 (*)(SSL *, int, int), as described somewhere. However, when calling the
3300 function (actually a macro) that sets it up, the type void(*)() is
3301 expected. I've put in a cast to prevent warnings from picky compilers.
3303 34. If a DNS black list lookup found a CNAME record, but there were no A
3304 records associated with the domain it pointed at, Exim crashed.
3306 35. If a DNS black list lookup returned more than one A record, Exim ignored
3307 all but the first. It now scans all returned addresses if a particular IP
3308 value is being sought. In this situation, the contents of the
3309 $dnslist_value variable are a list of all the addresses, separated by a
3312 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3313 to accept a host name and all its aliases if the forward lookup for any of
3314 them yielded the IP address of the incoming connection. Now it accepts only
3315 those names whose forward lookup yields the correct IP address. Any other
3316 names are discarded. This closes a loophole whereby a rogue DNS
3317 administrator could create reverse DNS records to break through a
3318 wildcarded host restriction in an ACL.
3320 37. If a user filter or a system filter that ran in a subprocess used any of
3321 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3322 the wrong values were passed to the pipe command ($thisaddress had the
3323 value of $0, $0 had the value of $1, etc). This bug was introduced by
3324 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3326 38. Improved the line breaking for long SMTP error messages from ACLs.
3327 Previously, if there was no break point between 40 and 75 characters, Exim
3328 left the rest of the message alone. Two changes have been made: (a) I've
3329 reduced the minimum length to 35 characters; (b) if it can't find a break
3330 point between 35 and 75 characters, it looks ahead and uses the first one
3331 that it finds. This may give the occasional overlong line, but at least the
3332 remaining text gets split now.
3334 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3335 file descriptors might be low, and that setting 1000 would always raise it.
3336 It turns out that in some environments, the limit is already over 1000 and
3337 that lowering it causes trouble. So now Exim takes care not to decrease it.
3339 40. When delivering a message, the value of $return_path is set to $sender_
3340 address at the start of routing (routers may change the value). By an
3341 oversight, this default was not being set up when an address was tested by
3342 -bt or -bv, which affected the outcome if any router or filter referred to
3345 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3346 log only when "message" or "log_message" are set. However, if one of the
3347 conditions was an address verification, or a call to a nested ACL, the
3348 messages generated by the underlying test were being passed through. This
3349 no longer happens. The underlying message is available in $acl_verify_
3350 message for both "message" and "log_message" expansions, so it can be
3351 passed through if needed.
3353 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3354 new expansion $bh_ to give the encoded byte string without charset
3355 translation. Translation happens only if iconv() is available; HAVE_ICONV
3356 indicates this at build time. HEADERS_CHARSET gives the charset to
3357 translate to; headers_charset can change it in the configuration, and
3358 "headers charset" can change it in an individual filter file.
3360 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3361 that creates RFC 2047 encoded "words" labels them as that charset instead
3362 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3363 expansion operator; (ii) when Exim creates a From: line for a local
3364 message; (iii) when a header line is rewritten to include a "phrase" part.
3366 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3367 buggy, causing it to skip the first lines of messages whose message ID
3368 ended in 'D'. This would not have bitten before Exim release 4.14, because
3369 message IDs were unlikely to end in 'D' before then. The effect was to have
3370 incorrect size information for certain domains.
3372 45. #include "config.h" was missing at the start of the crypt16.c module. This
3373 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3376 46. If there was a timeout during a "random" callout check, Exim treated it as
3377 a failure of the random address, and carried on sending RSET and the real
3378 address. If the delay was just some slowness somewhere, the response to the
3379 original RCPT would be taken as a response to RSET and so on, causing
3380 mayhem of various kinds.
3382 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3383 when I implemented it. It didn't allow for the fact that some option values
3384 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3385 the right test for positive values.
3387 48. Domain names in DNS records are case-independent. Exim always looks them up
3388 in lower case. Some resolvers return domain names in exactly the case they
3389 appear in the zone file, that is, they may contain uppercase letters. Not
3390 all resolvers do this - some return always lower case. Exim was treating a
3391 change of case by a resolver as a change of domain, similar to a widening
3392 of a domain abbreviation. This triggered its re-routing code and so it was
3393 trying to route what was effectively the same domain again. This normally
3394 caused routing to fail (because the router wouldn't handle the domain
3395 twice). Now Exim checks for this case specially, and just changes the
3396 casing of the domain that it ultimately uses when it transmits the message
3399 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3402 50. If a filter generated a file delivery with a non-absolute name (possible if
3403 no home directory exists for the router), the forbid_file option was not
3406 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3407 the existing equality matching.
3409 52. Exim was using ints instead of ino_t variables in some places where it was
3410 dealing with inode numbers.
3412 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3413 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3414 called TMPDIR, and if it finds it is different, it changes its value.
3416 54. The smtp_printf() function is now made available to local_scan() so
3417 additional output lines can be written before returning. There is also an
3418 smtp_fflush() function to enable the detection of a dropped connection.
3419 The variables smtp_input and smtp_batched_input are exported to
3422 55. Changed the default runtime configuration: the message "Unknown user"
3423 has been removed from the ACL, and instead placed on the localuser router,
3424 using the cannot_route_message feature. This means that any verification
3425 failures that generate their own messages won't get overridden. Similarly,
3426 the "Unrouteable address" message that was in the ACL for unverifiable
3427 relay addresses has also been removed.
3429 56. Added hosts_avoid_esmtp to the smtp transport.
3431 57. The exicyclog script was not checking for the esoteric option
3432 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3433 will work only if exicyclog is run under the appropriate euid.
3435 58. Following a discussion on the list, the rules by which Exim recognises line
3436 endings on incoming messages have been changed. The -dropcr and drop_cr
3437 options are now no-ops, retained only for backwards compatibility. The
3438 following line terminators are recognized: LF CRLF CR. However, special
3439 processing applies to CR:
3441 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3442 nor a local message in the state where . is a terminator.
3444 (ii) If a bare CR is encountered in a header line, an extra space is added
3445 after the line terminator so as not to end the header. The reasoning
3446 behind this is that bare CRs in header lines are most likely either
3447 to be mistakes, or people trying to play silly games.
3449 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3450 was being incorrectly given as 18 bytes larger than it should have been.
3451 This is a VOB (very old bug).
3453 60. This may never have affected anything current, but just in case it has:
3454 When the local host is found other than at the start of a list of hosts,
3455 the local host, those with the same MX, and any that follow, are discarded.
3456 When the list in question was part of a longer list of hosts, the following
3457 hosts (not currently being processed) were also being discarded. This no
3458 longer happens. I'm not sure if this situation could ever has previously
3461 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3464 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3465 header. This is something that is recommended in a new Internet Draft, and
3466 is something that is documented as being done by Sendmail. There are two
3467 possible values. For messages generated by the autoreply transport, Exim
3470 Auto-Submitted: auto-replied
3472 whereas for all other generated messages (e.g. bounces) it adds
3474 Auto-Submitted: auto-generated
3476 63. The "personal" condition in filters now includes a test for the
3477 Auto-Submitted: header. If it contains the string "auto-" the message it
3478 not considered personal.
3480 64. Added rcpt_include_affixes as a generic transport option.
3482 65. Added queue_only_override (default true).
3484 66. Added the syslog_duplication option.
3486 67. If what should have been the first header line of a message consisted of
3487 a space followed by a colon, Exim was mis-interpreting it as a header line.
3488 It isn't of course - it is syntactically invalid and should therefore be
3489 treated as the start of the message body. The misbehaviour could have
3490 caused a number of strange effects, including loss of data in subsequent
3491 header lines, and spool format errors.
3493 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3494 client host had authenticated. This control can now be exercised by an ACL
3495 for more flexibility.
3497 69. By default, callouts do not happen when testing with -bh. There is now a
3498 variant, -bhc, which does actually run the callout code, including
3499 consulting and updating the callout cache.
3501 70. Added support for saslauthd authentication, courtesy of Alexander
3504 71. If statvfs() failed on the spool or log directories while checking their
3505 size for availability, Exim confusingly gave the error "space shortage".
3506 Furthermore, in debugging mode it crashed with a floating point exception.
3507 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3508 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3509 serious problem, Exim now writes to the main and panic logs when this
3510 happens, with details of the failure. It then refuses to accept the
3511 incoming message, giving the message "spool directory problem" or "log
3512 directory problem" with a 421 code for SMTP messages.
3514 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3515 0, 1, and 2 exist, because some OS complain for execs without them (see
3516 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3517 descriptors. However, the code omitted to check that the open succeeded,
3518 causing mysterious errors if for some reason the permissions on /dev/null
3519 got screwed. Now Exim writes a message to the main and panic logs, and
3520 bombs out if it can't open /dev/null.
3522 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3523 interact so that it is all more flexible. It is supposed to remain
3524 backwards compatible. Also added extra_local_interfaces.
3526 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3527 to bomb out with an assertion failure - to the client this appears as a
3528 connection drop. This problem occurs in the part of the code that was taken
3529 from the Samba project. Fortunately, the assertion is in a very simple
3530 function, so I have fixed this by reproducing the function inline in the
3531 one place where it is called, and arranging for authentication to fail
3532 instead of killing the process with assert().
3534 75. The SPA client code was not working when the server requested OEM rather
3535 than Unicode encoding.
3537 76. Added code to make require_files with a specific uid setting more usable in
3538 the case where statting the file as root fails - usually a non-root-mounted
3539 NFS file system. When this happens and the failure is EACCES, Exim now
3540 forks a subprocess and does the per-uid checking as the relevant uid.
3542 77. Added process_log_path.
3544 78. If log_file_path was not explicitly set, a setting of check_log_space or
3545 check_log_inodes was ignored.
3547 79. If a space check for the spool or log partitions fails, the incident is now
3548 logged. Of course, in the latter case the data may get lost...
3550 80. Added the %p formatting code to string_format() so that it can be used to
3551 print addresses in debug_print(). Adjusted all the address printing in the
3552 debugging in store.c to use %p rather than %d.
3554 81. There was a concern that a line of code in smtp_in.c could overflow a
3555 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3556 initially expressed, the concern was not well-founded, because trailing
3557 spaces are removed early. However, if the trailing spaces were followed by
3558 a NULL, they did not get removed, so the overflow was possible. Two fixes
3561 (a) I re-wrote the offending code in a cleaner fashion.
3562 (b) If an incoming SMTP command contains a NULL character, it is rejected
3565 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3566 runs initgroups(), so that if the Exim user is in any additional groups,
3567 they will be used during message reception.
3573 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob