1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.270 2005/12/05 14:38:18 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
24 PH/01 Two changes to the default runtime configuration:
26 (1) Move the checks for relay_from_hosts and authenticated clients from
27 after to before the (commented out) DNS black list checks.
29 (2) Add control=submission to the relay_from_hosts and authenticated
30 clients checks, on the grounds that messages accepted by these
31 statements are most likely to be submissions.
33 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
35 (1) Generate an error if the third argument for the ${prvs expansion is
38 (2) Treat a missing third argument of ${prvscheck as if it were an empty
41 (3) Reset the variables that are obtained from the first argument of
42 ${prvscheck and used in the second argument before leaving the code,
43 because their memory is reclaimed, so using them afterwards may do
46 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
47 one (it's much easier than Tom thought :-).
49 (5) Because of (4), we can now allow for the use of $prvscheck_result
50 inside the third argument.
52 PH/03 For some reason, the default setting of PATH when running a command from
53 a pipe transport was just "/usr/bin". I have changed it to
56 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
57 anything to be listed in the output from -bV.
59 PH/05 When a filter generated an autoreply, the entire To: header line was
60 quoted in the delivery log line, like this:
62 => >A.N.Other <ano@some.domain> <original@ddress> ...
64 This has been changed so that it extracts the operative address. There
65 may be more than one such address. If so, they are comma-separated, like
68 => >ano@some.domain,ona@other.domain <original@ddress> ...
70 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
71 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
72 not being looked up in the rDNS to get a host name, Exim was showing the
73 IP address twice in Received: lines, even though the IP addresses were
74 identical. For example:
76 Received: from [1.2.3.4] (helo=[1.2.3.4])
78 However, if the real host name was known, it was omitting the HELO data
79 if it matched the actual IP address. This has been tidied up so that it
80 doesn't show the same IP address twice.
82 PH/07 When both +timestamp and +memory debugging was on, the value given by
83 $tod_xxx expansions could be wrong, because the tod_stamp() function was
84 called by the debug printing, thereby overwriting the timestamp buffer.
85 Debugging no longer uses the tod_stamp() function when +timestamp is set.
87 PH/08 When the original message was included in an autoreply transport, it
88 always said "this is a copy of the message, including all the headers",
89 even if body_only or headers_only was set. It now gives an appropriate
92 PH/09 Applied a patch from the Sieve maintainer which:
95 o adds the (disabled) notify extension core
96 o adds some debug output for the result of if/elsif tests
97 o points to the current vacation draft in the documentation
98 and documents the missing references header update
102 o fixes a bug in processing the envelope test (when testing
103 multiple envelope elements, the last element determinted the
106 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
107 Electronic Mail") by including:
109 Auto-submitted: auto-generated
111 in the messages that it generates (bounce messages and others, such as
112 warnings). In the case of bounce messages for non-SMTP mesages, there was
113 also a typo: it was using "Auto_submitted" (underscore instead of
114 hyphen). Since every message generated by Exim is necessarily in response
115 to another message, thes have all been changed to:
117 Auto-Submitted: auto-replied
119 in accordance with these statements in the RFC:
121 The auto-replied keyword:
123 - SHOULD be used on messages sent in direct response to another
124 message by an automatic process,
126 - MUST NOT be used on manually-generated messages,
128 - MAY be used on Delivery Status Notifications (DSNs) and Message
129 Disposition Notifications (MDNs),
131 - MUST NOT be used on messages generated by automatic or periodic
132 processes, except for messages which are automatic responses to
135 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
136 to the default Received: header definition.
138 PH/12 Added log selector acl_warn_skipped (default on).
140 PH/13 After a successful wildlsearch lookup, discard the values of numeric
141 variables because (a) they are in the wrong storage pool and (b) even if
142 they were copied, it wouldn't work properly because of the caching.
144 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
145 checking when decoding. Apparently there are clients that generate
146 overlong encoded strings. Why am I not surprised?
148 PH/15 If the first argument of "${if match_address" was not empty, but did not
149 contain an "@" character, Exim crashed. Now it writes a panic log message
150 and treats the condition as false.
152 PH/16 In autoreply, treat an empty string for "once" the same as unset.
154 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
155 extension "envelope-auth". The code is finished and in agreement with
156 other implementations, but there is no documentation so far and in fact,
157 nobody wrote the draft yet. This extension is currently #undef'ed, thus
158 not changing the active code.
160 Print executed "if" and "elsif" statements when debugging is used. This
161 helps a great deal to understand what a filter does.
163 Document more things not specified clearly in RFC3028. I had all this
164 sorted out, when out of a sudden new issues came to my mind. Oops."
166 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
169 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
170 canonical form (as documented). However, after a host name lookup from
171 the IP address, check_host() was doing a simple string comparison with
172 addresses acquired from the DNS when checking that the found name did
173 have the original IP as one of its addresses. Since any found IPv6
174 addresses are likely to be in abbreviated form, the comparison could
175 fail. Luckily, there already exists a function for doing the comparison
176 by converting both addresses to binary, so now that is used instead of
179 PH/20 There was another similar case to PH/19, when a complete host name was
180 given in a host list; looking up its IP address could give an abbreviated
181 form, whereas the current host's name might or might not be abbreviated.
182 The same fix has been applied.
188 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
189 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
192 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
193 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
195 PH/03 Typo: missing ".o" in src/pcre/Makefile.
197 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
198 header line, restrict the check to what is listed in RFCs 2369 and 2929.
199 Also, for "Auto-Submitted", treat anything other than "no" as
200 non-personal, in accordance with RFC 3834. (Previously it treated
201 anything starting "auto-" as non-personal.)
203 TF/01 The control=submission/name=... option had a problem with syntax
204 errors if the name included a slash character. The /name= option
205 now slurps the rest of the string, so it can include any characters
206 but it must come last in the list of options (after /sender_retain
209 PH/05 Some modifications to the interface to the fake nameserver for the new
212 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
213 into the default Received: header string.
220 TK/01 Added the "success_on_redirect" address verification option. See
221 NewStuff for rationale and an example.
223 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
225 PH/02 Patch to exigrep to allow it to work on syslog lines.
227 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
228 fread() to skip over the body file's header line, because in Cygwin the
229 header line is locked and is inaccessible.
231 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
232 co-exist for some time) to make it clear that it is the Exim ID that is
233 referenced, not the Message-ID: header line.
235 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
236 string_format() function, because snprintf() does not exist on all
239 PH/06 The use of forbid_filter_existstest now also locks out the use of the
240 ${stat: expansion item.
242 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
243 protocol synchronization error", to keep the pedants happy.
245 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
246 well as for IRIX systems, when gcc is being used. See the host.c source
249 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
251 PH/10 Named domain lists were not working if used in a queue_smtp_domains
254 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
255 transport and to the smtp transport in LMTP mode.
257 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
259 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
260 run a filter in a subprocess. This could lead to confusion in subsequent
261 lookups in the parent process. There should also be a search_tidyup() at
262 the end of the subprocess.
264 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
265 only if the host matched helo_try_verify_hosts, which caused the
266 verification to occur when the EHLO/HELO command was issued. The ACL just
267 tested the remembered result. Now, if a previous verification attempt has
268 not happened, "verify = helo" does it there and then.
270 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
272 TK/03 Fix log output including CR from clamd.
274 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
275 exist provoked a memory error which could cause a segfault.
277 PH/15 Installed PCRE 6.2
279 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
281 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
282 of the problem. Specifically, suggested +O2 rather than +O1 for the
285 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
287 PH/20 If a delivery was routed to a non-standard port by means of an SRV
288 record, the port was not correctly logged when the outgoing_port log
289 selector was set (it logged the transort's default port).
291 PH/21 Added support for host-specific ports to manualroute, queryprogram,
292 fallback_hosts, and "hosts" in the smtp transport.
294 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
295 host errors such as "Connection refused".
297 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
298 authentication with radiusclient 0.4.9:
300 - Error returned from rc_read_config was caught wrongly
301 - Username/password not passed on to radius server due to wrong length.
303 The presumption is that some radiusclient API changes for 4.51/PH/17
304 were not taken care of correctly. The code is still untested by me (my
305 Linux distribution still has 0.3.2 of radiusclient), but it was
306 contributed by a Radius user.
308 PH/24 When doing a callout, the value of $domain wasn't set correctly when
309 expanding the "port" option of the smtp transport.
311 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
312 while reading a MIME header. Thanks to Tom Hughes for a patch.
314 PH/24 Include config.h inside local_scan.h so that configuration settings are
317 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
318 that in an ACL for RCPT (for example), you can examine exactly what was
321 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
322 commands, but it was not correctly comparing the address with the actual
323 client host address. Thus, it would show the EHLO address in Received:
324 header lines when this was not necessary.
326 PH/27 Added the % operator to ${eval:}.
328 PH/28 Exim tries to create and chdir to its spool directory when it starts;
329 it should be ignoring failures (because with -C, for example, it has lost
330 privilege). It wasn't ignoring creation failures other than "already
333 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
336 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
337 because an input line was too long, either on its own, or by virtue of
338 too many continuations, the temporary file was not being removed, and the
339 return code was incorrect.
341 PH/31 Missing "BOOL" in function definition in filtertest.c.
343 PH/32 Applied Sieve patches from the maintainer.
345 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
347 PH/33 Added "verify = not_blind".
349 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
350 Local/Makefile (with some defaults set). These are used in built scripts
351 such as exicyclog, but they have never been used in the exim_install
352 script (though there are many overriding facilities there). I have
353 arranged that the exim_install script now takes note of these two
356 PH/35 Installed configuration files for Dragonfly.
358 PH/36 When a locally submitted message by a trusted user did not contain a
359 From: header, and the sender address was obtained from -f or from an SMTP
360 MAIL command, and the trusted user did not use -F to supply a sender
361 name, $originator_name was incorrectly used when constructing a From:
362 header. Furthermore, $originator_name was used for submission mode
363 messages from external hosts without From: headers in a similar way,
364 which is clearly wrong.
366 PH/37 Added control=suppress_local_fixups.
368 PH/38 When log_selector = +received_sender was set, and the addition of the
369 sender made the log line's construction buffer exactly full, or one byte
370 less than full, an overflow happened when the terminating "\n" was
373 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
374 when the result of a list match is failure because a DNS lookup failed.
376 PH/40 RM_COMMAND is now used in the building process.
378 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
379 the "build-* directories that it finds.
381 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
382 address in a domain literal was a prefix of an interface address.
384 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
385 when verifying a sender address, unless rewrite_headers is false.
387 PH/44 Wrote a long comment about why errors_to addresses are verified as
388 recipients, not senders.
390 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
391 the ratelimit ACL was added.
393 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
395 PH/46 Added extra information about PostgreSQL errors to the error string.
397 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
398 avoiding the need to install special zones in a real server. This is
399 backwards compatible; if it can't find the fake resolver, it drops back.
400 Thus, both old and new test suites can be run.
402 TF/02 Added util/ratelimit.pl
404 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
405 clock is set back in time.
407 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
408 Candler <B.Candler@pobox.com>.
410 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
411 OK for addresses that are the result of redirections.
413 PH/48 A number of further additions for the benefit of the new test suite,
414 including a fake gethostbyname() that interfaces to the fake DNS resolver
417 TF/06 The fix for widen_domains has also been applied to qualify_single and
418 search_parents which are the other dnslookup options that can cause
421 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
424 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
426 TF/07 Exim produced the error message "an SRV record indicated no SMTP
427 service" if it encountered an MX record with an empty target hostname.
428 The message is now "an MX or SRV record indicated no SMTP service".
430 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
431 if the DNS of the sending site is misconfigured. This is quite a
432 common situation. This change restores the behaviour of treating a
433 helo verification defer as a failure.
435 PH/51 If self=fail was set on a router, the bounce message did not include the
436 actual error message.
442 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
444 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
445 command itself ended in error, the underlying message about the transport
446 filter timeout was being overwritten with the pipe command error. Now the
447 underlying error message should be appended to the second error message.
449 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
451 PH/02 Reduce the amount of output that "make" produces by default. Full output
452 can still be requested.
454 PH/03 The warning log line about a condition test deferring for a "warn" verb
455 was being output only once per connection, rather than after each
456 occurrence (because it was using the same function as for successful
457 "warn" verbs). This seems wrong, so I have changed it.
459 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
460 it should not have, which might have caused a crash in the right
461 circumstances, but probably never did.
463 PH/04 Installed a modified version of Tony Finch's patch to make submission
464 mode fix the return path as well as the Sender: header line, and to
465 add a /name= option so that you can make the user's friendly name appear
468 TF/03 Added the control = fakedefer ACL modifier.
470 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
471 Mark Lowes for thorough testing.
473 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
475 TK/03 Merged latest SRS patch from Miles Wilton.
477 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
478 with the definition in sysexits.h (which is #included earlier).
479 Fortunately, Exim does not actually use EX_OK. The code used to try to
480 preserve the sysexits.h value, by assumimg that macro definitions were
481 scanned for macro replacements. I have been disabused of this notion,
482 so now the code just undefines EX_OK before #including unistd.h.
484 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
485 in the smtp transport. When a block could not be written in a single
486 write() function, the timeout was being re-applied to each part-write.
487 This seems wrong - if the receiver was accepting one byte at a time it
488 would take for ever. The timeout is now adjusted when this happens. It
489 doesn't have to be particularly precise.
491 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
492 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
494 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
495 without a domain if the check to <postmaster@domain> fails.
497 SC/01 Eximstats: added -xls and the ability to specify output files
498 (patch written by Frank Heydlauf).
500 SC/02 Eximstats: use FileHandles for outputing results.
502 SC/03 Eximstats: allow any combination of xls, txt, and html output.
504 SC/04 Eximstats: fixed display of large numbers with -nvr option
506 SC/05 Eximstats: fixed merging of reports with empty tables.
508 SC/06 Eximstats: added the -include_original_destination flag
510 SC/07 Eximstats: removed tabs and trailing whitespace.
512 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
514 TK/06 MBOX spool code: Add real "From " MBOX separator line
515 so the .eml file is really in mbox format (even though
516 most programs do not really care). Patch from Alex Miller.
518 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
519 The latter is generated from $received_to and is only set if the
520 message has one envelope recipient. SA can use these headers,
521 obviously out-of-the-box. Patch from Alex Miller.
523 PH/08 The ${def test on a variable was returning false if the variable's
524 value was "0", contrary to what the specification has always said!
525 The result should be true unless the variable is empty.
527 PH/09 The syntax error of a character other than { following "${if
528 def:variable_name" (after optional whitespace) was not being diagnosed.
529 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
530 accidental colon was present, for example, could give incorrect results.
532 PH/10 Tidied the code in a number of places where the st_size field of a stat()
533 result is used (not including appendfile, where other changes are about
536 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
537 This involved changing a lot of size variables from int to off_t. It
538 should work with maildirs and everything.
540 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
541 spamd dying while we are connected to it.
543 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
546 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
547 function for BDB 4.3.
549 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
551 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
552 the sources for PCRE have been re-arranged and more files are now
555 PH/15 The code I had for printing potentially long long variables in PH/11
556 above was not the best (it lost precision). The length of off_t variables
557 is now inspected at build time, and an appropriate printing format (%ld
558 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
559 to be "long long int" or "long int". This is needed for the internal
560 formatting function string_vformat().
562 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
563 the configuration file to be ":syslog", then the script "guesses" where
564 the logs files are, rather than using the compiled in default. In our
565 case the guess is not the same as the compiled default, so the script
566 suddenly stopped working when I started to use syslog. The patch checks
567 to see if log_file_path is "". If so, it attempts to read it from exim
568 with no configuration file to get the compiled in version, before it
569 falls back to the previous guessing code."
571 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
572 implementing BATV in an Exim configuration. See NewStuff for the gory
575 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
576 Makefile that are specific to HP-UX.
578 PH/18 If the "use_postmaster" option was set for a recipient callout together
579 with the "random" option, the postmaster address was used as the MAIL
580 FROM address for the random test, but not for the subsequent recipient
581 test. It is now used for both.
583 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
584 patch removes a few documentation additions to RFC 3028, because the
585 latest draft now contains them. It adds the new en;ascii-case comparator
586 and a new error check for 8bit text in MIME parts. Comparator and
587 require names are now matched exactly. I enabled the subaddress
588 extension, but it is not well tested yet (read: it works for me)."
590 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
591 rework some of the code of TK/09 above to avoid the hardwired use of
592 "%lld" and "long long". Replaced the call to snprintf() with a call to
595 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
596 records point to non-existent hosts", "retry timeout exceeded", and
597 "retry time not reached for any host after a long failure period".
599 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
600 experimental DomainKeys support:
602 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
603 (2) On an error such as an illegally used "control", the wrong name for
604 the control was given.
606 These problems did NOT occur unless DomainKeys support was compiled.
608 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
610 PH/24 Added ${if match_ip condition.
612 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
613 databases so that it will be absolutely obvious if a crash occurs in the
614 DB library. This is a regular occurrence (often caused by mis-matched
617 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
618 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
619 picked up on a user's system that detects such things. There doesn't seem
620 to be a gcc warning option for this - only an attribute that has to be
621 put on the function's prototype. It seems that in Fedora Core 4 they have
622 set this on a number of new functions. No doubt there will be more in due
625 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
626 specify a transport. However, if an address that was verified by such a
627 router was the subject of a callout, Exim crashed because it tried to
628 read the rcpt_include_affixes from the non-existent transport. Now it
629 just assumes that the setting of that option is false. This bug was
630 introduced by 4.51/PH/31.
632 PH/28 Changed -d+all to exclude +memory, because that information is very
633 rarely of interest, but it makes the output a lot bigger. People tend to
634 do -d+all out of habit.
636 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
637 code in os-type was giving problems when libc.so lives in lib64, like on
640 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
641 aren't the modern standard, and it seems that some systems' include files
642 don't always have them. Exim was already checking for some of the newer
643 ones like T_AAAA, and defining it itself. I've added checks for all the
644 record types that Exim uses.
646 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
647 not automatically generating a new one, as it is supposed to. This
648 prevented TLS from working. If the file did exist, but contained invalid
649 data, a new version was generated, as expected. It was only the case of a
650 non-existent file that was broken.
652 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
653 with a change in libdomainkeys > 0.64.
655 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
656 from DNS. If the selector record carries the flag, it now has
657 precedence over the domain-wide flag.
659 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
661 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
662 the use of a port name that isn't defined in /etc/services) occurred, the
663 message was deferred as in a normal delivery, and thus remained on the
664 spool, instead of being failed because of the mua_wrapper setting. This
665 is now fixed, and I tidied up some of the mua_wrapper messages at the
668 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
669 the messages in a hash of arrays rather than using individual hashes.
670 This is a bit cleaner and results in dramatic memory savings, albeit
671 at a slight CPU cost.
673 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
674 as requested by Marc Sherman.
676 SC/10 Eximstats: added histograms for user specified patterns as requested
679 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
681 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
682 fopen() in the content-scanning modules that did not already have it.
688 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
689 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
691 TK/02 Fix ACL "control" statement not being available in MIME ACL.
693 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
695 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
696 to test Sieve filters that use "vacation".
698 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
699 that changes the way the GnuTLS parameters are stored in the cache file.
700 The new format can be generated externally. For backward compatibility,
701 if the data in the cache doesn't make sense, Exim assumes it has read an
702 old-format file, and it generates new data and writes a new file. This
703 means that you can't go back to an older release without removing the
706 PH/03 A redirect router that has both "unseen" and "one_time" set does not
707 work if there are any delivery delays because "one_time" forces the
708 parent to be marked "delivered", so its unseen clone is never tried
709 again. For this reason, Exim now forbids the simultaneous setting of
712 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
713 redirected to themselves ("homonym" addresses). Read the long ChangeLog
714 entry if you want to know the details. The fix, however, neglected to
715 consider the case when local delivery batching is involved. The test for
716 "previously delivered" was not happening when checking to see if an
717 address could be batched with a previous (undelivered) one; under
718 certain circumstances this could lead to multiple deliveries to the same
721 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
722 in its include files, and this causes problems building Exim.
724 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
725 header_syntax) but Exim was just ignoring anything given after a slash.
726 In particular, this caused confusion with an attempt to use "verify =
727 reverse_host_lookup/defer_ok". An error is now given when options are
728 supplied for verify items that do not have them. (Maybe reverse_host_
729 lookup should have a defer_ok option, but that's a different point.)
731 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
732 defined by RFC 821) to 2048, because there were problems with some AUTH
733 commands, and RFC 1869 says the size should be increased for extended
734 SMTP commands that take arguments.
736 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
739 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
740 "unknown" error; now it says that the functionality isn't in the binary.
742 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
743 an address' error message when a string expansion fails (syntax or
744 whatever). Otherwise the password may appear in the log. Following change
745 PH/42 below, there is no longer a chance of it appearing in a bounce
748 PH/11 Installed exipick version 20050225.0 from John Jetmore.
750 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
751 of its addresses was ever tried. (Bugzilla bug #2.)
753 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
754 the result incorrectly in the debug output. (It correctly added a newline
755 to what was transported.)
757 TF/01 Added $received_time.
759 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
760 commented out examples of how to interface to a virus scanner and to
761 SpamAssassin. Also added commented examples of av_scanner and
762 spamd_address settings.
764 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
765 and controls are allowed in which ACLs. There were a couple of minor
766 errors. Some of the entries in the conditions table (which is a table of
767 where they are NOT allowed) were getting very unwieldy; rewrote them as a
768 negation of where the condition IS allowed.
770 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
772 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
773 header file does not have a version number, so I've had to invent a new
774 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
775 API. The code is untested by me (my Linux distribution still has 0.3.2 of
776 radiusclient), but it was contributed by a Radius user.
778 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
779 files or directories, for OpenSSL.
781 PH/19 When an Exim process that is running as root has to create an Exim log
782 file, it does so in a subprocess that runs as exim:exim so as to get the
783 ownership right at creation (otherwise, other Exim processes might see
784 the file with the wrong ownership). There was no test for failure of this
785 fork() call, which would lead to the process getting stuck as it waited
786 for a non-existent subprocess. Forks do occasionally fail when resources
787 run out. I reviewed all the other calls to fork(); they all seem to check
790 PH/20 When checking for unexpected SMTP input at connect time (before writing
791 the banner), Exim was not dealing correctly with a non-positive return
792 from the read() function. If the client had disconnected by this time,
793 the result was a log entry for a synchronization error with an empty
794 string after "input=" when read() returned zero. If read() returned -1
795 (an event I could not check), uninitialized data bytes were printed.
796 There were reports of junk text (parts of files, etc) appearing after
799 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
801 PH/22 Added support for macro redefinition, and (re)definition in between
802 driver and ACL definitions.
804 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
805 forgetting to use the resulting value; it was using the unexpanded value.
807 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
808 hadn't been configured. The fix is from Juergen Kreileder, who
809 understands it better than I do:
811 "Here's what I see happening with three configured cyrus_sasl
812 authenticators configured (plain, login, cram-md5):
814 On startup auth_cyrus_sasl_init() gets called for each of these.
815 This means three calls to sasl_listmech() without a specified mech_list.
816 => SASL tests which mechs of all available mechs actually work
817 => three warnings about OTP not working
818 => the returned list contains: plain, login, cram-md5, digest-md5, ...
820 With the patch, sasl_listmech() also gets called three times. But now
821 SASL's mech_list option is set to the server_mech specified in the the
822 authenticator. Or in other words, the answer from sasl_listmech()
823 gets limited to just the mech you're testing for (which is different
825 => the return list contains just 'plain' or 'login', 'cram-md5' or
826 nothing depending on the value of ob->server_mech.
828 I've just tested the patch: Authentication still works fine,
829 unavailable mechs specified in the exim configuration are still
830 caught, and the auth.log warnings about OTP are gone."
832 PH/25 When debugging is enabled, the contents of the command line are added
833 to the debugging output, even when log_selector=+arguments is not
836 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
837 answer is "GNU", and only if the return is "GNU/something" is the answer
840 PH/27 $acl_verify_message is now set immediately after the failure of a
841 verification in an ACL, and so is available in subsequent modifiers. In
842 particular, the message can be preserved by coding like this:
844 warn !verify = sender
845 set acl_m0 = $acl_verify_message
847 Previously, $acl_verify_message was set only while expanding "message"
848 and "log_message" when a very denied access.
850 PH/28 Modified OS/os.c-Linux with
852 -#ifndef OS_LOAD_AVERAGE
853 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
855 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
856 nomenclature these days.)
858 PH/29 Installed patch from the Sieve maintainer that adds the options
859 sieve_useraddress and sieve_subaddress to the redirect router.
861 PH/30 In these circumstances:
862 . Two addresses routed to the same list of hosts;
863 . First host does not offer TLS;
864 . First host accepts first address;
865 . First host gives temporary error to second address;
866 . Second host offers TLS and a TLS session is established;
867 . Second host accepts second address.
868 Exim incorrectly logged both deliveries with the TLS parameters (cipher
869 and peerdn, if requested) that were in fact used only for the second
872 PH/31 When doing a callout as part of verifying an address, Exim was not paying
873 attention to any local part prefix or suffix that was matched by the
874 router that accepted the address. It now behaves in the same way as it
875 does for delivery: the affixes are removed from the local part unless
876 rcpt_include_affixes is set on the transport.
878 PH/32 Add the sender address, as F=<...>, to the log line when logging a
879 timeout during the DATA phase of an incoming message.
881 PH/33 Sieve envelope tests were broken for match types other than :is. I have
882 applied a patch sanctioned by the Sieve maintainer.
884 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
885 the uid or gid is negative. A case of a negative gid caused this to be
886 noticed. The fix allows for either to be negative.
888 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
889 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
892 PH/36 The change PH/12 above was broken. Fixed it.
894 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
895 the grounds that routing the same address twice would always produce the
896 same answer. This might have been true once, but it is certainly no
897 longer true now. Routing a child address may depend on the previous
898 routing that produced that child. Some complicated redirection strategies
899 went wrong when messages had multiple recipients, and made Exim's
900 behaviour dependent on the order in which the addresses were given.
902 I have moved the duplicate checking until after the routing is complete.
903 Exim scans the addresses that are assigned to local and remote
904 transports, and removes any duplicates. This means that more work will be
905 done, as duplicates will always all be routed, but duplicates are
906 presumably rare, so I don't expect this is of any significance.
908 For deliveries to pipes, files, and autoreplies, the duplicate checking
909 still happens during the routing process, since they are not going to be
912 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
913 It corrects a timeout issue with spamd. This is Ian's comment: "The
914 background is that sometimes spamd either never reads data from a
915 connection it has accepted, or it never writes response data. The exiscan
916 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
917 blindly assumes that writes won't block so it may never time out."
919 PH/39 Allow G after quota size as well as K and M.
921 PH/40 The value set for $authenticated_id in an authenticator may not contain
922 binary zeroes or newlines because the value is written to log lines and
923 to spool files. There was no check on this. Now the value is run through
924 the string_printing() function so that such characters are converted to
925 printable escape sequences.
927 PH/41 $message_linecount is a new variable that contains the total number of
928 lines in the message. Compare $body_linecount, which is the count for the
931 PH/42 Exim no longer gives details of delivery errors for specific addresses in
932 bounce and delay warning messages, except in certain special cases, which
935 (a) An SMTP error message from a remote host;
936 (b) A message specified in a :fail: redirection;
937 (c) A message specified in a "fail" command in a system filter;
938 (d) A message specified in a FAIL return from the queryprogram router;
939 (e) A message specified by the cannot_route_message router option.
941 In these cases only, Exim does include the error details in bounce and
942 warning messages. There are also a few cases where bland messages such
943 as "unrouteable address" or "local delivery error" are given.
945 PH/43 $value is now also set for the "else" part of a ${run expansion.
947 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
948 being worked on, but at least Exim now implements the latest version to
951 PH/45 In a pipe transport, although a timeout while waiting for the pipe
952 process to complete was treated as a delivery failure, a timeout while
953 writing the message to the pipe was logged, but erroneously treated as a
954 successful delivery. Such timeouts include transport filter timeouts. For
955 consistency with the overall process timeout, these timeouts are now
956 treated as errors, giving rise to delivery failures by default. However,
957 there is now a new Boolean option for the pipe transport called
958 timeout_defer, which, if set TRUE, converts the failures into defers for
959 both kinds of timeout. A transport filter timeout is now identified in
962 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
963 systems where "make" and "gmake" are different, calling "gmake" at top
964 level broke things. I've arranged for the value of $(MAKE) to be passed
965 from the Makefile to this script so that it can call the same version of
969 A note about Exim versions 4.44 and 4.50
970 ----------------------------------------
972 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
973 changes of various kinds. As a consequence, a big documentation update was
974 needed. This delayed the release for rather longer than seemed good, especially
975 in the light of a couple of (minor) security issues. Therefore, the changes
976 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
977 release. So 4.44 and 4.50 are in effect two different branches that both start
980 I have left the 4.50 change log unchanged; it contains all the changes since
981 4.43. The change log for 4.44 is below; many of its items are identical to
982 those for 4.50. This seems to be the most sensible way to preserve the
983 historical information.
989 1. Minor wording change to the doc/README.SIEVE file.
991 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
992 computation of the current number of files was incorrect.
994 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
995 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
998 4. Give more explanation in the error message when the command for a transport
999 filter fails to execute.
1001 5. There are several places where Exim runs a non-Exim command in a
1002 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1003 was being done only for the command run by the queryprogram router. It is
1004 now done for all such subprocesses. The other cases are: ${run, transport
1005 filters, and the commands run by the lmtp and pipe transports.
1007 6. Added CONFIGURE_GROUP build-time option.
1009 7. Some older OS have a limit of 256 on the maximum number of file
1010 descriptors. Exim was using setrlimit() to set 1000 as a large value
1011 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1012 systems. I've change it so that if it can't get 1000, it tries for 256.
1014 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1015 was an oversight, and furthermore, ever since the addition of extra
1016 controls (e.g. 4.43/32), the checks on when to allow different forms of
1017 "control" were broken. There should now be diagnostics for all cases when a
1018 control that does not make sense is encountered.
1020 9. Added the /retain_sender option to "control=submission".
1022 10. $recipients is now available in the predata ACL (oversight).
1024 11. Tidy the search cache before the fork to do a delivery from a message
1025 received from the command line. Otherwise the child will trigger a lookup
1026 failure and thereby defer the delivery if it tries to use (for example) a
1027 cached ldap connection that the parent has called unbind on.
1029 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1030 of $address_data from the recipient verification was clobbered by the
1031 sender verification.
1033 13. The value of address_data from a sender verification is now available in
1034 $sender_address_data in subsequent conditions in the ACL statement.
1036 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1038 15. Added a new option "connect=<time>" to callout options, to set a different
1041 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1042 was its contents. (It was OK if the option was not defined at all.)
1044 17. A "Completed" log line is now written for messages that are removed from
1045 the spool by the -Mrm option.
1047 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1048 information about exactly what failed.
1050 19. Added -dd to debug only the daemon process.
1052 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1053 handles timeouts, both on the server side and network timeouts. Renamed the
1054 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1056 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1057 It is now set to "smtps".
1059 22. $host_address is now set to the target address during the checking of
1060 ignore_target_hosts.
1062 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1063 being passed; this would have caused $sender_host_name to have been used if
1064 matching the list had actually called for a host name (not very likely,
1065 since this list is usually IP addresses). A host name is now passed as
1068 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1069 code that specifies a non-restarting handler (typically sigaction() in
1070 modern systems) in an attempt to fix a rare and obscure crash bug.
1072 25. Narrowed the window for a race in the daemon that could cause it to ignore
1073 SIGCHLD signals. This is not a major problem, because they are used only to
1074 wake it up if nothing else does.
1076 26. A malformed maildirsize file could cause Exim to calculate negative values
1077 for the mailbox size or file count. Odd effects could occur as a result.
1078 The maildirsize information is now recalculated if the size or filecount
1081 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1082 support for a long time. Removed HAVE_SYS_VFS_H.
1084 28. Installed the latest version of exipick from John Jetmore.
1086 29. In an address list, if the pattern was not a regular expression, an empty
1087 subject address (from a bounce message) matched only if the pattern was an
1088 empty string. Non-empty patterns were not even tested. This was the wrong
1089 because it is perfectly reasonable to use an empty address as part of a
1090 database query. An empty address is now tested by patterns that are
1091 lookups. However, all the other forms of pattern expect the subject to
1092 contain a local part and a domain, and therefore, for them, an empty
1093 address still always fails if the pattern is not itself empty.
1095 30. Exim went into a mad DNS loop when attempting to do a callout where the
1096 host was specified on an smtp transport, and looking it up yielded more
1097 than one IP address.
1099 31. Re-factored the code for checking spool and log partition space into a
1100 function that finds that data and another that does the check. The former
1101 is then used to implement four new variables: $spool_space, $log_space,
1102 $spool_inodes, and $log_inodes.
1104 32. The RFC2047 encoding function was originally intended for short strings
1105 such as real names; it was not keeping to the 75-character limit for
1106 encoded words that the RFC imposes. It now respects the limit, and
1107 generates multiple encoded words if necessary. To be on the safe side, I
1108 have increased the buffer size for the ${rfc2047: expansion operator from
1111 33. It is now permitted to omit both strings after an "if" condition; if the
1112 condition is true, the result is "true". As before, when the second string
1113 is omitted, a false condition yields an empty string. This makes it less
1114 cumbersome to write custom ACL and router conditions.
1116 34. Failure to deliver a bounce message always caused it to be frozen, even if
1117 there was an errors_to setting on the router. The errors_to setting is now
1120 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1121 canonical form (fully expanded) before being placed in
1122 $sender_host_address.
1124 36. The table in the code that translates DNS record types into text (T_A to
1125 "A" for instance) was missing entries for NS and CNAME. It is just possible
1126 that this could have caused confusion if both these types were looked up
1127 for the same domain, because the text type is used as part of Exim's
1128 per-process caching. But the chance of anyone hitting this buglet seems
1131 37. The dnsdb lookup has been extended in a number of ways.
1133 (1) There is a new type, "zns", which walks up the domain tree until it
1134 finds some nameserver records. It should be used with care.
1136 (2) There is a new type, "mxh", which is like "mx" except that it returns
1137 just the host names, not the priorities.
1139 (3) It is now possible to give a list of domains (or IP addresses) to be
1140 looked up. The behaviour when one of the lookups defers can be
1141 controlled by a keyword.
1143 (4) It is now possible to specify the separator character for use when
1144 multiple records are returned.
1146 38. The dnslists ACL condition has been extended: it is now possible to supply
1147 a list of IP addresses and/or domains to be looked up in a particular DNS
1150 39. Added log_selector=+queue_time_overall.
1152 40. When running the queue in the test harness, wait just a tad after forking a
1153 delivery process, to get repeatability of debugging output.
1155 41. Include certificate and key file names in error message when GnuTLS fails
1156 to set them up, because the GnuTLS error message doesn't include the name
1157 of the failing file when there is a problem reading it.
1159 42. Allow both -bf and -bF in the same test run.
1161 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1163 44. The "Exiscan patch" is now merged into the mainline Exim source.
1165 45. Sometimes the final signoff response after QUIT could fail to get
1166 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1167 < 0 before doing a fflush(). This bug looks as though it goes back to the
1168 introduction of TLS in release 3.20, but "sometimes" must have been rare
1169 because the tests only now provoked it.
1171 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1172 (this can affect the format of dates).
1174 47. exim_tidydb, when checking for the continued existence of a message for
1175 which it has found a message-specific retry record, was not finding
1176 messages that were in split spool directories. Consequently, it was
1177 deleting retry records that should have stayed in existence.
1179 48. Steve fixed some bugs in eximstats.
1181 49. The SPA authentication driver was not abandoning authentication and moving
1182 on to the next authenticator when an expansion was forced to fail,
1183 contradicting the general specification for all authenticators. Instead it
1184 was generating a temporary error. It now behaves as specified.
1186 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1187 (the order specifies the preference for clients). The order is now AES256,
1188 AES128, 3DES, ARCFOUR128.
1190 51. Small patch to Sieve code - explicitly set From: when generating an
1193 52. Exim crashed if a remote delivery caused a very long error message to be
1194 recorded - for instance if somebody sent an entire SpamAssassin report back
1195 as a large number of 550 error lines. This bug was coincidentally fixed by
1196 increasing the size of one of Exim's internal buffers (big_buffer) that
1197 happened as part of the Exiscan merge. However, to be on the safe side, I
1198 have made the code more robust (and fixed the comments that describe what
1201 53. Now that there can be additional text after "Completed" in log lines (if
1202 the queue_time_overall log selector is set), a one-byte patch to exigrep
1203 was needed to allow it to recognize "Completed" as not the last thing in
1206 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1207 patch that reportedly fixes this has been added. I am not expert enough to
1208 create a test for it. This is what the patch creator wrote:
1210 "I found a little strange behaviour of ldap code when working with
1211 Windows 2003 AD Domain, where users was placed in more than one
1212 Organization Units. When I tried to give exim partial DN, the exit code
1213 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1214 But simultaneously result of request was absolutely normal ldap result,
1215 so I produce this patch..."
1217 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1218 REFERENCE, so I have modified the code to exclude the patch when that macro
1221 55. Some experimental protocols are using DNS PTR records for new purposes. The
1222 keys for these records are domain names, not reversed IP addresses. The
1223 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1224 leaves it alone. Component reversal etc. now happens only for IP addresses.
1227 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1229 57. Double the size of the debug message buffer (to 2048) so that more of very
1230 long debug lines gets shown.
1232 58. The exicyclog utility now does better if the number of log files to keep
1233 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1235 59. Two changes related to the smtp_active_hostname option:
1237 (1) $smtp_active_hostname is now available as a variable.
1238 (2) The default for smtp_banner uses $smtp_active_hostname instead
1239 of $primary_hostname.
1241 60. The host_aton() function is supposed to be passed a string that is known
1242 to be a valid IP address. However, in the case of IPv6 addresses, it was
1243 not checking this. This is a hostage to fortune. Exim now panics and dies
1244 if the condition is not met. A case was found where this could be provoked
1245 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1246 components; fortuitously, this particular loophole had already been fixed
1247 by change 4.50/55 above.
1249 If there are any other similar loopholes, the new check in host_aton()
1250 itself should stop them being exploited. The report I received stated that
1251 data on the command line could provoke the exploit when Exim was running as
1252 exim, but did not say which command line option was involved. All I could
1253 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1254 running as the user.
1257 61. There was a buffer overflow vulnerability in the SPA authentication code
1258 (which came originally from the Samba project). I have added a test to the
1259 spa_base64_to_bits() function which I hope fixes it.
1262 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1263 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1265 63. The daemon start-up calls getloadavg() while still root for those OS that
1266 need the first call to be done as root, but it missed one case: when
1267 deliver_queue_load_max is set with deliver_drop_privilege. This is
1268 necessary for the benefit of the queue runner, because there is no re-exec
1269 when deliver_drop_privilege is set.
1271 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1272 This has been fixed.
1274 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1275 was in use, was not putting the data itself into the right store pool;
1276 consequently, it could be overwritten for a subsequent message in the same
1277 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1280 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1282 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1283 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1284 the function were treating the return as a boolean value, which happened to
1285 work because 0=false and not-0=true, but is not correct code.
1287 68. The host_aton() function was not handling scoped IPv6 addresses (those
1288 with, for example, "%eth0" on the end) correctly.
1290 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1291 negated items (that is, ~something) in unsigned ints. Some compilers
1292 apparently mutter when there is no cast.
1294 70. If an address verification called from an ACL failed, and did not produce a
1295 user-specific message (i.e. there was only a "system" message), nothing was
1296 put in $acl_verify_message. In this situation, it now puts the system
1299 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1300 session; change 4.31/43 added the unwanted input to the log line - except
1301 that it did not do this in the start of session case. It now does.
1303 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1304 This is wrong and can cause the other end to generate a synchronization
1305 error if it is another Exim or anything else that does the synchronization
1306 check. A QUIT command is no longer sent after a timeout.
1308 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1309 during host lookups.
1311 74. The defer_ok option of callout verification was not working if it was used
1312 when verifying addresses in header lines, that is, for this case:
1314 verify = header_sender/callout=defer_ok
1316 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1317 those file descriptors could be used for SMTP connections. If anything
1318 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1319 could be sent to the SMTP client, causing chaos. The daemon now opens
1320 stdin, stdout, and stderr to /dev/null when it puts itself into the
1323 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1324 log by default. The user can override this with suitable Perl magic.
1326 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1327 the log message when discard triggers, was not working for the DATA ACL or
1328 for the non-SMTP ACL.
1330 78. Error message wording change in sieve.c.
1332 79. If smtp_accept_max_per_host was set, the number of connections could be
1333 restricted to fewer than expected, because the daemon was trying to set up
1334 a new connection before checking whether the processes handling previous
1335 connections had finished. The check for completed processes is now done
1336 earlier. On busy systems, this bug wouldn't be noticed because something
1337 else would have woken the daemon, and it would have reaped the completed
1340 80. If a message was submitted locally by a user whose login name contained one
1341 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1342 It caused a spool format error. I have fixed the spool reading code. A
1343 related problem was that the "from" clause in the Received: line became
1344 illegal because of the space(s). It is now covered by ${quote_local_part.
1346 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1349 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1351 83. Patch from Sieve maintainer to fix unterminated string problem in
1352 "vacation" handling.
1354 84. Some minor changes to the Linux configuration files to help with other
1355 OS variants using glibc.
1357 85. One more patch for Sieve to update vacation handling to latest spec.
1360 ----------------------------------------------------
1361 See the note above about the 4.44 and 4.50 releases.
1362 ----------------------------------------------------
1368 1. Change 4.43/35 introduced a bug that caused file counts to be
1369 incorrectly computed when quota_filecount was set in an appendfile
1372 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1373 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1376 3. Give more explanation in the error message when the command for a transport
1377 filter fails to execute.
1379 4. There are several places where Exim runs a non-Exim command in a
1380 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1381 was being done only for the command run by the queryprogram router. It is
1382 now done for all such subprocesses. The other cases are: ${run, transport
1383 filters, and the commands run by the lmtp and pipe transports.
1385 5. Some older OS have a limit of 256 on the maximum number of file
1386 descriptors. Exim was using setrlimit() to set 1000 as a large value
1387 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1388 systems. I've change it so that if it can't get 1000, it tries for 256.
1390 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1391 was an oversight, and furthermore, ever since the addition of extra
1392 controls (e.g. 4.43/32), the checks on when to allow different forms of
1393 "control" were broken. There should now be diagnostics for all cases when a
1394 control that does not make sense is encountered.
1396 7. $recipients is now available in the predata ACL (oversight).
1398 8. Tidy the search cache before the fork to do a delivery from a message
1399 received from the command line. Otherwise the child will trigger a lookup
1400 failure and thereby defer the delivery if it tries to use (for example) a
1401 cached ldap connection that the parent has called unbind on.
1403 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1404 of $address_data from the recipient verification was clobbered by the
1405 sender verification.
1407 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1408 was its contents. (It was OK if the option was not defined at all.)
1410 11. A "Completed" log line is now written for messages that are removed from
1411 the spool by the -Mrm option.
1413 12. $host_address is now set to the target address during the checking of
1414 ignore_target_hosts.
1416 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1417 being passed; this would have caused $sender_host_name to have been used if
1418 matching the list had actually called for a host name (not very likely,
1419 since this list is usually IP addresses). A host name is now passed as
1422 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1423 code that specifies a non-restarting handler (typically sigaction() in
1424 modern systems) in an attempt to fix a rare and obscure crash bug.
1426 15. Narrowed the window for a race in the daemon that could cause it to ignore
1427 SIGCHLD signals. This is not a major problem, because they are used only to
1428 wake it up if nothing else does.
1430 16. A malformed maildirsize file could cause Exim to calculate negative values
1431 for the mailbox size or file count. Odd effects could occur as a result.
1432 The maildirsize information is now recalculated if the size or filecount
1435 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1436 support for a long time. Removed HAVE_SYS_VFS_H.
1438 18. Updated exipick to current release from John Jetmore.
1440 19. Allow an empty sender to be matched against a lookup in an address list.
1441 Previously the only cases considered were a regular expression, or an
1444 20. Exim went into a mad DNS lookup loop when doing a callout where the
1445 host was specified on the transport, if the DNS lookup yielded more than
1448 21. The RFC2047 encoding function was originally intended for short strings
1449 such as real names; it was not keeping to the 75-character limit for
1450 encoded words that the RFC imposes. It now respects the limit, and
1451 generates multiple encoded words if necessary. To be on the safe side, I
1452 have increased the buffer size for the ${rfc2047: expansion operator from
1455 22. Failure to deliver a bounce message always caused it to be frozen, even if
1456 there was an errors_to setting on the router. The errors_to setting is now
1459 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1460 canonical form (fully expanded) before being placed in
1461 $sender_host_address.
1463 24. Updated eximstats to version 1.33
1465 25. Include certificate and key file names in error message when GnuTLS fails
1466 to set them up, because the GnuTLS error message doesn't include the name
1467 of the failing file when there is a problem reading it.
1469 26. Expand error message when OpenSSL has problems setting up cert/key files.
1472 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1473 (this can affect the format of dates).
1475 28. exim_tidydb, when checking for the continued existence of a message for
1476 which it has found a message-specific retry record, was not finding
1477 messages that were in split spool directories. Consequently, it was
1478 deleting retry records that should have stayed in existence.
1480 29. eximstats updated to version 1.35
1481 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1482 1.35 - bugfix such that pie charts by volume are generated correctly
1484 30. The SPA authentication driver was not abandoning authentication and moving
1485 on to the next authenticator when an expansion was forced to fail,
1486 contradicting the general specification for all authenticators. Instead it
1487 was generating a temporary error. It now behaves as specified.
1489 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1490 (the order specifies the preference for clients). The order is now AES256,
1491 AES128, 3DES, ARCFOUR128.
1493 31. Small patch to Sieve code - explicitly set From: when generating an
1496 32. Exim crashed if a remote delivery caused a very long error message to be
1497 recorded - for instance if somebody sent an entire SpamAssassin report back
1498 as a large number of 550 error lines. This bug was coincidentally fixed by
1499 increasing the size of one of Exim's internal buffers (big_buffer) that
1500 happened as part of the Exiscan merge. However, to be on the safe side, I
1501 have made the code more robust (and fixed the comments that describe what
1504 33. Some experimental protocols are using DNS PTR records for new purposes. The
1505 keys for these records are domain names, not reversed IP addresses. The
1506 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1507 leaves it alone. Component reversal etc. now happens only for IP addresses.
1510 34. The host_aton() function is supposed to be passed a string that is known
1511 to be a valid IP address. However, in the case of IPv6 addresses, it was
1512 not checking this. This is a hostage to fortune. Exim now panics and dies
1513 if the condition is not met. A case was found where this could be provoked
1514 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1515 components; fortuitously, this particular loophole had already been fixed
1516 by change 4.50/55 or 4.44/33 above.
1518 If there are any other similar loopholes, the new check in host_aton()
1519 itself should stop them being exploited. The report I received stated that
1520 data on the command line could provoke the exploit when Exim was running as
1521 exim, but did not say which command line option was involved. All I could
1522 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1523 running as the user.
1526 35. There was a buffer overflow vulnerability in the SPA authentication code
1527 (which came originally from the Samba project). I have added a test to the
1528 spa_base64_to_bits() function which I hope fixes it.
1531 36. The daemon start-up calls getloadavg() while still root for those OS that
1532 need the first call to be done as root, but it missed one case: when
1533 deliver_queue_load_max is set with deliver_drop_privilege. This is
1534 necessary for the benefit of the queue runner, because there is no re-exec
1535 when deliver_drop_privilege is set.
1537 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1538 was in use, was not putting the data itself into the right store pool;
1539 consequently, it could be overwritten for a subsequent message in the same
1540 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1543 38. Sometimes the final signoff response after QUIT could fail to get
1544 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1545 < 0 before doing a fflush(). This bug looks as though it goes back to the
1546 introduction of TLS in release 3.20, but "sometimes" must have been rare
1547 because the tests only now provoked it.
1553 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1554 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1555 Now it returns an integer. A number of calls were still expecting a T/F
1556 return. Fortuitously, in all cases, the tests worked in OK situations,
1557 which is the norm. However, things would have gone wrong on any write
1558 failures on the smtp file descriptor. This function is used when sending
1559 messages over SMTP and also when doing verify callouts.
1561 2. When Exim is called to do synchronous delivery of a locally submitted
1562 message (the -odf or -odi options), it no longer closes stderr before doing
1565 3. Implemented the mua_wrapper option.
1567 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1569 5. Implemented the functions header_remove(), header_testname(),
1570 header_add_at_position(), and receive_remove_recipient(), and exported them
1573 6. If an ACL "warn" statement specified the addition of headers, Exim already
1574 inserted X-ACL-Warn: at the start if there was no header name. However, it
1575 was not making this test for the second and subsequent header lines if
1576 there were newlines in the string. This meant that an invalid header could
1577 be inserted if Exim was badly configured.
1579 7. Allow an ACL "warn" statement to add header lines at the start or after all
1580 the Received: headers, as well as at the end.
1582 8. Added the rcpt_4xx retry error code.
1584 9. Added postmaster_mailfrom=xxx to callout verification option.
1586 10. Added mailfrom=xxxx to the callout verification option, for verify=
1589 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1590 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1592 12. Inserted some casts to stop certain compilers warning when using pointer
1593 differences as field lengths or precisions in printf-type calls (mostly
1594 affecting debugging statements).
1596 13. Added optional readline() support for -be (dynamically loaded).
1598 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1599 same clock tick as a message's arrival, so that its received time was the
1600 same as the "first fail" time on the retry record, and that message
1601 remained on the queue past the ultimate address timeout, every queue runner
1602 would try a delivery (because it was past the ultimate address timeout) but
1603 after another failure, the ultimate address timeout, which should have then
1604 bounced the address, did not kick in. This was a "< instead of <=" error;
1605 in most cases the first failure would have been in the next clock tick
1606 after the received time, and all would be well.
1608 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1609 being recognized when the domain list was tested by the match_domain
1610 condition in an expansion string.
1612 16. Added the ${str2b64: operator.
1614 17. Exim was always calling setrlimit() to set a large limit for the number of
1615 processes, without checking whether the existing limit was already
1616 adequate. (It did check for the limit on file descriptors.) Furthermore,
1617 errors from getrlimit() and setrlimit() were being ignored. Now they are
1618 logged to the main and panic logs, but Exim does carry on, to try to do its
1619 job under whatever limits there are.
1621 18. Imported PCRE 5.0.
1623 19. Trivial typo in log message " temporarily refused connection" (the leading
1626 20. If the log selector return_path_on_delivery was set and an address was
1627 redirected to /dev/null, the delivery process crashed because it assumed
1628 that a return path would always be set for a "successful" delivery. In this
1629 case, the whole delivery is bypassed as an optimization, and therefore no
1632 21. Internal re-arrangement: the function for sending a challenge and reading
1633 a response while authentication was assuming a zero-terminated challenge
1634 string. It's now changed to take a pointer and a length, to allow for
1635 binary data in such strings.
1637 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1639 23. Exim was not respecting finduser_retries when seeking the login of the
1640 uid under which it was called; it was always trying 10 times. (The default
1641 setting of finduser_retries is zero.) Also, it was sleeping after the final
1642 failure, which is pointless.
1644 24. Implemented tls_on_connect_ports.
1646 25. Implemented acl_smtp_predata.
1648 26. If the domain in control=submission is set empty, Exim assumes that the
1649 authenticated id is a complete email address when it generates From: or
1650 Sender: header lines.
1652 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1653 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1654 chown and chgrp in /bin and hostname in /usr/bin.
1656 28. Exim was keeping the "process log" file open after each use, just as it
1657 does for the main log. This opens the possibility of it remaining open for
1658 long periods when the USR1 signal hits a daemon. Occasional processlog
1659 errors were reported, that could have been caused by this. Anyway, it seems
1660 much more sensible not to leave this file open at all, so that is what now
1663 29. The long-running daemon process does not normally write to the log once it
1664 has entered its main loop, and it closes the log before doing so. This is
1665 so that log files can straightforwardly be renamed and moved. However,
1666 there are a couple of unusual error situations where the daemon does write
1667 log entries, and I had neglected to close the log afterwards.
1669 30. The text of an SMTP error response that was received during a remote
1670 delivery was being truncated at 512 bytes. This is too short for some of
1671 the long messages that one sometimes sees. I've increased the limit to
1674 31. It is now possible to make retry rules that apply only when a message has a
1675 specific sender, in particular, an empty sender.
1677 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1678 it possible to be selective about when SMTP synchronization is enforced.
1680 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1682 32. Implemented hosts_connection_nolog.
1684 33. Added an ACL for QUIT.
1686 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1689 35. Added mailbox_size and mailbox_filecount to appendfile.
1691 36. Added control = no_multiline_responses to ACLs.
1693 37. There was a bug in the logic of the code that waits for the clock to tick
1694 in the case where the clock went backwards by a substantial amount such
1695 that the microsecond fraction of "now" was more than the microsecond
1696 fraction of "then" (but the whole seconds number was less).
1698 38. Added support for the libradius Radius client library this is found on
1699 FreeBSD (previously only the radiusclient library was supported).
1705 1. When certain lookups returned multiple values in the form name=value, the
1706 quoting of the values was not always being done properly. Specifically:
1707 (a) If the value started with a double quote, but contained no whitespace,
1709 (b) If the value contained whitespace other than a space character (i.e.
1710 tabs or newlines or carriage returns) it was not quoted.
1711 This fix has been applied to the mysql and pgsql lookups by writing a
1712 separate quoting function and calling it from the lookup code. The fix
1713 should probably also be applied to nisplus, ibase and oracle lookups, but
1714 since I cannot test any of those, I have not disturbed their existing code.
1716 2. A hit in the callout cache for a specific address caused a log line with no
1717 reason for rejecting RCPT. Now it says "Previous (cached) callout
1718 verification failure".
1720 3. There was an off-by-one bug in the queryprogram router. An over-long
1721 return line was truncated at 256 instead of 255 characters, thereby
1722 overflowing its buffer with the terminating zero. As well as fixing this, I
1723 have increased the buffer size to 1024 (and made a note to document this).
1725 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1726 when Exim is waiting for an SMTP response from a remote server, Exim
1727 restarts its select() call on the socket, thereby resetting its timeout.
1728 This is not a problem when such interrupts are rare. Somebody set up a cron
1729 job to run exiwhat every 2 minutes, which is less than the normal select()
1730 timeout (5 or 10 minutes). This meant that the select() timeout never
1731 kicked in because it was always reset. I have fixed this by comparing the
1732 time when an interrupt arrives with the time at the start of the first call
1733 to select(). If more time than the timeout has elapsed, the interrupt is
1734 treated as a timeout.
1736 5. Some internal re-factoring in preparation for the addition of Sieve
1737 extensions (by MH). In particular, the "personal" test is moved to a
1738 separate function, and given an option for scanning Cc: and Bcc: (which is
1739 not set for Exim filters).
1741 6. When Exim created an email address using the login of the caller as the
1742 local part (e.g. when creating a From: or Sender: header line), it was not
1743 quoting the local part when it contained special characters such as @.
1745 7. Installed new OpenBSD configuration files.
1747 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1748 try to make them clearer.
1750 9. Callout options, other than the timeout value, were being ignored when
1751 verifying sender addresses in header lines. For example, when using
1753 verify = header_sender/callout=no_cache
1755 the cache was (incorrectly) being used.
1757 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1758 only the Cygwin environment.
1760 11. When return_path_on_delivery was set as a log selector, if different remote
1761 addresses in the same message used different return paths and parallel
1762 remote delivery occurred, the wrong values would sometimes be logged.
1763 (Whenever a remote delivery process finished, the return path value from
1764 the most recently started remote delivery process was logged.)
1766 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1767 lines when AUTH and/or TLS are in use. This is the "received protocol"
1768 field. Exim used to use "asmtp" for authenticated SMTP, without any
1769 indication (in the protocol name) for TLS use. Now it follows the RFC and
1770 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1771 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1772 names appear in log lines as well as in Received: header lines.
1774 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1775 extensions, and comparison tests, and to fix some bugs.
1777 14. Changes to the "personal" filter test:
1779 (1) The test was buggy in that it was just doing the equivalent of
1780 "contains" tests on header lines. For example, if a user's address was
1781 anne@some.where, the "personal" test would incorrectly be true for
1783 To: susanne@some.where
1785 This test is now done by extracting each address from the header in turn,
1786 and checking the entire address. Other tests that are part of "personal"
1787 are now done using regular expressions (for example, to check local parts
1788 of addresses in From: header lines).
1790 (2) The list of non-personal local parts in From: addresses has been
1791 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1792 taken from the Sieve specification recommendations.
1794 (3) If the message contains any header line starting with "List-" it is
1795 treated as non-personal.
1797 (4) The test for "circular" in the Subject: header line has been removed
1798 because it now seems ill-conceived.
1800 15. Minor typos in src/EDITME comments corrected.
1802 16. Installed latest exipick from John Jetmore.
1804 17. If headers_add on a router specified a text string that was too long for
1805 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1806 of string_sprintf() is now avoided.
1808 18. $message_body_size was not set (it was always zero) when running the DATA
1809 ACL and the local_scan() function.
1811 19. For the "mail" command in an Exim filter, no default was being set for
1812 the once_repeat time, causing a random time value to be used if "once" was
1813 specified. (If the value happened to be <= 0, no repeat happened.) The
1814 default is now 0s, meaning "never repeat". The "vacation" command was OK
1815 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1816 (I found it when inspecting the code).
1818 20. There is now an overall timeout for performing a callout verification. It
1819 defaults to 4 times the callout timeout, which applies to individual SMTP
1820 commands during the callout. The overall timeout applies when there is more
1821 than one host that can be tried. The timeout is checked before trying the
1822 next host. This prevents very long delays if there are a large number of
1823 hosts and all are timing out (e.g. when the network connections are timing
1824 out). The value of the overall timeout can be changed by specifying an
1825 additional sub-option for "callout", called "maxwait". For example:
1827 verify = sender/callout=5s,maxwait=20s
1829 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1830 to the end before writing, but this should make it even safer).
1832 22. Exim was forgetting that it had advertised PIPELINING for the second and
1833 subsequent messages on an SMTP connection. It was also not resetting its
1834 memory on STARTTLS and an internal HELO.
1836 23. When Exim logs an SMTP synchronization error within a session, it now
1837 records whether PIPELINING has been advertised or not.
1839 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1840 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1841 rather than long int.
1843 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1845 26. Added the never_mail option to autoreply.
1851 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1852 crash if the getsockname() call failed; this can happen if a connection is
1853 closed very soon after it is established. The problem was simply in the
1854 order in which certain operations were done, causing Exim to try to write
1855 to the SMTP stream before it had set up the file descriptor. The bug has
1856 been fixed by making things happen in the correct order.
1862 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1863 before the connection was closed, thus losing the rejection response.
1865 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1866 some early Solaris releases, but causes trouble in current releases where
1867 socklen_t is defined.
1869 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1872 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1875 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1876 that releasing the top of it at the end releases what was used for sub-
1877 expansions (unless the block got too big). However, discard this block if
1878 the first thing is a variable or header, so that we can use its block when
1879 it is dynamic (useful for very large $message_headers, for example).
1881 6. Lookups now cache *every* query, not just the most recent. A new, separate
1882 store pool is used for this. It can be recovered when all lookup caches are
1883 flushed. Lookups now release memory at the end of their result strings.
1884 This has involved some general refactoring of the lookup sources.
1886 7. Some code has been added to the store_xxx() functions to reduce the amount
1887 of flapping under certain conditions.
1889 8. log_incoming_interface used to affect only the <= reception log lines. Now
1890 it causes the local interface and port to be added to several more SMTP log
1891 lines, for example "SMTP connection from", and rejection lines.
1893 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1895 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1897 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1898 could be overwritten at the end of the current message (or the start of a
1899 new message if it was set in a HELO ACL). The value is now preserved for
1900 the duration of the SMTP connection.
1902 12. If a transport had a headers_rewrite setting, and a matching header line
1903 contained an unqualified address, that address was qualified, even if it
1904 did not match any rewriting rules. The underlying bug was that the values
1905 of the flags that permit the existence of unqualified sender and recipient
1906 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1907 non-local messages, and by -bnq for local messages) were not being
1908 preserved with the message after it was received.
1910 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1911 "next input=" as part of the text comprising the host identity instead of
1912 the correct text. The code was using the same buffer for two different
1913 strings. However, depending on which order the printing function evaluated
1914 its arguments, the bug did not always show up. Under Linux, for example, my
1915 test suite worked just fine.
1917 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1918 doesn't work with some older versions of Perl. It has been changed to "my",
1919 which in any case is probably the better facility to use.
1921 15. A really picky compiler found some instances of statements for creating
1922 error messages that either had too many or two few arguments for the format
1925 16. The size of the buffer for calls to the DNS resolver has been increased
1926 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1927 for addresses that have a lot of PTR records. This alleviates a problem; it
1928 does not fully solve it.
1930 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1931 buffer now truncates the list and logs the incident, which is the same
1932 action as happens when Exim is looking up a host name and its aliases.
1933 Previously in this situation something unpredictable would happen;
1934 sometimes it was "internal error: store_reset failed".
1936 18. If a server dropped the connection unexpectedly when an Exim client was
1937 using GnuTLS and trying to read a response, the client delivery process
1938 crashed while trying to generate an error log message.
1940 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1941 string, for example:
1943 warn message = H1: something\nH2: something
1945 the text was added as a single header line from Exim's point of view
1946 though it ended up OK in the delivered message. However, searching for the
1947 second and subsequent header lines using $h_h2: did not work. This has been
1948 fixed. Similarly, if a system filter added multiple headers in this way,
1949 the routers could not see them.
1951 20. Expanded the error message when iplsearch is called with an invalid key to
1952 suggest using net-iplsearch in a host list.
1954 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1955 ACLs are no longer actually imposed (and a message to that effect is
1958 22. If a "gecos" field in a passwd entry contained escaped characters, in
1959 particular, if it contained a \" sequence, Exim got it wrong when building
1960 a From: or a Sender: header from that name. A second bug also caused
1961 incorrect handling when an unquoted " was present following a character
1962 that needed quoting.
1964 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1965 was not being matched caselessly.
1967 24. Arranged for all hyphens in the exim.8 source to be escaped with
1970 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1971 properly. Recipient callout cache records were still being keyed to include
1972 the sender, even when use_sender was set false. This led to far more
1973 callouts that were necessary. The sender is no longer included in the key
1974 when use_sender is false.
1976 26. Added "control = submission" modifier to ACLs.
1978 27. Added the ${base62d: operator to decode base 62 numbers.
1980 28. dnsdb lookups can now access SRV records.
1982 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1983 the configuration file.
1985 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1986 (-v) mode. This makes the output for a verbose queue run more intelligible.
1988 31. Added a use_postmaster feature to recipient callouts.
1990 32. Added the $body_zerocount variable, containing the number of binary zero
1991 bytes in the message body.
1993 33. The time of last modification of the "new" subdirectory is now used as the
1994 "mailbox time last read" when there is a quota error for a maildir
1997 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1999 35. Added +ignore_unknown as a special item in host lists.
2001 36. Code for decoding IPv6 addresses in host lists is now included, even if
2002 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2003 address was recognized as an IP address, but was then not correctly decoded
2004 into binary, causing unexpected and incorrect effects when compared with
2011 1. Very minor rewording of debugging text in manualroute to say "list of
2012 hosts" instead of "hostlist".
2014 2. If verify=header_syntax was set, and a header line with an unqualified
2015 address (no domain) and a large number of spaces between the end of the
2016 name and the colon was received, the reception process suffered a buffer
2017 overflow, and (when I tested it) crashed. This was caused by some obsolete
2018 code that should have been removed. The fix is to remove it!
2020 3. When running in the test harness, delay a bit after writing a bounce
2021 message to get a bit more predictability in the log output.
2023 4. Added a call to search_tidyup() just before forking a reception process. In
2024 theory, someone could use a lookup in the expansion of smtp_accept_max_
2025 per_host which, without the tidyup, could leave open a database connection.
2027 5. Added the variables $recipient_data and $sender_data which get set from a
2028 lookup success in an ACL "recipients" or "senders" condition, or a router
2029 "senders" option, similar to $domain_data and $local_part_data.
2031 6. Moved the writing of debug_print from before to after the "senders" test
2034 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2035 problems for message scanning, either using a data ACL, or using
2036 local_scan() because the Received: header was not generated till after they
2037 were called (in order to set the time as the time of reception completion).
2038 I have revised the way this works. The header is now generated after the
2039 body is received, but before the ACL or local_scan() are called. After they
2040 are run, the timestamp in the header is updated.
2046 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2047 before starting a queue runner without re-exec. This happened only when
2048 deliver_drop_privilege was set or when the Exim user was set to root. The
2049 effect of the bug was that timeouts during subsequent deliveries caused
2050 crashes instead of being properly handled. The handler is now left at its
2051 default (and expected) setting.
2053 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2054 message, again when deliver_drop_privilege is set or Exim is run as root.
2055 The bug described in (1) was not present in this case, but the tidying up
2056 of the other signals was missing. I have made the two cases consistent.
2058 3. The ignore_target_hosts setting on a manualroute router was being ignored
2059 for hosts that were looked up using the /MX notation.
2061 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2064 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2065 operated on the sender address. After changing the $sender_address to <>
2066 for the sender address verify, Exim was re-instated it as the original
2067 (before rewriting) address, but remembering that it had rewritten it, so it
2068 wasn't rewriting it again. This bug also had the effect of breaking the
2069 sender address verification caching when the sender address was rewritten.
2071 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2072 This has been changed so that if the ip literal address matches
2073 ignore_target_hosts, the router declines.
2075 7. Added expansion conditions match_domain, match_address, and match_local_
2076 part (NOT match_host).
2078 8. The placeholder for the Received: header didn't have a length field set.
2080 9. Added code to Exim itself and to exim_lock to test for a specific race
2081 condition that could lead to file corruption when using MBX delivery. The
2082 issue is with the lockfile that is created in /tmp. If this file is removed
2083 after a process has opened it but before that process has acquired a lock,
2084 there is the potential for a second process to recreate the file and also
2085 acquire a lock. This could lead to two Exim processes writing to the file
2086 at the same time. The added code performs the same test as UW imapd; it
2087 checks after acquiring the lock that its file descriptor still refers to
2088 the same named file.
2090 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2091 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2094 11. Added the smtp_active_hostname option. If used, this will typically be made
2095 to depend on the incoming interface address. Because $interface_address is
2096 not set up until the daemon has forked a reception process, error responses
2097 that can happen earlier (such as "too many connections") no longer contain
2100 12. If an expansion in a condition on a "warn" statement fails because a lookup
2101 defers, the "warn" statement is abandoned, and the next ACL statement is
2102 processed. Previously this caused the whole ACL to be aborted.
2104 13. Added the iplsearch lookup type.
2106 14. Added ident_timeout as a log selector.
2108 15. Added tls_certificate_verified as a log selector.
2110 16. Added a global option tls_require_ciphers (compare the smtp transport
2111 option of the same name). This controls incoming TLS connections.
2113 17. I finally figured out how to make tls_require_ciphers do a similar thing
2114 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2115 before starting the TLS session.
2117 18. Tabs are now shown as \t in -bP output.
2119 19. If the log selector return_path_on_delivery was set, Exim crashed when
2120 bouncing a message because it had too many Received: header lines.
2122 20. If two routers both had headers_remove settings, and the first one included
2123 a superfluous trailing colon, the final name in the first list and the
2124 first name in the second list were incorrectly joined into one item (with a
2125 colon in the middle).
2131 1. Added -C and -D options to the exinext utility, mainly to make it easier
2132 to include in the automated testing, but these could be helpful when
2133 multiple configurations are in use.
2135 2. The exinext utility was not formatting the output nicely when there was
2136 an alternate port involved in the retry record key, nor when there was a
2137 message id as well (for retries that were specific to a specific message
2138 and a specific host). It was also confused by IPv6 addresses, because of
2139 the additional colons they contain. I have fixed the IPv4 problem, and
2140 patched it up to do a reasonable job for IPv6.
2142 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2143 delivery, the log line now contains "pipelined" if PIPELINING was used.
2145 4. An SMTP transport process used to panic and die if the bind() call to set
2146 an explicit outgoing interface failed. This has been changed; it is now
2147 treated in the same way as a connect() failure.
2149 5. A reference to $sender_host_name in the part of a conditional expansion
2150 that was being skipped was still causing a DNS lookup. This no longer
2153 6. The def: expansion condition was not recognizing references to header lines
2154 that used bh_ and bheader_.
2156 7. Added the _cache feature to named lists.
2158 8. The code for checking quota_filecount in the appendfile transport was
2159 allowing one more file than it should have been.
2161 9. For compatibility with Sendmail, the command line option
2169 and sets the incoming protocol and host name (for trusted callers). The
2170 host name and its colon can be omitted when only the protocol is to be set.
2171 Note the Exim already has two private options, -pd and -ps, that refer to
2172 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2173 "s", but I don't think that's a major issue.
2175 10. A number of refactoring changes to the code, none of which should affect
2178 (a) The number of logging options was getting close to filling up the
2179 32-bit word that was used as a bit map. I have split them into two classes:
2180 those that are passed in the argument to log_write(), and those that are
2181 only ever tested independently outside of that function. These are now in
2182 separate 32-bit words, so there is plenty of room for expansion again.
2183 There is no change in the user interface or the logging behaviour.
2185 (b) When building, for example, log lines, the code previously used a
2186 macro that called string_cat() twice, in order to add two strings. This is
2187 not really sufficiently general. Furthermore, there was one instance where
2188 it was actually wrong because one of the argument was used twice, and in
2189 one call a function was used. (As it happened, calling the function twice
2190 did not affect the overall behaviour.) The macro has been replaced by a
2191 function that can join an arbitrary number of extra strings onto a growing
2194 (c) The code for expansion conditions now uses a table and a binary chop
2195 instead of a serial search (which was left over from when there were very
2196 few conditions). Also, it now recognizes conditions like "pam" even when
2197 the relevant support is not compiled in: a suitably worded error message is
2198 given if an attempt is made to use such a condition.
2200 11. Added ${time_interval:xxxxx}.
2202 12. A bug was causing one of the ddress fields not to be passed back correctly
2203 from remote delivery subprocesses. The field in question was not being
2204 subsequently used, so this caused to problems in practice.
2206 13. Added new log selectors queue_time and deliver_time.
2208 14. Might have fixed a bug in maildirsizefile handling that threw up
2209 "unexpected character" debug warnings, and recalculated the data
2210 unnecessarily. In any case, I expanded the warning message to give more
2213 15. Added the message "Restricted characters in address" to the statements in
2214 the default ACL that block characters like @ and % in local parts.
2216 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2217 Three changes have been made:
2219 (a) There was a serious bug; a negative response to MAIL caused the whole
2220 recipient domain to be cached as invalid, thereby blocking all messages
2221 to all local parts at the same domain, from all senders. This bug has
2222 been fixed. The domain is no longer cached after a negative response to
2223 MAIL if the sender used is not empty.
2225 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2228 (c) A new callout option, "use_sender" has been added for people who want
2229 the modified behaviour.
2235 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2238 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2239 indeed breaks things for older releases.
2241 3. Added additional logging to the case where there is a problem reading data
2242 from a filter that is running in a subprocess using a pipe, in order to
2243 try to track down a specific problem.
2245 4. Testing facility fudge: when running in the test harness and attempting
2246 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2247 getting "No route to host". Convert this to a timeout.
2249 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2252 6. Some OS don't have socklen_t but use size_t instead. This affects the
2253 fifth argument of getsockopt() amongst other things. This is now
2254 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2255 can be set for individual OS. I have set it for SunOS5, OSF1, and
2256 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2257 some earlier ones do not.
2259 7. Change 4.30/15 was not doing the test caselessly.
2261 8. The standard form for an IPv6 address literal was being rejected by address
2262 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2263 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2264 this, as well as the form without the "IPv6" on the front (but only when
2265 address literals are enabled, of course).
2267 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2269 10. Exim crashed if a message with an empty sender address specified by -f
2270 encountered a router with an errors_to setting. This could be provoked only
2271 by a command such as
2275 where an empty string was supplied; "<>" did not hit this bug.
2277 11. Installed PCRE release 4.5.
2279 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2280 remained set. It is now erased.
2282 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2283 times from message ids (which are base 36 rather than the normal 62).
2285 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2286 were being counted as actual protocol errors, and logged if the log
2287 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2288 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2289 and DATA following a set of rejected RCPTs do not count as protocol errors.
2290 In other words, Exim assumes they were pipelined, though this may not
2291 actually be the case. Of course, in all cases the client gets an
2292 appropriate error code.
2294 15. If a lookup fails in an ACL condition, a message about the failure may
2295 be available; it is used if testing the ACL cannot continue, because most
2296 such messages specify what the cause of the deferral is. However, some
2297 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2298 that caused an old message to be retained and used if a later statement
2299 caused a defer, replacing the real cause of the deferral.
2301 16. If an IP address had so many PTR records that the DNS lookup buffer
2302 was not large enough to hold them, Exim could crash while trying to process
2303 the truncated data. It now detects and logs this case.
2305 17. Further to 4.21/58, another change has been made: if (and only if) the
2306 first line of a message (the first header line) ends with CRLF, a bare LF
2307 in a subsequent header line has a space inserted after it, so as not to
2308 terminate the header.
2310 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2311 unnecessarily, used atoi() instead of strtol(), and didn't check the
2312 termination when getting file sizes from file names by regex.
2314 19. Completely re-implemented the support for maildirsize files, in the light
2315 of a number of problems with the previous contributed implementation
2316 (4.30/29). In particular:
2318 . If the quota is zero, the maildirsize file is maintained, but no quota is
2321 . If the maildir directory does not exist, it is created before any attempt
2322 to write a maildirsize file.
2324 . The quota value in the file is just a cache; if the quota is changed in
2325 the transport, the new value overrides.
2327 . A regular expression is available for excluding directories from the
2330 20. The autoreply transport checks the characters in options that define the
2331 message's headers; it allows continued headers, but it was checking with
2332 isspace() after an embedded newline instead of explicitly looking for a
2335 21. If all the "regular" hosts to which an address was routed had passed their
2336 expiry times, and had not reached their retry times, the address was
2337 bounced, even if fallback hosts were defined. Now Exim should go on to try
2340 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2341 equivalent code in the SMTP transport. Some hosts send humungous responses
2342 to HELO/EHLO, more than 1024 it seems.
2344 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2345 gives compiler warnings in some environments. I've now done it "properly",
2348 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2349 (because of problems with the built-in one) was declared to return uschar *
2350 instead of char *, causing compiler failure.
2352 25. Fixed a file descriptor leak when processing alias/forward files.
2354 26. Fixed a minor format string issue in dbfn.c.
2356 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2358 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2359 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2362 29. When a sender address is verified, it is cached, to save repeating the test
2363 when there is more than one recipient in a message. However, when the
2364 verification involves a callout, it is possible for different callout
2365 options to be set for different recipients. It is too complicated to keep
2366 track of this in the cache, so now Exim always runs a verification when a
2367 callout is required, relying on the callout cache for the optimization.
2368 The overhead is duplication of the address routing, but this should not be
2371 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2372 to be verified with callout=postmaster, and the main callout worked but the
2373 postmaster check failed, the verification correctly failed. However, if a
2374 subsequent RCPT command asked for sender verification *without* the
2375 postmaster check, incorrect caching caused this verification also to fail,
2378 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2379 it was not caching the DNS options (qualify_single, search_parents) that
2380 were used when the lookup failed. A subsequent lookup with different
2381 options therefore always gave the same answer, though there were cases
2382 where it should not have. (Example: a "domains = !$mx_any" option on a
2383 dnslookup router: the "domains" option is always processed without any
2384 widening, but the router might have qualify_single set.) Now Exim uses the
2385 cached value only when the same options are set.
2387 32. Added John Jetmore's "exipick" utility to the distribution.
2389 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2390 than a timeout (e.g. a certificate is required, and is not provided), an
2391 Exim server now closes the connection immediately. Previously it waited for
2392 the client to close - but if the client is SSL, it seems that they each
2393 wait for each other, leading to a delay before one of them times out.
2395 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2396 maintained 0.8.x compatibility because I don't think many are using it, and
2397 it is clearly obsolete.
2399 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2402 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2403 client certificate was expired. A simple patch fixes this, though I don't
2404 understand the full logic of why the verify callback is called multiple
2407 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2408 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2409 which causes problems with some clients (such as the Certicom SSL Plus
2410 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2411 disables the coutermeasure allowing Eudora to connect."
2413 38. Exim was not checking that a write() to a log file succeeded. This could
2414 lead to Bad Things if a log got too big, in particular if it hit a file
2415 size limit. Exim now panics and dies if it cannot write to a log file, just
2416 as it does if it cannot open a log file.
2418 39. Modified OS/Makefile-Linux so that it now contains
2420 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2422 The two -D definitions ensure that Exim is compiled with large file
2423 support, which makes it possible to handle log files that are bigger than
2426 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2427 instance) a domain was checked against a named list that involved a lookup,
2428 causing $domain_data to be set, then another domain was checked against the
2429 same list, then the first domain was re-checked, the value of $domain_data
2430 after the final check could be wrong. In particular, if the second check
2431 failed, it could be set empty. This bug probably also applied to
2434 41. The strip_trailing_dot option was not being applied to the address given
2435 with the -f command-line option.
2437 42. The code for reading a message's header from the spool was incrementing
2438 $received_count, but never initializing it. This meant that the value was
2439 incorrect (doubled) while delivering a message in the same process in which
2440 it was received. In the most common configuration of Exim, this never
2441 happens - a fresh exec is done - but it can happen when
2442 deliver_drop_privilege is set.
2444 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2445 it now includes up to 150 characters of the unexpected data in the log
2448 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2449 and building data strings. The size of both of these buffers was 10 000
2450 bytes - far larger than anybody would *ever* want, thought I. Needless to
2451 say, somebody hit the limit. I have increased the maximum line length to
2452 20 000 and the maximum data length of concatenated lines to 100 000. I have
2453 also fixed two bugs, because there was no checking on these buffers. Tsk,
2454 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2455 buffer is too small.
2457 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2458 lsearch lookups. Now it does.
2460 46. When parsing a route_list item in a manualroute router, a fixed-length
2461 buffer was used for the list of hosts. I made this 1024 bytes long,
2462 thinking that nobody would ever have a list of hosts that long. Wrong.
2463 Somebody had a whole pile of complicated expansion conditions, and the
2464 string was silently truncated, leading to an expansion error. It turns out
2465 that it is easier to change to an unlimited length (owing to other changes
2466 that have happened since this code was originally written) than to build
2467 structure for giving a limitation error. The length of the item that
2468 expands into the list of hosts is now unlimited.
2470 47. The lsearch lookup could not handle data where the length of text line was
2471 more than 4095 characters. Such lines were truncated, leading to shortened
2472 data being returned. It should now handle lines of any length.
2474 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2475 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2478 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2479 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2480 no longer happens in comment lines. A list of replacements is now placed
2481 at the head of all of the source files, except those whose only change is
2482 to replace PERL_COMMAND in the very first #! line.
2484 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2485 messages on the queue, with a bottom-up merge sort, using code contributed
2486 by Michael Haardt. This should make operations like -bp somewhat faster on
2487 large queues. It won't affect queue runners, except when queue_run_in_order
2490 51. Installed eximstats 1.31 in the distribution.
2492 52. Added support for SRV lookups to the dnslookup router.
2494 53. If an ACL referred to $message_body or $message_body_end, the value was not
2495 reset for any messages that followed in the same SMTP session.
2497 54. The store-handling optimization for building very long strings was not
2498 differentiating between the different store pools. I don't think this
2499 actually made any difference in practice, but I've tidied it.
2501 55. While running the routers to verify a sender address, $sender_address
2502 was still set to the sender address. This is wrong, because when routing to
2503 send a bounce to the sender, it would be empty. Therefore, I have changed
2504 it so that, while verifying a sender address, $sender_address is set to <>.
2505 (There is no change to what happens when verifying a recipient address.)
2507 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2508 target A or AAAA records (if not already returned) without resetting the
2509 qualify_single or search_parents options of the DNS resolver. These are
2510 inappropriate in this case because the targets of MX and SRV records must
2511 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2512 target that, when qualified, matched something in the local domain. These
2513 two options are now turned off when doing these lookups.
2515 57. It seems that at least some releases of Reiserfs (which does not have the
2516 concept of a fixed number of inodes) returns zero and not -1 for the
2517 number of available inodes. This interacted badly with check_spool_inodes,
2518 which assumed that -1 was the "no such thing" setting. What I have done is
2519 to check that the total number of inodes is greater than zero before doing
2520 the test of how many are available.
2522 58. When a "warn" ACL statement has a log_message modifier, the message is
2523 remembered, and not repeated. This is to avoid a lot of repetition when a
2524 message has many recipients that cause the same warning to be written.
2525 Howewer, Exim was preserving the list of already written lines for an
2526 entire SMTP session, which doesn't seem right. The memory is now reset if a
2527 new message is started.
2529 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2530 debugging output unless log_rewrite was also set.
2532 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2533 of (int)(handle) when we know that handle contains (void *)(-1).
2535 61. The Exim daemon panic-logs an error return when it closes the incoming
2536 connection. However "connection reset by peer" seems to be common, and
2537 isn't really an error worthy of noting specially, so that particular error
2540 62. When Exim is trying to find all the local interfaces, it used to panic and
2541 die if the ioctl to get the interface flags failed. However, it seems that
2542 on at least one OS (Solaris 9) it is possible to have an interface that is
2543 included in the list of interfaces, but for which you get a failure error
2544 for this call. This happens when the interface is not "plumbed" into a
2545 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2546 failure of the "get flags" call assumes that the interface is down.
2548 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2549 makes life easier for people who are doing arithmetic on fields extracted
2550 from dates, where you often get leading zeros that should not be
2551 interpreted as octal.
2553 64. Added qualify_domain to the redirect router, to override the global
2556 65. If a pathologically long header line contained very many addresses (the
2557 report of this problem mentioned 10 000) and each of them was rewritten,
2558 Exim could use up a very large amount of memory. (It kept on making new
2559 copies of the header line as it rewrote, and never released the old ones.)
2560 At the expense of a bit more processing, the header rewriting function has
2561 been changed so that it no longer eats memory in this way.
2563 66. The generation of the Received: header has been moved from the time that a
2564 message starts to be received, to the time that it finishes. The timestamp
2565 in the Received: header should now be very close to that of the <= log
2566 line. There are two side-effects of this change:
2568 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2569 logged header lines no longer include the local Received: line, because
2570 it has not yet been created. The same applies to a copy of the message
2571 that is returned to a non-SMTP sender when a message is rejected.
2573 (b) When a filter file is tested using -bf, no additional Received: header
2574 is added to the test message. After some thought, I decided that this
2577 This change does not affect the value of $received_for. It is still set
2578 after address rewriting, but before local_scan() is called.
2580 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2582 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2583 gave an unhelpful panic error message, and a defer error. I have managed to
2584 change this behaviour so that it now rejects any supplied certificate,
2585 which seems right, as the list of acceptable certificates is empty.
2587 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2588 gave an unhelpful defer error. I have not managed to make this reject any
2589 supplied certificates, but the error message it gives is "no certificate
2590 supplied", which is not helpful.
2592 70. exigrep's output now also includes lines that are not associated with any
2593 message, but which match the given pattern. Implemented by a patch from
2594 Martin Sluka, which also tidied up the Perl a bit.
2596 71. Recipient callout verification, like sender verification, was using <> in
2597 the MAIL FROM command. This isn't really the right thing, since the actual
2598 sender may affect whether the remote host accepts the recipient or not. I
2599 have changed it to use the actual sender in the callout; this means that
2600 the cache record is now keyed on a recipient/sender pair, not just the
2601 recipient address. There doesn't seem to be a real danger of callout loops,
2602 since a callout by the remote host to check the sender would use <>.
2603 [SEE ABOVE: changed after hitting problems.]
2605 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2606 temporary errors. However, in the case of such a code being given after
2607 the end of a data transmission (i.e. after ".") Exim was failing to write
2608 a retry record for the message. (Yes, there was some broken host that was
2609 actually sending 8xx at this point.)
2611 73. An unknown lookup type in a host list could cause Exim to panic-die when
2612 the list was checked. (An example that provoked this was putting <; in the
2613 middle of a list instead of at the start.) If this happened during a DATA
2614 ACL check, a -D file could be left lying around. This kind of configuration
2615 error no longer causes Exim to die; instead it causes a defer errror. The
2616 incident is still logged to the main and panic logs.
2618 74. Buglet left over from Exim 3 conversion. The message "too many messages
2619 in one connection" was written to the rejectlog but not the mainlog, except
2620 when address rewriting (yes!) was being logged.
2622 75. Added write_rejectlog option.
2624 76. When a system filter was run not as root (that is, when system_filter_user
2625 was set), the values of the $n variables were not being returned to the
2626 main process; thus, they were not subsequently available in the $sn
2629 77. Added +return_path_on_delivery log selector.
2631 78. A connection timeout was being treated differently from recipients deferred
2632 when testing hosts_max_try with a message that was older than the host's
2633 retry timeout. (The host should not be counted, thus allowing all hosts to
2634 be tried at least once before bouncing.) This may have been the cause of an
2635 occasionally reported bug whereby a message would remain on the queue
2636 longer than the retry timeout, but would be bounced if a delivery was
2637 forced. I say "may" because I never totally pinned down the problem;
2638 setting up timeout/retry tests is difficult. See also the next item.
2640 79. The ultimate address timeout was not being applied to errors that involved
2641 a combination of host plus message (for example, a timeout on a MAIL
2642 command). When an address resolved to a number of possible hosts, and they
2643 were not all tried for each delivery (e.g. because of hosts_max_try), a
2644 message could remain on the queue longer than the retry timeout.
2646 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2649 81. Fixed an obscure SMTP outgoing bug which required at least the following
2650 conditions: (a) there was another message waiting for the same server;
2651 (b) the server returned 5xx to all RCPT commands in the first message so
2652 that the message was not completed; (c) the server dropped the connection
2653 or gave a negative response to the RSET that Exim sends to abort the
2654 transaction. The observed case was a dropped connection after DATA that had
2655 been sent in pipelining mode. That is, the server had advertised PIPELINING
2656 but was not implementing it correctly. The effect of the bug was incorrect
2657 behaviour, such as trying another host, and this could lead to a crash.
2663 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2664 and daemon.c were passed as pointers to ints; they should have been
2665 pointers to socklen_t variables (which are typically unsigned ints).
2667 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2670 3. Fixed a really odd bug that affected only the testing scheme; patching a
2671 certain fixed string in the binary changed the value of another string that
2672 happened to be identical to the end of the original first string.
2674 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2675 name", it returns that address as the IP address. On some operating
2676 systems (e.g. Solaris), it also passes back the IP address string as the
2677 "host name". However, on others (e.g. Linux), it passes back an empty
2678 string. Exim wasn't checking for this, and was changing the host name to an
2679 empty string, assuming it had been canonicized.
2681 5. Although rare, it is permitted to have more than one PTR record for a given
2682 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2683 all the names associated with an address, because they do in Solaris.
2684 However, it seems that they do not in Linux for data that comes from the
2685 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2686 I found this out when I moved to a new Linux workstation and tried to run
2687 the Exim test suite.
2689 To get round this problem I have changed the code so that it now does its
2690 own call to the DNS to look up PTR records when searching for a host name.
2691 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2692 addresses that are only in /etc/hosts are still found.
2694 This behaviour is, however, controlled by an option called host_lookup_
2695 order, which defaults to "bydns:byaddr". If people want to use the other
2696 order, or indeed, just use one or the other means of lookup, they can
2697 specify it in this variable.
2699 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2700 some operating systems, this comes back from gethostbyaddr() as an empty
2701 string, and this is what Exim used to test for. However, it seems that in
2702 other systems, "." is yielded. Exim now tests for this case too.
2704 7. The values of check_spool_space and check_log_space are now held internally
2705 as a number of kilobytes instead of an absolute number of bytes. If a
2706 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2707 kilobyte. This means that much larger values can be stored.
2709 8. Exim monitor: an attempt to get the action menu when not actually pointing
2710 at a message produces an empty menu entitled "No message selected". This
2711 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2712 no entries in it ("Shell widget menu has zero width and/or height"). So I
2713 have added a single, blank menu entry in this case.
2715 9. Added ${quote_local_part.
2717 10. MIME decoding is now applied to the contents of Subject: header lines when
2720 11. Now that a reference to $sender_host_address automatically causes a reverse
2721 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2722 host lookup before query-style lookups in lists that might use this
2723 variable. This has therefore been abolished, and the "net-" prefix is no
2724 longer necessary for query-style lookups.
2726 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2727 have been a typo for LFLAGS, so it has been changed.
2729 13. The install script calls Exim with "-C /dev/null" in order to find the
2730 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2731 to be output. Howeve, since Exim outputs its version number before the
2732 error, it didn't break the script. It just looked ugly. I fixed this by
2733 always allowing "-C /dev/null" if the caller is root.
2735 14. Ignore overlarge ACL variable number when reading spool file - insurance
2736 against a later release with more variables having written the file.
2738 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2739 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2740 this, as well as the form without the "IPv6" on the front.
2742 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2743 OS/Makefile-Darwin file.
2745 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2746 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2748 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2749 variable. However, some OS use ints for inodes. I've added cast to long int
2750 to get rid of the compiler warning.
2752 19. I had forgotten to lock out "/../" in configuration file names when
2753 ALT_CONFIG_PREFIX was set.
2755 20. Routers used for verification do not need to specify transports. However,
2756 if such a router generated a host list, and callout was configured, Exim
2757 crashed, because it could not find a port number from the (non-existent)
2758 transport. It now assumes port 25 in this circumstance.
2760 21. Added the -t option to exigrep.
2762 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2763 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2764 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2765 them is set, LOOKUP_LSEARCH is forced.
2767 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2769 24. Added sender and host information to the "rejected by local_scan()" log
2770 line; previously there was no indication of these.
2772 25. Added .include_if_exists.
2774 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2775 Linux. It turns out that not all file systems support this. Apparently some
2776 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2777 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2778 not supported on the file descriptor, is now ignored when Exim is trying to
2779 sync a directory. This applies only to Linux.
2781 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2783 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2784 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2787 29. Added support for maildirsize files from supplied patch (modified a bit).
2789 30. The use of :fail: followed by an empty string could lead Exim to respond to
2790 sender verification failures with (e.g.):
2792 550 Verification failed for <xxx>
2793 550 Sender verify failed
2795 where the first response line was missing the '-' that indicates it is not
2796 the final line of the response.
2798 31. The loop for finding the name of the user that called Exim had a hardwired
2799 limit of 10; it now uses the value of finduser_retries, which is used for
2800 all other user lookups.
2802 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2805 33. Exim was neglecting to zero errno before one call of strtol() when
2806 expanding a string and expecting an integer value. On some systems this
2807 resulted in spurious "integer overflow" errors. Also, it was casting the
2808 result into an int without checking.
2810 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2811 did not work. The code looks as if it has *never* worked, though it appears
2812 to have been documented since at least releast 1.62. I have made it work.
2814 35. The "timeout_DNS" error in retry rules, also documented since at least
2815 1.62, also never worked. As it isn't clear exactly what this means, and
2816 clearly it isn't a major issue, I have abolished the feature by treating it
2817 as "timeout", and writing a warning to the main and panic logs.
2819 36. The display of retry rules for -brt wasn't always showing the error code
2822 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2823 timeout_connect_A, timeout_connect_MX.
2825 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2826 to the empty sender.
2828 39. The daemon was not analysing the content of -oX till after it had closed
2829 stderr and disconnected from the controlling terminal. This meant that any
2830 syntax errors were only noted on the panic log, and the return code from
2831 the command was 0. By re-arranging the code a little, I've made the
2832 decoding happen first, so such errors now appear on stderr, and the return
2833 code is 1. However, the actual setting up of the sockets still happens in
2834 the disconnected process, so errors there are still only recorded on the
2837 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2838 connections (as happens on some IP stacks) was logged at start up time as
2839 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2840 it from "IPv6 and IPv4", which means that two separate sockets are being
2843 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2844 says whether AF_INET or AF_INET6 was passed as an argument.
2846 42. Exiwhat output was messed up when time zones were included in log
2849 43. Exiwhat now gives more information about the daemon's listening ports,
2850 and whether -tls-on-connect was used.
2852 44. The "port" option of the smtp transport is now expanded.
2854 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2855 silently ignored. Now an error message is written to the main and panic
2858 46. There's a new ACL modifier called "logwrite" which writes to a log file
2859 as soon as it is encountered.
2861 47. Added $local_user_uid and $local_user_gid at routing time.
2863 48. Exim crashed when trying to verify a sender address that was being
2866 49. Exim was recognizing only a space character after ".include". It now also
2867 recognizes a tab character.
2869 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2870 extracting the relevant information from the specification. The man page no
2871 longer contains scrambled data for the -d option, and I've added a section
2872 at the front about calling Exim under different names.
2874 51. Added "extra_headers" argument to the "mail" command in filter files.
2876 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2879 53. Installed eximstats 1.29.
2881 54. Added transport_filter_timeout as a generic transport option.
2883 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2884 Cc: header lines. This was required by RFC 822, but it not required by RFC
2887 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2888 incoming messages that did not have them. Now it does so only if the
2889 message originates locally, that is, if there is no associated remote host
2890 address. When Resent- header lines are present, this applies to the Resent-
2891 lines rather than the non-Resent- lines.
2893 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2894 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2896 58. Messages for configuration errors now include the name of the main
2897 configuration file - useful now that there may be more than one file in a
2898 list (.included file names were always shown).
2900 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2901 for those rare installations that do not start the daemon as root or run it
2902 setuid root. I've cut out the call to initgroups() if the daemon is not
2905 60. The Exim user and group can now be bound into the binary as text strings
2906 that are looked up at the start of Exim's processing.
2908 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2910 62. Added $mailstore_basename variable.
2912 63. Installed patch to sieve.c from Michael Haardt.
2914 64. When Exim failed to open the panic log after failing to open the main log,
2915 the original message it was trying to log was written to stderr and debug
2916 output, but if they were not available (the usual case in production), it
2917 was lost. Now it is written to syslog before the two lines that record the
2918 failures to open the logs.
2920 65. Users' Exim filters run in subprocesses under the user's uid. It is
2921 possible for a "deliver" command or an alias in a "personal" command to
2922 provoke an address rewrite. If logging of address rewriting is configured,
2923 this fails because the process is not running as root or exim. There may be
2924 a better way of dealing with this, but for the moment (because 4.30 needs
2925 to be released), I have disabled address rewrite logging when running a
2926 filter in a non-root, non-exim process.
2932 1. The buildconfig auxiliary program wasn't quoting the value set for
2933 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2934 not defined. This bug was masked in 4.22 by the effect that was fixed in
2937 2. Some messages that were rejected after a message id was allocated were
2938 shown as "incomplete" by exigrep. It no longer does this for messages that
2939 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2941 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2942 have allow_domain_literals set, the ID did not get logged in the <= line.
2943 Domain literals are now always recognized in Message-ID: header lines.
2945 4. The first argument for a ${extract expansion item is the key name or field
2946 number. Leading and trailing spaces in this item were not being ignored,
2947 causing some misleading effects.
2949 5. When deliver_drop_privilege was set, single queue runner processes started
2950 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2951 same command in the process it spins off) were not dropping privilege.
2953 6. When the daemon running as "exim" started a queue runner, it always
2954 re-executed Exim in the spun-off process. This is a waste of effort when
2955 deliver_drop_privilege is set. The new process now just calls the
2956 queue-runner function directly.
2962 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2965 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2966 ignored. Though the use of -oP was forcing the writing of a pid file, it
2967 was always written to the default place.
2969 3. If the message "no IP address found for host xxxx" is generated during
2970 incoming verification, it is now followed by identification of the incoming
2971 connection (so you can more easily find what provoked it).
2973 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2975 5. Added some features to "harden" Exim a bit more against certain attacks:
2977 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2978 be put in Local/Makefile. This is like the never_users runtime option,
2979 but it cannot be overridden. The default setting is "root".
2981 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2982 prefix string with which any file named in a -C command line option
2985 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2986 is retained for -C and -D only if the caller of Exim is root. Without
2987 it, the exim user may also use -C and -D and retain privilege.
2989 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2990 command line option is disabled.
2992 6. Macro names set by the -D option must start with an upper case letter, just
2993 like macro names defined in the configuration file.
2995 7. Added "dereference=" facility to LDAP.
2997 8. Two instances of the typo "uknown" in the source files are fixed.
2999 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3000 the Configure-Makefile script screwed up while processing it.
3002 10. Incorporated PCRE 4.4.
3004 11. The SMTP synchronization check was not operating right at the start of an
3005 SMTP session. For example, it could not catch a HELO sent before the client
3006 waited for the greeting. There is now a check for outstanding input at the
3007 point when the greeting is written. Because of the duplex, asynchronous
3008 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3009 way, but not yet received, when the check is performed.
3011 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3012 on TCP/IP sockets, because this apparently causes some broken clients to
3015 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3016 unchanged) from the Cygwin maintainer.
3018 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3019 was supported instead of testing for mbx. Effectively a typo.
3021 15. The spa authenticator server code was not checking that the input it
3022 received was valid base64.
3024 16. The debug output line for the "set" modifier in ACLs was not showing the
3025 name of the variable that was being set.
3027 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3029 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3030 lookup on its own. Something else was needed to trigger the lookup. For
3031 example, a match in host_lookup or the need for a host name in a host list.
3032 Now, if $sender_host_name is referenced and the host name has not yet been
3033 looked up, a lookup is performed. If the lookup fails, the variable remains
3034 empty, and $host_lookup_failed is set to "1".
3036 19. Added "eqi" as a case-independent comparison operator.
3038 20. The saslauthd authentication condition could segfault if neither service
3039 nor realm was specified.
3041 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3042 error message that was logged was misleading, and incoming SMTP
3043 connections were dropped. The message is now more accurate, and temporary
3044 errors are given to SMTP connections.
3046 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3047 (except RSET) with a 421 error, until QUIT is received. However, it was
3048 failing to send a response to QUIT.
3050 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3051 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3052 ACL is now run after the helo_try_verify_hosts code.
3054 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3055 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3056 case-independent, so other case variants are also recognized). Apparently
3057 some systems use these upper case variants.
3059 25. If more than two messages were waiting for the same host, and a transport
3060 filter was specified for the transport, Exim sent two messages over the
3061 same TCP/IP connection, and then failed with "socket operation on non-
3062 socket" when it tried to send the third.
3064 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3066 27. The extern definition of crypt16() in expand.c was not being excluded when
3067 the OS had its own crypt16() function.
3069 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3070 as a preferred synonym for return_size_limit, both as an option and as an
3073 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3075 30. Changed the default configuration ACL to relax the local part checking rule
3076 for addresses that are not in any local domains. For these addresses,
3077 slashes and pipe symbols are allowed within local parts, but the sequence
3078 /../ is explicitly forbidden.
3080 31. SPA server authentication was not clearing the challenge buffer before
3083 32. log_message in a "warn" ACL statement was writing to the reject log as
3084 well as to the main log, which contradicts the documentation and doesn't
3085 seem right (because no rejection is happening). So I have stopped it.
3087 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3088 However, I am unable to do any testing of this.
3090 34. Fixed an infelicity in the appendfile transport. When checking directories
3091 for a mailbox, to see if any needed to be created, it was accidentally
3092 using path names with one or more superfluous leading slashes; tracing
3093 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3095 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3096 contents are added to the log line that is written for every discarded
3097 recipient. (Previously a log_message setting was ignored.)
3099 36. The ${quote: operator now quotes the string if it is empty.
3101 37. The install script runs exim in order to find its version number. If for
3102 some reason other than non-existence or emptiness, which it checks, it
3103 could not run './exim', it was installing it with an empty version number,
3104 i.e. as "exim-". This error state is now caught, and the installation is
3107 38. An argument was missing from the function that creates an error message
3108 when Exim fails to connect to the socket for saslauthd authentication.
3109 This could cause Exim to crash, or give a corrupted message.
3111 39. Added isip, isip4, and isip6 to ${if conditions.
3113 40. The ACL variables $acl_xx are now saved with the message, and can be
3114 accessed later in routers, transports, and filters.
3116 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3117 strings in the file are not string-expanded.
3119 42. If a MAIL command specified a SIZE value that was too large to fit into an
3120 int variable, the check against message_size_limit failed. Such values are
3121 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3122 day this will have to be increased, but I don't think I want to be around
3123 when emails are that large.
3130 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3131 iconv() is not standard in FreeBSD.
3133 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3134 IPv6 enabled. The observed symptom was a segmentation fault on return from
3135 the function os_common_find_running_interfaces() in src/os.c.
3137 3. In the check_special_case() function in daemon.c I had used "errno" as an
3138 argument name, which causes warnings on some systems. This was basically a
3139 typo, since it was named "eno" in the comments!
3141 4. The code that waits for the clock to tick (at a resolution of some fraction
3142 of a second) so as to ensure message-id uniqueness was always waiting for
3143 at least one whole tick, when it could have waited for less. [This is
3144 almost certainly not relevant at current processor speeds, where it is
3145 unlikely to ever wait at all. But we try to future-proof.]
3147 5. The function that sleeps for a time interval that includes fractions of a
3148 second contained a race. It did not block SIGALRM between setting the
3149 timer, and suspending (a couple of lines later). If the interval was short
3150 and the sigsuspend() was delayed until after it had expired, the suspension
3151 never ended. On busy systems this could lead to processes getting stuck for
3154 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3155 process, before it forks any delivery processes. The open lookup caching
3156 mechanism meant that the open file or database connection was passed into
3157 the delivery process. The problem was that delivery processes always tidy
3158 up cached lookup data. This could cause a problem for the next delivery
3159 process started by the queue runner, because the external queue runner
3160 process does not know about the closure. So the next delivery process
3161 still has data in the lookup cache. In the case of a file lookup, there was
3162 no problem because closing a file descriptor in a subprocess doesn't affect
3163 the parent. However, if the lookup was caching a connection to a database,
3164 the connection was closed, and the second delivery process was likely to
3165 see errors such as "PGSQL: query failed: server closed the connection
3166 unexpectedly". The problem has been fixed by closing all cached lookups
3167 in a queue runner before running a delivery process.
3169 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3170 seem to have the "const" qualifier which it has on other OS. I've
3173 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3174 *of the same type* (client or server) with the same public name that an
3175 error should be diagnosed.
3177 9. When Exim looked up a host name for an IP address, but failed to find the
3178 original IP address when looking up the host name (a safety check), it
3179 output the message "<ip address> does not match any IP for NULL", which was
3180 confusing, to say the least. The bug was that the host name should have
3181 appeared instead of "NULL".
3183 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3184 user that is built into the binary, and the -C or -D options is used, root
3185 privilege is dropped before the configuration file is read. In addition,
3186 logging is switched to stderr instead of the normal log files. If the
3187 configuration then re-defines the Exim user, the unprivileged environment
3188 is probably not what is expected, so Exim logs a panic warning message (but
3191 However, if deliver_drop_privilege is set, the unprivileged state may well
3192 be exactly what is intended, so the warning has been cut out in that case,
3193 and Exim is allowed to try to write to its normal log files.
3199 1. smtp_return_error_details was not giving details for temporary sender
3200 or receiver verification errors.
3202 2. Diagnose a configuration error if two authenticators have the same public
3205 3. Exim used not to create the message log file for a message until the first
3206 delivery attempt. This could be confusing when incoming messages were held
3207 for policy or load reasons. The message log file is now created at the time
3208 the message is received, and an initial "Received" line is written to it.
3210 4. The automatically generated man page for command line options had a minor
3211 bug that caused no ill effects; however, a more serious problem was that
3212 the procedure for building the man page automatically didn't always
3213 operate. Consequently, release 4.20 contains an out-of-date version. This
3214 shouldn't happen again.
3216 5. When building Exim with embedded Perl support, the script that builds the
3217 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3218 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3220 6. The freeze_tell option was not being used for messages that were frozen on
3221 arrival, either by an ACL or by local_scan().
3223 7. Added the smtp_incomplete_transaction log selector.
3225 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3226 was accepting AUTH without a new EHLO.
3228 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3229 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3232 10. Logging of TCP/IP connections (when configured) now happens in the main
3233 daemon process instead of the child process, so that the TCP/IP connection
3234 count is more accurate (but it can never be perfect).
3236 11. The use of "drop" in a nested ACL was not being handled correctly in the
3237 outer ACL. Now, if condition failure induced by the nested "drop" causes
3238 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3239 or "require"), the connection is dropped.
3241 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3242 that yield "discard" can now be used with an "accept" or a "discard" verb,
3243 but an error is generated for any others (because I can't see a useful way
3244 to define what should happen).
3246 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3247 are now processed in the same way as lines in the Exim configuration file.
3248 In particular, continuation lines are supported.
3250 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3252 15. Added -ti meaning -t -i.
3254 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3255 domains, and warn by logging if others are found.
3257 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3258 returned from GIFCONF when Exim is trying to find the list of interfaces on
3259 a host. The code in os.c has been modified to copy each ifreq to an aligned
3260 structure in all cases.
3262 Also, in some cases, the returned ifreq's were being copied to a 'struct
3263 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3264 means the last couple of bytes of an IPv6 address could be chopped if the
3265 ifreq contained only a normal sockaddr (14 bytes storage).
3267 18. Named domain lists were not supported in the hosts_treat_as_local option.
3268 An entry such as +xxxx was not recognized, and was treated as a literal
3271 19. Ensure that header lines added by a DATA ACL are included in the reject log
3272 if the ACL subsequently rejects the message.
3274 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3275 MD5 (which is deprecated).
3277 21. When testing a filter file using -bf, Exim was writing a message when it
3278 took the sender from a "From " line in the message, but it was not doing so
3279 when it took $return_path from a Return-Path: header line. It now does.
3281 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3282 with a valid header line field name (a series of printing characters
3283 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3285 23. Changed "disc" in the source to "disk" to conform to the documentation and
3286 the book and for uniformity.
3288 24. Ignore Sendmail's -Ooption=value command line item.
3290 25. When execve() failed while trying to run a command in a pipe transport,
3291 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3292 could be confused with a return value of 69 from the command itself. This
3293 has been changed to 127, the value the shell returns if it is asked to run
3294 a non-existent command. The wording for the related log line suggests a
3295 non-existent command as the problem.
3297 26. If received_header_text expands to an empty string, do not add a Received:
3298 header line to the message. (Well, it adds a token one on the spool, but
3299 marks it "old" so that it doesn't get used or transmitted.)
3301 27. Installed eximstats 1.28 (addition of -nt option).
3303 28. There was no check for failure on the call to getsockname() in the daemon
3304 code. This can fail if there is a shortage of resources on the system, with
3305 ENOMEM, for example. A temporary error is now given on failure.
3307 29. Contrary to the C standard, it seems that in some environments, the
3308 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3309 program. Exim now does this explicitly; it affects the formatting of
3310 timestamps using strftime().
3312 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3313 complaints. I've now initialized all the variables, to avoid this.
3315 32. Header lines added by a system filter were not being "seen" during
3316 transport-time rewrites.
3318 33. The info_callback() function passed to OpenSSL is set up with type void
3319 (*)(SSL *, int, int), as described somewhere. However, when calling the
3320 function (actually a macro) that sets it up, the type void(*)() is
3321 expected. I've put in a cast to prevent warnings from picky compilers.
3323 34. If a DNS black list lookup found a CNAME record, but there were no A
3324 records associated with the domain it pointed at, Exim crashed.
3326 35. If a DNS black list lookup returned more than one A record, Exim ignored
3327 all but the first. It now scans all returned addresses if a particular IP
3328 value is being sought. In this situation, the contents of the
3329 $dnslist_value variable are a list of all the addresses, separated by a
3332 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3333 to accept a host name and all its aliases if the forward lookup for any of
3334 them yielded the IP address of the incoming connection. Now it accepts only
3335 those names whose forward lookup yields the correct IP address. Any other
3336 names are discarded. This closes a loophole whereby a rogue DNS
3337 administrator could create reverse DNS records to break through a
3338 wildcarded host restriction in an ACL.
3340 37. If a user filter or a system filter that ran in a subprocess used any of
3341 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3342 the wrong values were passed to the pipe command ($thisaddress had the
3343 value of $0, $0 had the value of $1, etc). This bug was introduced by
3344 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3346 38. Improved the line breaking for long SMTP error messages from ACLs.
3347 Previously, if there was no break point between 40 and 75 characters, Exim
3348 left the rest of the message alone. Two changes have been made: (a) I've
3349 reduced the minimum length to 35 characters; (b) if it can't find a break
3350 point between 35 and 75 characters, it looks ahead and uses the first one
3351 that it finds. This may give the occasional overlong line, but at least the
3352 remaining text gets split now.
3354 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3355 file descriptors might be low, and that setting 1000 would always raise it.
3356 It turns out that in some environments, the limit is already over 1000 and
3357 that lowering it causes trouble. So now Exim takes care not to decrease it.
3359 40. When delivering a message, the value of $return_path is set to $sender_
3360 address at the start of routing (routers may change the value). By an
3361 oversight, this default was not being set up when an address was tested by
3362 -bt or -bv, which affected the outcome if any router or filter referred to
3365 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3366 log only when "message" or "log_message" are set. However, if one of the
3367 conditions was an address verification, or a call to a nested ACL, the
3368 messages generated by the underlying test were being passed through. This
3369 no longer happens. The underlying message is available in $acl_verify_
3370 message for both "message" and "log_message" expansions, so it can be
3371 passed through if needed.
3373 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3374 new expansion $bh_ to give the encoded byte string without charset
3375 translation. Translation happens only if iconv() is available; HAVE_ICONV
3376 indicates this at build time. HEADERS_CHARSET gives the charset to
3377 translate to; headers_charset can change it in the configuration, and
3378 "headers charset" can change it in an individual filter file.
3380 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3381 that creates RFC 2047 encoded "words" labels them as that charset instead
3382 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3383 expansion operator; (ii) when Exim creates a From: line for a local
3384 message; (iii) when a header line is rewritten to include a "phrase" part.
3386 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3387 buggy, causing it to skip the first lines of messages whose message ID
3388 ended in 'D'. This would not have bitten before Exim release 4.14, because
3389 message IDs were unlikely to end in 'D' before then. The effect was to have
3390 incorrect size information for certain domains.
3392 45. #include "config.h" was missing at the start of the crypt16.c module. This
3393 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3396 46. If there was a timeout during a "random" callout check, Exim treated it as
3397 a failure of the random address, and carried on sending RSET and the real
3398 address. If the delay was just some slowness somewhere, the response to the
3399 original RCPT would be taken as a response to RSET and so on, causing
3400 mayhem of various kinds.
3402 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3403 when I implemented it. It didn't allow for the fact that some option values
3404 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3405 the right test for positive values.
3407 48. Domain names in DNS records are case-independent. Exim always looks them up
3408 in lower case. Some resolvers return domain names in exactly the case they
3409 appear in the zone file, that is, they may contain uppercase letters. Not
3410 all resolvers do this - some return always lower case. Exim was treating a
3411 change of case by a resolver as a change of domain, similar to a widening
3412 of a domain abbreviation. This triggered its re-routing code and so it was
3413 trying to route what was effectively the same domain again. This normally
3414 caused routing to fail (because the router wouldn't handle the domain
3415 twice). Now Exim checks for this case specially, and just changes the
3416 casing of the domain that it ultimately uses when it transmits the message
3419 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3422 50. If a filter generated a file delivery with a non-absolute name (possible if
3423 no home directory exists for the router), the forbid_file option was not
3426 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3427 the existing equality matching.
3429 52. Exim was using ints instead of ino_t variables in some places where it was
3430 dealing with inode numbers.
3432 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3433 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3434 called TMPDIR, and if it finds it is different, it changes its value.
3436 54. The smtp_printf() function is now made available to local_scan() so
3437 additional output lines can be written before returning. There is also an
3438 smtp_fflush() function to enable the detection of a dropped connection.
3439 The variables smtp_input and smtp_batched_input are exported to
3442 55. Changed the default runtime configuration: the message "Unknown user"
3443 has been removed from the ACL, and instead placed on the localuser router,
3444 using the cannot_route_message feature. This means that any verification
3445 failures that generate their own messages won't get overridden. Similarly,
3446 the "Unrouteable address" message that was in the ACL for unverifiable
3447 relay addresses has also been removed.
3449 56. Added hosts_avoid_esmtp to the smtp transport.
3451 57. The exicyclog script was not checking for the esoteric option
3452 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3453 will work only if exicyclog is run under the appropriate euid.
3455 58. Following a discussion on the list, the rules by which Exim recognises line
3456 endings on incoming messages have been changed. The -dropcr and drop_cr
3457 options are now no-ops, retained only for backwards compatibility. The
3458 following line terminators are recognized: LF CRLF CR. However, special
3459 processing applies to CR:
3461 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3462 nor a local message in the state where . is a terminator.
3464 (ii) If a bare CR is encountered in a header line, an extra space is added
3465 after the line terminator so as not to end the header. The reasoning
3466 behind this is that bare CRs in header lines are most likely either
3467 to be mistakes, or people trying to play silly games.
3469 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3470 was being incorrectly given as 18 bytes larger than it should have been.
3471 This is a VOB (very old bug).
3473 60. This may never have affected anything current, but just in case it has:
3474 When the local host is found other than at the start of a list of hosts,
3475 the local host, those with the same MX, and any that follow, are discarded.
3476 When the list in question was part of a longer list of hosts, the following
3477 hosts (not currently being processed) were also being discarded. This no
3478 longer happens. I'm not sure if this situation could ever has previously
3481 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3484 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3485 header. This is something that is recommended in a new Internet Draft, and
3486 is something that is documented as being done by Sendmail. There are two
3487 possible values. For messages generated by the autoreply transport, Exim
3490 Auto-Submitted: auto-replied
3492 whereas for all other generated messages (e.g. bounces) it adds
3494 Auto-Submitted: auto-generated
3496 63. The "personal" condition in filters now includes a test for the
3497 Auto-Submitted: header. If it contains the string "auto-" the message it
3498 not considered personal.
3500 64. Added rcpt_include_affixes as a generic transport option.
3502 65. Added queue_only_override (default true).
3504 66. Added the syslog_duplication option.
3506 67. If what should have been the first header line of a message consisted of
3507 a space followed by a colon, Exim was mis-interpreting it as a header line.
3508 It isn't of course - it is syntactically invalid and should therefore be
3509 treated as the start of the message body. The misbehaviour could have
3510 caused a number of strange effects, including loss of data in subsequent
3511 header lines, and spool format errors.
3513 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3514 client host had authenticated. This control can now be exercised by an ACL
3515 for more flexibility.
3517 69. By default, callouts do not happen when testing with -bh. There is now a
3518 variant, -bhc, which does actually run the callout code, including
3519 consulting and updating the callout cache.
3521 70. Added support for saslauthd authentication, courtesy of Alexander
3524 71. If statvfs() failed on the spool or log directories while checking their
3525 size for availability, Exim confusingly gave the error "space shortage".
3526 Furthermore, in debugging mode it crashed with a floating point exception.
3527 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3528 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3529 serious problem, Exim now writes to the main and panic logs when this
3530 happens, with details of the failure. It then refuses to accept the
3531 incoming message, giving the message "spool directory problem" or "log
3532 directory problem" with a 421 code for SMTP messages.
3534 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3535 0, 1, and 2 exist, because some OS complain for execs without them (see
3536 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3537 descriptors. However, the code omitted to check that the open succeeded,
3538 causing mysterious errors if for some reason the permissions on /dev/null
3539 got screwed. Now Exim writes a message to the main and panic logs, and
3540 bombs out if it can't open /dev/null.
3542 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3543 interact so that it is all more flexible. It is supposed to remain
3544 backwards compatible. Also added extra_local_interfaces.
3546 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3547 to bomb out with an assertion failure - to the client this appears as a
3548 connection drop. This problem occurs in the part of the code that was taken
3549 from the Samba project. Fortunately, the assertion is in a very simple
3550 function, so I have fixed this by reproducing the function inline in the
3551 one place where it is called, and arranging for authentication to fail
3552 instead of killing the process with assert().
3554 75. The SPA client code was not working when the server requested OEM rather
3555 than Unicode encoding.
3557 76. Added code to make require_files with a specific uid setting more usable in
3558 the case where statting the file as root fails - usually a non-root-mounted
3559 NFS file system. When this happens and the failure is EACCES, Exim now
3560 forks a subprocess and does the per-uid checking as the relevant uid.
3562 77. Added process_log_path.
3564 78. If log_file_path was not explicitly set, a setting of check_log_space or
3565 check_log_inodes was ignored.
3567 79. If a space check for the spool or log partitions fails, the incident is now
3568 logged. Of course, in the latter case the data may get lost...
3570 80. Added the %p formatting code to string_format() so that it can be used to
3571 print addresses in debug_print(). Adjusted all the address printing in the
3572 debugging in store.c to use %p rather than %d.
3574 81. There was a concern that a line of code in smtp_in.c could overflow a
3575 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3576 initially expressed, the concern was not well-founded, because trailing
3577 spaces are removed early. However, if the trailing spaces were followed by
3578 a NULL, they did not get removed, so the overflow was possible. Two fixes
3581 (a) I re-wrote the offending code in a cleaner fashion.
3582 (b) If an incoming SMTP command contains a NULL character, it is rejected
3585 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3586 runs initgroups(), so that if the Exim user is in any additional groups,
3587 they will be used during message reception.
3593 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob