1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.271 2005/12/06 10:25:59 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
17 PH/02 The ipliteral router was not recognizing addresses of the form user@
18 [ipv6:....] because it didn't know about the "ipv6:" prefix.
20 PH/03 Added disable_ipv6.
26 PH/01 Two changes to the default runtime configuration:
28 (1) Move the checks for relay_from_hosts and authenticated clients from
29 after to before the (commented out) DNS black list checks.
31 (2) Add control=submission to the relay_from_hosts and authenticated
32 clients checks, on the grounds that messages accepted by these
33 statements are most likely to be submissions.
35 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
37 (1) Generate an error if the third argument for the ${prvs expansion is
40 (2) Treat a missing third argument of ${prvscheck as if it were an empty
43 (3) Reset the variables that are obtained from the first argument of
44 ${prvscheck and used in the second argument before leaving the code,
45 because their memory is reclaimed, so using them afterwards may do
48 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
49 one (it's much easier than Tom thought :-).
51 (5) Because of (4), we can now allow for the use of $prvscheck_result
52 inside the third argument.
54 PH/03 For some reason, the default setting of PATH when running a command from
55 a pipe transport was just "/usr/bin". I have changed it to
58 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
59 anything to be listed in the output from -bV.
61 PH/05 When a filter generated an autoreply, the entire To: header line was
62 quoted in the delivery log line, like this:
64 => >A.N.Other <ano@some.domain> <original@ddress> ...
66 This has been changed so that it extracts the operative address. There
67 may be more than one such address. If so, they are comma-separated, like
70 => >ano@some.domain,ona@other.domain <original@ddress> ...
72 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
73 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
74 not being looked up in the rDNS to get a host name, Exim was showing the
75 IP address twice in Received: lines, even though the IP addresses were
76 identical. For example:
78 Received: from [1.2.3.4] (helo=[1.2.3.4])
80 However, if the real host name was known, it was omitting the HELO data
81 if it matched the actual IP address. This has been tidied up so that it
82 doesn't show the same IP address twice.
84 PH/07 When both +timestamp and +memory debugging was on, the value given by
85 $tod_xxx expansions could be wrong, because the tod_stamp() function was
86 called by the debug printing, thereby overwriting the timestamp buffer.
87 Debugging no longer uses the tod_stamp() function when +timestamp is set.
89 PH/08 When the original message was included in an autoreply transport, it
90 always said "this is a copy of the message, including all the headers",
91 even if body_only or headers_only was set. It now gives an appropriate
94 PH/09 Applied a patch from the Sieve maintainer which:
97 o adds the (disabled) notify extension core
98 o adds some debug output for the result of if/elsif tests
99 o points to the current vacation draft in the documentation
100 and documents the missing references header update
104 o fixes a bug in processing the envelope test (when testing
105 multiple envelope elements, the last element determinted the
108 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
109 Electronic Mail") by including:
111 Auto-submitted: auto-generated
113 in the messages that it generates (bounce messages and others, such as
114 warnings). In the case of bounce messages for non-SMTP mesages, there was
115 also a typo: it was using "Auto_submitted" (underscore instead of
116 hyphen). Since every message generated by Exim is necessarily in response
117 to another message, thes have all been changed to:
119 Auto-Submitted: auto-replied
121 in accordance with these statements in the RFC:
123 The auto-replied keyword:
125 - SHOULD be used on messages sent in direct response to another
126 message by an automatic process,
128 - MUST NOT be used on manually-generated messages,
130 - MAY be used on Delivery Status Notifications (DSNs) and Message
131 Disposition Notifications (MDNs),
133 - MUST NOT be used on messages generated by automatic or periodic
134 processes, except for messages which are automatic responses to
137 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
138 to the default Received: header definition.
140 PH/12 Added log selector acl_warn_skipped (default on).
142 PH/13 After a successful wildlsearch lookup, discard the values of numeric
143 variables because (a) they are in the wrong storage pool and (b) even if
144 they were copied, it wouldn't work properly because of the caching.
146 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
147 checking when decoding. Apparently there are clients that generate
148 overlong encoded strings. Why am I not surprised?
150 PH/15 If the first argument of "${if match_address" was not empty, but did not
151 contain an "@" character, Exim crashed. Now it writes a panic log message
152 and treats the condition as false.
154 PH/16 In autoreply, treat an empty string for "once" the same as unset.
156 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
157 extension "envelope-auth". The code is finished and in agreement with
158 other implementations, but there is no documentation so far and in fact,
159 nobody wrote the draft yet. This extension is currently #undef'ed, thus
160 not changing the active code.
162 Print executed "if" and "elsif" statements when debugging is used. This
163 helps a great deal to understand what a filter does.
165 Document more things not specified clearly in RFC3028. I had all this
166 sorted out, when out of a sudden new issues came to my mind. Oops."
168 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
171 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
172 canonical form (as documented). However, after a host name lookup from
173 the IP address, check_host() was doing a simple string comparison with
174 addresses acquired from the DNS when checking that the found name did
175 have the original IP as one of its addresses. Since any found IPv6
176 addresses are likely to be in abbreviated form, the comparison could
177 fail. Luckily, there already exists a function for doing the comparison
178 by converting both addresses to binary, so now that is used instead of
181 PH/20 There was another similar case to PH/19, when a complete host name was
182 given in a host list; looking up its IP address could give an abbreviated
183 form, whereas the current host's name might or might not be abbreviated.
184 The same fix has been applied.
190 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
191 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
194 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
195 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
197 PH/03 Typo: missing ".o" in src/pcre/Makefile.
199 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
200 header line, restrict the check to what is listed in RFCs 2369 and 2929.
201 Also, for "Auto-Submitted", treat anything other than "no" as
202 non-personal, in accordance with RFC 3834. (Previously it treated
203 anything starting "auto-" as non-personal.)
205 TF/01 The control=submission/name=... option had a problem with syntax
206 errors if the name included a slash character. The /name= option
207 now slurps the rest of the string, so it can include any characters
208 but it must come last in the list of options (after /sender_retain
211 PH/05 Some modifications to the interface to the fake nameserver for the new
214 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
215 into the default Received: header string.
222 TK/01 Added the "success_on_redirect" address verification option. See
223 NewStuff for rationale and an example.
225 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
227 PH/02 Patch to exigrep to allow it to work on syslog lines.
229 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
230 fread() to skip over the body file's header line, because in Cygwin the
231 header line is locked and is inaccessible.
233 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
234 co-exist for some time) to make it clear that it is the Exim ID that is
235 referenced, not the Message-ID: header line.
237 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
238 string_format() function, because snprintf() does not exist on all
241 PH/06 The use of forbid_filter_existstest now also locks out the use of the
242 ${stat: expansion item.
244 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
245 protocol synchronization error", to keep the pedants happy.
247 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
248 well as for IRIX systems, when gcc is being used. See the host.c source
251 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
253 PH/10 Named domain lists were not working if used in a queue_smtp_domains
256 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
257 transport and to the smtp transport in LMTP mode.
259 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
261 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
262 run a filter in a subprocess. This could lead to confusion in subsequent
263 lookups in the parent process. There should also be a search_tidyup() at
264 the end of the subprocess.
266 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
267 only if the host matched helo_try_verify_hosts, which caused the
268 verification to occur when the EHLO/HELO command was issued. The ACL just
269 tested the remembered result. Now, if a previous verification attempt has
270 not happened, "verify = helo" does it there and then.
272 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
274 TK/03 Fix log output including CR from clamd.
276 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
277 exist provoked a memory error which could cause a segfault.
279 PH/15 Installed PCRE 6.2
281 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
283 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
284 of the problem. Specifically, suggested +O2 rather than +O1 for the
287 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
289 PH/20 If a delivery was routed to a non-standard port by means of an SRV
290 record, the port was not correctly logged when the outgoing_port log
291 selector was set (it logged the transort's default port).
293 PH/21 Added support for host-specific ports to manualroute, queryprogram,
294 fallback_hosts, and "hosts" in the smtp transport.
296 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
297 host errors such as "Connection refused".
299 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
300 authentication with radiusclient 0.4.9:
302 - Error returned from rc_read_config was caught wrongly
303 - Username/password not passed on to radius server due to wrong length.
305 The presumption is that some radiusclient API changes for 4.51/PH/17
306 were not taken care of correctly. The code is still untested by me (my
307 Linux distribution still has 0.3.2 of radiusclient), but it was
308 contributed by a Radius user.
310 PH/24 When doing a callout, the value of $domain wasn't set correctly when
311 expanding the "port" option of the smtp transport.
313 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
314 while reading a MIME header. Thanks to Tom Hughes for a patch.
316 PH/24 Include config.h inside local_scan.h so that configuration settings are
319 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
320 that in an ACL for RCPT (for example), you can examine exactly what was
323 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
324 commands, but it was not correctly comparing the address with the actual
325 client host address. Thus, it would show the EHLO address in Received:
326 header lines when this was not necessary.
328 PH/27 Added the % operator to ${eval:}.
330 PH/28 Exim tries to create and chdir to its spool directory when it starts;
331 it should be ignoring failures (because with -C, for example, it has lost
332 privilege). It wasn't ignoring creation failures other than "already
335 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
338 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
339 because an input line was too long, either on its own, or by virtue of
340 too many continuations, the temporary file was not being removed, and the
341 return code was incorrect.
343 PH/31 Missing "BOOL" in function definition in filtertest.c.
345 PH/32 Applied Sieve patches from the maintainer.
347 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
349 PH/33 Added "verify = not_blind".
351 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
352 Local/Makefile (with some defaults set). These are used in built scripts
353 such as exicyclog, but they have never been used in the exim_install
354 script (though there are many overriding facilities there). I have
355 arranged that the exim_install script now takes note of these two
358 PH/35 Installed configuration files for Dragonfly.
360 PH/36 When a locally submitted message by a trusted user did not contain a
361 From: header, and the sender address was obtained from -f or from an SMTP
362 MAIL command, and the trusted user did not use -F to supply a sender
363 name, $originator_name was incorrectly used when constructing a From:
364 header. Furthermore, $originator_name was used for submission mode
365 messages from external hosts without From: headers in a similar way,
366 which is clearly wrong.
368 PH/37 Added control=suppress_local_fixups.
370 PH/38 When log_selector = +received_sender was set, and the addition of the
371 sender made the log line's construction buffer exactly full, or one byte
372 less than full, an overflow happened when the terminating "\n" was
375 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
376 when the result of a list match is failure because a DNS lookup failed.
378 PH/40 RM_COMMAND is now used in the building process.
380 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
381 the "build-* directories that it finds.
383 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
384 address in a domain literal was a prefix of an interface address.
386 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
387 when verifying a sender address, unless rewrite_headers is false.
389 PH/44 Wrote a long comment about why errors_to addresses are verified as
390 recipients, not senders.
392 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
393 the ratelimit ACL was added.
395 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
397 PH/46 Added extra information about PostgreSQL errors to the error string.
399 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
400 avoiding the need to install special zones in a real server. This is
401 backwards compatible; if it can't find the fake resolver, it drops back.
402 Thus, both old and new test suites can be run.
404 TF/02 Added util/ratelimit.pl
406 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
407 clock is set back in time.
409 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
410 Candler <B.Candler@pobox.com>.
412 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
413 OK for addresses that are the result of redirections.
415 PH/48 A number of further additions for the benefit of the new test suite,
416 including a fake gethostbyname() that interfaces to the fake DNS resolver
419 TF/06 The fix for widen_domains has also been applied to qualify_single and
420 search_parents which are the other dnslookup options that can cause
423 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
426 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
428 TF/07 Exim produced the error message "an SRV record indicated no SMTP
429 service" if it encountered an MX record with an empty target hostname.
430 The message is now "an MX or SRV record indicated no SMTP service".
432 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
433 if the DNS of the sending site is misconfigured. This is quite a
434 common situation. This change restores the behaviour of treating a
435 helo verification defer as a failure.
437 PH/51 If self=fail was set on a router, the bounce message did not include the
438 actual error message.
444 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
446 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
447 command itself ended in error, the underlying message about the transport
448 filter timeout was being overwritten with the pipe command error. Now the
449 underlying error message should be appended to the second error message.
451 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
453 PH/02 Reduce the amount of output that "make" produces by default. Full output
454 can still be requested.
456 PH/03 The warning log line about a condition test deferring for a "warn" verb
457 was being output only once per connection, rather than after each
458 occurrence (because it was using the same function as for successful
459 "warn" verbs). This seems wrong, so I have changed it.
461 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
462 it should not have, which might have caused a crash in the right
463 circumstances, but probably never did.
465 PH/04 Installed a modified version of Tony Finch's patch to make submission
466 mode fix the return path as well as the Sender: header line, and to
467 add a /name= option so that you can make the user's friendly name appear
470 TF/03 Added the control = fakedefer ACL modifier.
472 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
473 Mark Lowes for thorough testing.
475 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
477 TK/03 Merged latest SRS patch from Miles Wilton.
479 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
480 with the definition in sysexits.h (which is #included earlier).
481 Fortunately, Exim does not actually use EX_OK. The code used to try to
482 preserve the sysexits.h value, by assumimg that macro definitions were
483 scanned for macro replacements. I have been disabused of this notion,
484 so now the code just undefines EX_OK before #including unistd.h.
486 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
487 in the smtp transport. When a block could not be written in a single
488 write() function, the timeout was being re-applied to each part-write.
489 This seems wrong - if the receiver was accepting one byte at a time it
490 would take for ever. The timeout is now adjusted when this happens. It
491 doesn't have to be particularly precise.
493 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
494 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
496 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
497 without a domain if the check to <postmaster@domain> fails.
499 SC/01 Eximstats: added -xls and the ability to specify output files
500 (patch written by Frank Heydlauf).
502 SC/02 Eximstats: use FileHandles for outputing results.
504 SC/03 Eximstats: allow any combination of xls, txt, and html output.
506 SC/04 Eximstats: fixed display of large numbers with -nvr option
508 SC/05 Eximstats: fixed merging of reports with empty tables.
510 SC/06 Eximstats: added the -include_original_destination flag
512 SC/07 Eximstats: removed tabs and trailing whitespace.
514 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
516 TK/06 MBOX spool code: Add real "From " MBOX separator line
517 so the .eml file is really in mbox format (even though
518 most programs do not really care). Patch from Alex Miller.
520 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
521 The latter is generated from $received_to and is only set if the
522 message has one envelope recipient. SA can use these headers,
523 obviously out-of-the-box. Patch from Alex Miller.
525 PH/08 The ${def test on a variable was returning false if the variable's
526 value was "0", contrary to what the specification has always said!
527 The result should be true unless the variable is empty.
529 PH/09 The syntax error of a character other than { following "${if
530 def:variable_name" (after optional whitespace) was not being diagnosed.
531 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
532 accidental colon was present, for example, could give incorrect results.
534 PH/10 Tidied the code in a number of places where the st_size field of a stat()
535 result is used (not including appendfile, where other changes are about
538 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
539 This involved changing a lot of size variables from int to off_t. It
540 should work with maildirs and everything.
542 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
543 spamd dying while we are connected to it.
545 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
548 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
549 function for BDB 4.3.
551 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
553 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
554 the sources for PCRE have been re-arranged and more files are now
557 PH/15 The code I had for printing potentially long long variables in PH/11
558 above was not the best (it lost precision). The length of off_t variables
559 is now inspected at build time, and an appropriate printing format (%ld
560 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
561 to be "long long int" or "long int". This is needed for the internal
562 formatting function string_vformat().
564 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
565 the configuration file to be ":syslog", then the script "guesses" where
566 the logs files are, rather than using the compiled in default. In our
567 case the guess is not the same as the compiled default, so the script
568 suddenly stopped working when I started to use syslog. The patch checks
569 to see if log_file_path is "". If so, it attempts to read it from exim
570 with no configuration file to get the compiled in version, before it
571 falls back to the previous guessing code."
573 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
574 implementing BATV in an Exim configuration. See NewStuff for the gory
577 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
578 Makefile that are specific to HP-UX.
580 PH/18 If the "use_postmaster" option was set for a recipient callout together
581 with the "random" option, the postmaster address was used as the MAIL
582 FROM address for the random test, but not for the subsequent recipient
583 test. It is now used for both.
585 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
586 patch removes a few documentation additions to RFC 3028, because the
587 latest draft now contains them. It adds the new en;ascii-case comparator
588 and a new error check for 8bit text in MIME parts. Comparator and
589 require names are now matched exactly. I enabled the subaddress
590 extension, but it is not well tested yet (read: it works for me)."
592 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
593 rework some of the code of TK/09 above to avoid the hardwired use of
594 "%lld" and "long long". Replaced the call to snprintf() with a call to
597 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
598 records point to non-existent hosts", "retry timeout exceeded", and
599 "retry time not reached for any host after a long failure period".
601 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
602 experimental DomainKeys support:
604 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
605 (2) On an error such as an illegally used "control", the wrong name for
606 the control was given.
608 These problems did NOT occur unless DomainKeys support was compiled.
610 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
612 PH/24 Added ${if match_ip condition.
614 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
615 databases so that it will be absolutely obvious if a crash occurs in the
616 DB library. This is a regular occurrence (often caused by mis-matched
619 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
620 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
621 picked up on a user's system that detects such things. There doesn't seem
622 to be a gcc warning option for this - only an attribute that has to be
623 put on the function's prototype. It seems that in Fedora Core 4 they have
624 set this on a number of new functions. No doubt there will be more in due
627 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
628 specify a transport. However, if an address that was verified by such a
629 router was the subject of a callout, Exim crashed because it tried to
630 read the rcpt_include_affixes from the non-existent transport. Now it
631 just assumes that the setting of that option is false. This bug was
632 introduced by 4.51/PH/31.
634 PH/28 Changed -d+all to exclude +memory, because that information is very
635 rarely of interest, but it makes the output a lot bigger. People tend to
636 do -d+all out of habit.
638 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
639 code in os-type was giving problems when libc.so lives in lib64, like on
642 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
643 aren't the modern standard, and it seems that some systems' include files
644 don't always have them. Exim was already checking for some of the newer
645 ones like T_AAAA, and defining it itself. I've added checks for all the
646 record types that Exim uses.
648 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
649 not automatically generating a new one, as it is supposed to. This
650 prevented TLS from working. If the file did exist, but contained invalid
651 data, a new version was generated, as expected. It was only the case of a
652 non-existent file that was broken.
654 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
655 with a change in libdomainkeys > 0.64.
657 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
658 from DNS. If the selector record carries the flag, it now has
659 precedence over the domain-wide flag.
661 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
663 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
664 the use of a port name that isn't defined in /etc/services) occurred, the
665 message was deferred as in a normal delivery, and thus remained on the
666 spool, instead of being failed because of the mua_wrapper setting. This
667 is now fixed, and I tidied up some of the mua_wrapper messages at the
670 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
671 the messages in a hash of arrays rather than using individual hashes.
672 This is a bit cleaner and results in dramatic memory savings, albeit
673 at a slight CPU cost.
675 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
676 as requested by Marc Sherman.
678 SC/10 Eximstats: added histograms for user specified patterns as requested
681 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
683 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
684 fopen() in the content-scanning modules that did not already have it.
690 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
691 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
693 TK/02 Fix ACL "control" statement not being available in MIME ACL.
695 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
697 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
698 to test Sieve filters that use "vacation".
700 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
701 that changes the way the GnuTLS parameters are stored in the cache file.
702 The new format can be generated externally. For backward compatibility,
703 if the data in the cache doesn't make sense, Exim assumes it has read an
704 old-format file, and it generates new data and writes a new file. This
705 means that you can't go back to an older release without removing the
708 PH/03 A redirect router that has both "unseen" and "one_time" set does not
709 work if there are any delivery delays because "one_time" forces the
710 parent to be marked "delivered", so its unseen clone is never tried
711 again. For this reason, Exim now forbids the simultaneous setting of
714 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
715 redirected to themselves ("homonym" addresses). Read the long ChangeLog
716 entry if you want to know the details. The fix, however, neglected to
717 consider the case when local delivery batching is involved. The test for
718 "previously delivered" was not happening when checking to see if an
719 address could be batched with a previous (undelivered) one; under
720 certain circumstances this could lead to multiple deliveries to the same
723 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
724 in its include files, and this causes problems building Exim.
726 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
727 header_syntax) but Exim was just ignoring anything given after a slash.
728 In particular, this caused confusion with an attempt to use "verify =
729 reverse_host_lookup/defer_ok". An error is now given when options are
730 supplied for verify items that do not have them. (Maybe reverse_host_
731 lookup should have a defer_ok option, but that's a different point.)
733 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
734 defined by RFC 821) to 2048, because there were problems with some AUTH
735 commands, and RFC 1869 says the size should be increased for extended
736 SMTP commands that take arguments.
738 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
741 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
742 "unknown" error; now it says that the functionality isn't in the binary.
744 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
745 an address' error message when a string expansion fails (syntax or
746 whatever). Otherwise the password may appear in the log. Following change
747 PH/42 below, there is no longer a chance of it appearing in a bounce
750 PH/11 Installed exipick version 20050225.0 from John Jetmore.
752 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
753 of its addresses was ever tried. (Bugzilla bug #2.)
755 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
756 the result incorrectly in the debug output. (It correctly added a newline
757 to what was transported.)
759 TF/01 Added $received_time.
761 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
762 commented out examples of how to interface to a virus scanner and to
763 SpamAssassin. Also added commented examples of av_scanner and
764 spamd_address settings.
766 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
767 and controls are allowed in which ACLs. There were a couple of minor
768 errors. Some of the entries in the conditions table (which is a table of
769 where they are NOT allowed) were getting very unwieldy; rewrote them as a
770 negation of where the condition IS allowed.
772 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
774 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
775 header file does not have a version number, so I've had to invent a new
776 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
777 API. The code is untested by me (my Linux distribution still has 0.3.2 of
778 radiusclient), but it was contributed by a Radius user.
780 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
781 files or directories, for OpenSSL.
783 PH/19 When an Exim process that is running as root has to create an Exim log
784 file, it does so in a subprocess that runs as exim:exim so as to get the
785 ownership right at creation (otherwise, other Exim processes might see
786 the file with the wrong ownership). There was no test for failure of this
787 fork() call, which would lead to the process getting stuck as it waited
788 for a non-existent subprocess. Forks do occasionally fail when resources
789 run out. I reviewed all the other calls to fork(); they all seem to check
792 PH/20 When checking for unexpected SMTP input at connect time (before writing
793 the banner), Exim was not dealing correctly with a non-positive return
794 from the read() function. If the client had disconnected by this time,
795 the result was a log entry for a synchronization error with an empty
796 string after "input=" when read() returned zero. If read() returned -1
797 (an event I could not check), uninitialized data bytes were printed.
798 There were reports of junk text (parts of files, etc) appearing after
801 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
803 PH/22 Added support for macro redefinition, and (re)definition in between
804 driver and ACL definitions.
806 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
807 forgetting to use the resulting value; it was using the unexpanded value.
809 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
810 hadn't been configured. The fix is from Juergen Kreileder, who
811 understands it better than I do:
813 "Here's what I see happening with three configured cyrus_sasl
814 authenticators configured (plain, login, cram-md5):
816 On startup auth_cyrus_sasl_init() gets called for each of these.
817 This means three calls to sasl_listmech() without a specified mech_list.
818 => SASL tests which mechs of all available mechs actually work
819 => three warnings about OTP not working
820 => the returned list contains: plain, login, cram-md5, digest-md5, ...
822 With the patch, sasl_listmech() also gets called three times. But now
823 SASL's mech_list option is set to the server_mech specified in the the
824 authenticator. Or in other words, the answer from sasl_listmech()
825 gets limited to just the mech you're testing for (which is different
827 => the return list contains just 'plain' or 'login', 'cram-md5' or
828 nothing depending on the value of ob->server_mech.
830 I've just tested the patch: Authentication still works fine,
831 unavailable mechs specified in the exim configuration are still
832 caught, and the auth.log warnings about OTP are gone."
834 PH/25 When debugging is enabled, the contents of the command line are added
835 to the debugging output, even when log_selector=+arguments is not
838 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
839 answer is "GNU", and only if the return is "GNU/something" is the answer
842 PH/27 $acl_verify_message is now set immediately after the failure of a
843 verification in an ACL, and so is available in subsequent modifiers. In
844 particular, the message can be preserved by coding like this:
846 warn !verify = sender
847 set acl_m0 = $acl_verify_message
849 Previously, $acl_verify_message was set only while expanding "message"
850 and "log_message" when a very denied access.
852 PH/28 Modified OS/os.c-Linux with
854 -#ifndef OS_LOAD_AVERAGE
855 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
857 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
858 nomenclature these days.)
860 PH/29 Installed patch from the Sieve maintainer that adds the options
861 sieve_useraddress and sieve_subaddress to the redirect router.
863 PH/30 In these circumstances:
864 . Two addresses routed to the same list of hosts;
865 . First host does not offer TLS;
866 . First host accepts first address;
867 . First host gives temporary error to second address;
868 . Second host offers TLS and a TLS session is established;
869 . Second host accepts second address.
870 Exim incorrectly logged both deliveries with the TLS parameters (cipher
871 and peerdn, if requested) that were in fact used only for the second
874 PH/31 When doing a callout as part of verifying an address, Exim was not paying
875 attention to any local part prefix or suffix that was matched by the
876 router that accepted the address. It now behaves in the same way as it
877 does for delivery: the affixes are removed from the local part unless
878 rcpt_include_affixes is set on the transport.
880 PH/32 Add the sender address, as F=<...>, to the log line when logging a
881 timeout during the DATA phase of an incoming message.
883 PH/33 Sieve envelope tests were broken for match types other than :is. I have
884 applied a patch sanctioned by the Sieve maintainer.
886 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
887 the uid or gid is negative. A case of a negative gid caused this to be
888 noticed. The fix allows for either to be negative.
890 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
891 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
894 PH/36 The change PH/12 above was broken. Fixed it.
896 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
897 the grounds that routing the same address twice would always produce the
898 same answer. This might have been true once, but it is certainly no
899 longer true now. Routing a child address may depend on the previous
900 routing that produced that child. Some complicated redirection strategies
901 went wrong when messages had multiple recipients, and made Exim's
902 behaviour dependent on the order in which the addresses were given.
904 I have moved the duplicate checking until after the routing is complete.
905 Exim scans the addresses that are assigned to local and remote
906 transports, and removes any duplicates. This means that more work will be
907 done, as duplicates will always all be routed, but duplicates are
908 presumably rare, so I don't expect this is of any significance.
910 For deliveries to pipes, files, and autoreplies, the duplicate checking
911 still happens during the routing process, since they are not going to be
914 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
915 It corrects a timeout issue with spamd. This is Ian's comment: "The
916 background is that sometimes spamd either never reads data from a
917 connection it has accepted, or it never writes response data. The exiscan
918 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
919 blindly assumes that writes won't block so it may never time out."
921 PH/39 Allow G after quota size as well as K and M.
923 PH/40 The value set for $authenticated_id in an authenticator may not contain
924 binary zeroes or newlines because the value is written to log lines and
925 to spool files. There was no check on this. Now the value is run through
926 the string_printing() function so that such characters are converted to
927 printable escape sequences.
929 PH/41 $message_linecount is a new variable that contains the total number of
930 lines in the message. Compare $body_linecount, which is the count for the
933 PH/42 Exim no longer gives details of delivery errors for specific addresses in
934 bounce and delay warning messages, except in certain special cases, which
937 (a) An SMTP error message from a remote host;
938 (b) A message specified in a :fail: redirection;
939 (c) A message specified in a "fail" command in a system filter;
940 (d) A message specified in a FAIL return from the queryprogram router;
941 (e) A message specified by the cannot_route_message router option.
943 In these cases only, Exim does include the error details in bounce and
944 warning messages. There are also a few cases where bland messages such
945 as "unrouteable address" or "local delivery error" are given.
947 PH/43 $value is now also set for the "else" part of a ${run expansion.
949 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
950 being worked on, but at least Exim now implements the latest version to
953 PH/45 In a pipe transport, although a timeout while waiting for the pipe
954 process to complete was treated as a delivery failure, a timeout while
955 writing the message to the pipe was logged, but erroneously treated as a
956 successful delivery. Such timeouts include transport filter timeouts. For
957 consistency with the overall process timeout, these timeouts are now
958 treated as errors, giving rise to delivery failures by default. However,
959 there is now a new Boolean option for the pipe transport called
960 timeout_defer, which, if set TRUE, converts the failures into defers for
961 both kinds of timeout. A transport filter timeout is now identified in
964 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
965 systems where "make" and "gmake" are different, calling "gmake" at top
966 level broke things. I've arranged for the value of $(MAKE) to be passed
967 from the Makefile to this script so that it can call the same version of
971 A note about Exim versions 4.44 and 4.50
972 ----------------------------------------
974 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
975 changes of various kinds. As a consequence, a big documentation update was
976 needed. This delayed the release for rather longer than seemed good, especially
977 in the light of a couple of (minor) security issues. Therefore, the changes
978 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
979 release. So 4.44 and 4.50 are in effect two different branches that both start
982 I have left the 4.50 change log unchanged; it contains all the changes since
983 4.43. The change log for 4.44 is below; many of its items are identical to
984 those for 4.50. This seems to be the most sensible way to preserve the
985 historical information.
991 1. Minor wording change to the doc/README.SIEVE file.
993 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
994 computation of the current number of files was incorrect.
996 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
997 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1000 4. Give more explanation in the error message when the command for a transport
1001 filter fails to execute.
1003 5. There are several places where Exim runs a non-Exim command in a
1004 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1005 was being done only for the command run by the queryprogram router. It is
1006 now done for all such subprocesses. The other cases are: ${run, transport
1007 filters, and the commands run by the lmtp and pipe transports.
1009 6. Added CONFIGURE_GROUP build-time option.
1011 7. Some older OS have a limit of 256 on the maximum number of file
1012 descriptors. Exim was using setrlimit() to set 1000 as a large value
1013 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1014 systems. I've change it so that if it can't get 1000, it tries for 256.
1016 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1017 was an oversight, and furthermore, ever since the addition of extra
1018 controls (e.g. 4.43/32), the checks on when to allow different forms of
1019 "control" were broken. There should now be diagnostics for all cases when a
1020 control that does not make sense is encountered.
1022 9. Added the /retain_sender option to "control=submission".
1024 10. $recipients is now available in the predata ACL (oversight).
1026 11. Tidy the search cache before the fork to do a delivery from a message
1027 received from the command line. Otherwise the child will trigger a lookup
1028 failure and thereby defer the delivery if it tries to use (for example) a
1029 cached ldap connection that the parent has called unbind on.
1031 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1032 of $address_data from the recipient verification was clobbered by the
1033 sender verification.
1035 13. The value of address_data from a sender verification is now available in
1036 $sender_address_data in subsequent conditions in the ACL statement.
1038 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1040 15. Added a new option "connect=<time>" to callout options, to set a different
1043 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1044 was its contents. (It was OK if the option was not defined at all.)
1046 17. A "Completed" log line is now written for messages that are removed from
1047 the spool by the -Mrm option.
1049 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1050 information about exactly what failed.
1052 19. Added -dd to debug only the daemon process.
1054 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1055 handles timeouts, both on the server side and network timeouts. Renamed the
1056 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1058 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1059 It is now set to "smtps".
1061 22. $host_address is now set to the target address during the checking of
1062 ignore_target_hosts.
1064 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1065 being passed; this would have caused $sender_host_name to have been used if
1066 matching the list had actually called for a host name (not very likely,
1067 since this list is usually IP addresses). A host name is now passed as
1070 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1071 code that specifies a non-restarting handler (typically sigaction() in
1072 modern systems) in an attempt to fix a rare and obscure crash bug.
1074 25. Narrowed the window for a race in the daemon that could cause it to ignore
1075 SIGCHLD signals. This is not a major problem, because they are used only to
1076 wake it up if nothing else does.
1078 26. A malformed maildirsize file could cause Exim to calculate negative values
1079 for the mailbox size or file count. Odd effects could occur as a result.
1080 The maildirsize information is now recalculated if the size or filecount
1083 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1084 support for a long time. Removed HAVE_SYS_VFS_H.
1086 28. Installed the latest version of exipick from John Jetmore.
1088 29. In an address list, if the pattern was not a regular expression, an empty
1089 subject address (from a bounce message) matched only if the pattern was an
1090 empty string. Non-empty patterns were not even tested. This was the wrong
1091 because it is perfectly reasonable to use an empty address as part of a
1092 database query. An empty address is now tested by patterns that are
1093 lookups. However, all the other forms of pattern expect the subject to
1094 contain a local part and a domain, and therefore, for them, an empty
1095 address still always fails if the pattern is not itself empty.
1097 30. Exim went into a mad DNS loop when attempting to do a callout where the
1098 host was specified on an smtp transport, and looking it up yielded more
1099 than one IP address.
1101 31. Re-factored the code for checking spool and log partition space into a
1102 function that finds that data and another that does the check. The former
1103 is then used to implement four new variables: $spool_space, $log_space,
1104 $spool_inodes, and $log_inodes.
1106 32. The RFC2047 encoding function was originally intended for short strings
1107 such as real names; it was not keeping to the 75-character limit for
1108 encoded words that the RFC imposes. It now respects the limit, and
1109 generates multiple encoded words if necessary. To be on the safe side, I
1110 have increased the buffer size for the ${rfc2047: expansion operator from
1113 33. It is now permitted to omit both strings after an "if" condition; if the
1114 condition is true, the result is "true". As before, when the second string
1115 is omitted, a false condition yields an empty string. This makes it less
1116 cumbersome to write custom ACL and router conditions.
1118 34. Failure to deliver a bounce message always caused it to be frozen, even if
1119 there was an errors_to setting on the router. The errors_to setting is now
1122 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1123 canonical form (fully expanded) before being placed in
1124 $sender_host_address.
1126 36. The table in the code that translates DNS record types into text (T_A to
1127 "A" for instance) was missing entries for NS and CNAME. It is just possible
1128 that this could have caused confusion if both these types were looked up
1129 for the same domain, because the text type is used as part of Exim's
1130 per-process caching. But the chance of anyone hitting this buglet seems
1133 37. The dnsdb lookup has been extended in a number of ways.
1135 (1) There is a new type, "zns", which walks up the domain tree until it
1136 finds some nameserver records. It should be used with care.
1138 (2) There is a new type, "mxh", which is like "mx" except that it returns
1139 just the host names, not the priorities.
1141 (3) It is now possible to give a list of domains (or IP addresses) to be
1142 looked up. The behaviour when one of the lookups defers can be
1143 controlled by a keyword.
1145 (4) It is now possible to specify the separator character for use when
1146 multiple records are returned.
1148 38. The dnslists ACL condition has been extended: it is now possible to supply
1149 a list of IP addresses and/or domains to be looked up in a particular DNS
1152 39. Added log_selector=+queue_time_overall.
1154 40. When running the queue in the test harness, wait just a tad after forking a
1155 delivery process, to get repeatability of debugging output.
1157 41. Include certificate and key file names in error message when GnuTLS fails
1158 to set them up, because the GnuTLS error message doesn't include the name
1159 of the failing file when there is a problem reading it.
1161 42. Allow both -bf and -bF in the same test run.
1163 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1165 44. The "Exiscan patch" is now merged into the mainline Exim source.
1167 45. Sometimes the final signoff response after QUIT could fail to get
1168 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1169 < 0 before doing a fflush(). This bug looks as though it goes back to the
1170 introduction of TLS in release 3.20, but "sometimes" must have been rare
1171 because the tests only now provoked it.
1173 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1174 (this can affect the format of dates).
1176 47. exim_tidydb, when checking for the continued existence of a message for
1177 which it has found a message-specific retry record, was not finding
1178 messages that were in split spool directories. Consequently, it was
1179 deleting retry records that should have stayed in existence.
1181 48. Steve fixed some bugs in eximstats.
1183 49. The SPA authentication driver was not abandoning authentication and moving
1184 on to the next authenticator when an expansion was forced to fail,
1185 contradicting the general specification for all authenticators. Instead it
1186 was generating a temporary error. It now behaves as specified.
1188 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1189 (the order specifies the preference for clients). The order is now AES256,
1190 AES128, 3DES, ARCFOUR128.
1192 51. Small patch to Sieve code - explicitly set From: when generating an
1195 52. Exim crashed if a remote delivery caused a very long error message to be
1196 recorded - for instance if somebody sent an entire SpamAssassin report back
1197 as a large number of 550 error lines. This bug was coincidentally fixed by
1198 increasing the size of one of Exim's internal buffers (big_buffer) that
1199 happened as part of the Exiscan merge. However, to be on the safe side, I
1200 have made the code more robust (and fixed the comments that describe what
1203 53. Now that there can be additional text after "Completed" in log lines (if
1204 the queue_time_overall log selector is set), a one-byte patch to exigrep
1205 was needed to allow it to recognize "Completed" as not the last thing in
1208 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1209 patch that reportedly fixes this has been added. I am not expert enough to
1210 create a test for it. This is what the patch creator wrote:
1212 "I found a little strange behaviour of ldap code when working with
1213 Windows 2003 AD Domain, where users was placed in more than one
1214 Organization Units. When I tried to give exim partial DN, the exit code
1215 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1216 But simultaneously result of request was absolutely normal ldap result,
1217 so I produce this patch..."
1219 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1220 REFERENCE, so I have modified the code to exclude the patch when that macro
1223 55. Some experimental protocols are using DNS PTR records for new purposes. The
1224 keys for these records are domain names, not reversed IP addresses. The
1225 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1226 leaves it alone. Component reversal etc. now happens only for IP addresses.
1229 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1231 57. Double the size of the debug message buffer (to 2048) so that more of very
1232 long debug lines gets shown.
1234 58. The exicyclog utility now does better if the number of log files to keep
1235 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1237 59. Two changes related to the smtp_active_hostname option:
1239 (1) $smtp_active_hostname is now available as a variable.
1240 (2) The default for smtp_banner uses $smtp_active_hostname instead
1241 of $primary_hostname.
1243 60. The host_aton() function is supposed to be passed a string that is known
1244 to be a valid IP address. However, in the case of IPv6 addresses, it was
1245 not checking this. This is a hostage to fortune. Exim now panics and dies
1246 if the condition is not met. A case was found where this could be provoked
1247 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1248 components; fortuitously, this particular loophole had already been fixed
1249 by change 4.50/55 above.
1251 If there are any other similar loopholes, the new check in host_aton()
1252 itself should stop them being exploited. The report I received stated that
1253 data on the command line could provoke the exploit when Exim was running as
1254 exim, but did not say which command line option was involved. All I could
1255 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1256 running as the user.
1259 61. There was a buffer overflow vulnerability in the SPA authentication code
1260 (which came originally from the Samba project). I have added a test to the
1261 spa_base64_to_bits() function which I hope fixes it.
1264 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1265 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1267 63. The daemon start-up calls getloadavg() while still root for those OS that
1268 need the first call to be done as root, but it missed one case: when
1269 deliver_queue_load_max is set with deliver_drop_privilege. This is
1270 necessary for the benefit of the queue runner, because there is no re-exec
1271 when deliver_drop_privilege is set.
1273 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1274 This has been fixed.
1276 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1277 was in use, was not putting the data itself into the right store pool;
1278 consequently, it could be overwritten for a subsequent message in the same
1279 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1282 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1284 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1285 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1286 the function were treating the return as a boolean value, which happened to
1287 work because 0=false and not-0=true, but is not correct code.
1289 68. The host_aton() function was not handling scoped IPv6 addresses (those
1290 with, for example, "%eth0" on the end) correctly.
1292 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1293 negated items (that is, ~something) in unsigned ints. Some compilers
1294 apparently mutter when there is no cast.
1296 70. If an address verification called from an ACL failed, and did not produce a
1297 user-specific message (i.e. there was only a "system" message), nothing was
1298 put in $acl_verify_message. In this situation, it now puts the system
1301 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1302 session; change 4.31/43 added the unwanted input to the log line - except
1303 that it did not do this in the start of session case. It now does.
1305 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1306 This is wrong and can cause the other end to generate a synchronization
1307 error if it is another Exim or anything else that does the synchronization
1308 check. A QUIT command is no longer sent after a timeout.
1310 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1311 during host lookups.
1313 74. The defer_ok option of callout verification was not working if it was used
1314 when verifying addresses in header lines, that is, for this case:
1316 verify = header_sender/callout=defer_ok
1318 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1319 those file descriptors could be used for SMTP connections. If anything
1320 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1321 could be sent to the SMTP client, causing chaos. The daemon now opens
1322 stdin, stdout, and stderr to /dev/null when it puts itself into the
1325 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1326 log by default. The user can override this with suitable Perl magic.
1328 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1329 the log message when discard triggers, was not working for the DATA ACL or
1330 for the non-SMTP ACL.
1332 78. Error message wording change in sieve.c.
1334 79. If smtp_accept_max_per_host was set, the number of connections could be
1335 restricted to fewer than expected, because the daemon was trying to set up
1336 a new connection before checking whether the processes handling previous
1337 connections had finished. The check for completed processes is now done
1338 earlier. On busy systems, this bug wouldn't be noticed because something
1339 else would have woken the daemon, and it would have reaped the completed
1342 80. If a message was submitted locally by a user whose login name contained one
1343 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1344 It caused a spool format error. I have fixed the spool reading code. A
1345 related problem was that the "from" clause in the Received: line became
1346 illegal because of the space(s). It is now covered by ${quote_local_part.
1348 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1351 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1353 83. Patch from Sieve maintainer to fix unterminated string problem in
1354 "vacation" handling.
1356 84. Some minor changes to the Linux configuration files to help with other
1357 OS variants using glibc.
1359 85. One more patch for Sieve to update vacation handling to latest spec.
1362 ----------------------------------------------------
1363 See the note above about the 4.44 and 4.50 releases.
1364 ----------------------------------------------------
1370 1. Change 4.43/35 introduced a bug that caused file counts to be
1371 incorrectly computed when quota_filecount was set in an appendfile
1374 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1375 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1378 3. Give more explanation in the error message when the command for a transport
1379 filter fails to execute.
1381 4. There are several places where Exim runs a non-Exim command in a
1382 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1383 was being done only for the command run by the queryprogram router. It is
1384 now done for all such subprocesses. The other cases are: ${run, transport
1385 filters, and the commands run by the lmtp and pipe transports.
1387 5. Some older OS have a limit of 256 on the maximum number of file
1388 descriptors. Exim was using setrlimit() to set 1000 as a large value
1389 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1390 systems. I've change it so that if it can't get 1000, it tries for 256.
1392 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1393 was an oversight, and furthermore, ever since the addition of extra
1394 controls (e.g. 4.43/32), the checks on when to allow different forms of
1395 "control" were broken. There should now be diagnostics for all cases when a
1396 control that does not make sense is encountered.
1398 7. $recipients is now available in the predata ACL (oversight).
1400 8. Tidy the search cache before the fork to do a delivery from a message
1401 received from the command line. Otherwise the child will trigger a lookup
1402 failure and thereby defer the delivery if it tries to use (for example) a
1403 cached ldap connection that the parent has called unbind on.
1405 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1406 of $address_data from the recipient verification was clobbered by the
1407 sender verification.
1409 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1410 was its contents. (It was OK if the option was not defined at all.)
1412 11. A "Completed" log line is now written for messages that are removed from
1413 the spool by the -Mrm option.
1415 12. $host_address is now set to the target address during the checking of
1416 ignore_target_hosts.
1418 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1419 being passed; this would have caused $sender_host_name to have been used if
1420 matching the list had actually called for a host name (not very likely,
1421 since this list is usually IP addresses). A host name is now passed as
1424 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1425 code that specifies a non-restarting handler (typically sigaction() in
1426 modern systems) in an attempt to fix a rare and obscure crash bug.
1428 15. Narrowed the window for a race in the daemon that could cause it to ignore
1429 SIGCHLD signals. This is not a major problem, because they are used only to
1430 wake it up if nothing else does.
1432 16. A malformed maildirsize file could cause Exim to calculate negative values
1433 for the mailbox size or file count. Odd effects could occur as a result.
1434 The maildirsize information is now recalculated if the size or filecount
1437 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1438 support for a long time. Removed HAVE_SYS_VFS_H.
1440 18. Updated exipick to current release from John Jetmore.
1442 19. Allow an empty sender to be matched against a lookup in an address list.
1443 Previously the only cases considered were a regular expression, or an
1446 20. Exim went into a mad DNS lookup loop when doing a callout where the
1447 host was specified on the transport, if the DNS lookup yielded more than
1450 21. The RFC2047 encoding function was originally intended for short strings
1451 such as real names; it was not keeping to the 75-character limit for
1452 encoded words that the RFC imposes. It now respects the limit, and
1453 generates multiple encoded words if necessary. To be on the safe side, I
1454 have increased the buffer size for the ${rfc2047: expansion operator from
1457 22. Failure to deliver a bounce message always caused it to be frozen, even if
1458 there was an errors_to setting on the router. The errors_to setting is now
1461 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1462 canonical form (fully expanded) before being placed in
1463 $sender_host_address.
1465 24. Updated eximstats to version 1.33
1467 25. Include certificate and key file names in error message when GnuTLS fails
1468 to set them up, because the GnuTLS error message doesn't include the name
1469 of the failing file when there is a problem reading it.
1471 26. Expand error message when OpenSSL has problems setting up cert/key files.
1474 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1475 (this can affect the format of dates).
1477 28. exim_tidydb, when checking for the continued existence of a message for
1478 which it has found a message-specific retry record, was not finding
1479 messages that were in split spool directories. Consequently, it was
1480 deleting retry records that should have stayed in existence.
1482 29. eximstats updated to version 1.35
1483 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1484 1.35 - bugfix such that pie charts by volume are generated correctly
1486 30. The SPA authentication driver was not abandoning authentication and moving
1487 on to the next authenticator when an expansion was forced to fail,
1488 contradicting the general specification for all authenticators. Instead it
1489 was generating a temporary error. It now behaves as specified.
1491 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1492 (the order specifies the preference for clients). The order is now AES256,
1493 AES128, 3DES, ARCFOUR128.
1495 31. Small patch to Sieve code - explicitly set From: when generating an
1498 32. Exim crashed if a remote delivery caused a very long error message to be
1499 recorded - for instance if somebody sent an entire SpamAssassin report back
1500 as a large number of 550 error lines. This bug was coincidentally fixed by
1501 increasing the size of one of Exim's internal buffers (big_buffer) that
1502 happened as part of the Exiscan merge. However, to be on the safe side, I
1503 have made the code more robust (and fixed the comments that describe what
1506 33. Some experimental protocols are using DNS PTR records for new purposes. The
1507 keys for these records are domain names, not reversed IP addresses. The
1508 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1509 leaves it alone. Component reversal etc. now happens only for IP addresses.
1512 34. The host_aton() function is supposed to be passed a string that is known
1513 to be a valid IP address. However, in the case of IPv6 addresses, it was
1514 not checking this. This is a hostage to fortune. Exim now panics and dies
1515 if the condition is not met. A case was found where this could be provoked
1516 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1517 components; fortuitously, this particular loophole had already been fixed
1518 by change 4.50/55 or 4.44/33 above.
1520 If there are any other similar loopholes, the new check in host_aton()
1521 itself should stop them being exploited. The report I received stated that
1522 data on the command line could provoke the exploit when Exim was running as
1523 exim, but did not say which command line option was involved. All I could
1524 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1525 running as the user.
1528 35. There was a buffer overflow vulnerability in the SPA authentication code
1529 (which came originally from the Samba project). I have added a test to the
1530 spa_base64_to_bits() function which I hope fixes it.
1533 36. The daemon start-up calls getloadavg() while still root for those OS that
1534 need the first call to be done as root, but it missed one case: when
1535 deliver_queue_load_max is set with deliver_drop_privilege. This is
1536 necessary for the benefit of the queue runner, because there is no re-exec
1537 when deliver_drop_privilege is set.
1539 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1540 was in use, was not putting the data itself into the right store pool;
1541 consequently, it could be overwritten for a subsequent message in the same
1542 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1545 38. Sometimes the final signoff response after QUIT could fail to get
1546 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1547 < 0 before doing a fflush(). This bug looks as though it goes back to the
1548 introduction of TLS in release 3.20, but "sometimes" must have been rare
1549 because the tests only now provoked it.
1555 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1556 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1557 Now it returns an integer. A number of calls were still expecting a T/F
1558 return. Fortuitously, in all cases, the tests worked in OK situations,
1559 which is the norm. However, things would have gone wrong on any write
1560 failures on the smtp file descriptor. This function is used when sending
1561 messages over SMTP and also when doing verify callouts.
1563 2. When Exim is called to do synchronous delivery of a locally submitted
1564 message (the -odf or -odi options), it no longer closes stderr before doing
1567 3. Implemented the mua_wrapper option.
1569 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1571 5. Implemented the functions header_remove(), header_testname(),
1572 header_add_at_position(), and receive_remove_recipient(), and exported them
1575 6. If an ACL "warn" statement specified the addition of headers, Exim already
1576 inserted X-ACL-Warn: at the start if there was no header name. However, it
1577 was not making this test for the second and subsequent header lines if
1578 there were newlines in the string. This meant that an invalid header could
1579 be inserted if Exim was badly configured.
1581 7. Allow an ACL "warn" statement to add header lines at the start or after all
1582 the Received: headers, as well as at the end.
1584 8. Added the rcpt_4xx retry error code.
1586 9. Added postmaster_mailfrom=xxx to callout verification option.
1588 10. Added mailfrom=xxxx to the callout verification option, for verify=
1591 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1592 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1594 12. Inserted some casts to stop certain compilers warning when using pointer
1595 differences as field lengths or precisions in printf-type calls (mostly
1596 affecting debugging statements).
1598 13. Added optional readline() support for -be (dynamically loaded).
1600 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1601 same clock tick as a message's arrival, so that its received time was the
1602 same as the "first fail" time on the retry record, and that message
1603 remained on the queue past the ultimate address timeout, every queue runner
1604 would try a delivery (because it was past the ultimate address timeout) but
1605 after another failure, the ultimate address timeout, which should have then
1606 bounced the address, did not kick in. This was a "< instead of <=" error;
1607 in most cases the first failure would have been in the next clock tick
1608 after the received time, and all would be well.
1610 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1611 being recognized when the domain list was tested by the match_domain
1612 condition in an expansion string.
1614 16. Added the ${str2b64: operator.
1616 17. Exim was always calling setrlimit() to set a large limit for the number of
1617 processes, without checking whether the existing limit was already
1618 adequate. (It did check for the limit on file descriptors.) Furthermore,
1619 errors from getrlimit() and setrlimit() were being ignored. Now they are
1620 logged to the main and panic logs, but Exim does carry on, to try to do its
1621 job under whatever limits there are.
1623 18. Imported PCRE 5.0.
1625 19. Trivial typo in log message " temporarily refused connection" (the leading
1628 20. If the log selector return_path_on_delivery was set and an address was
1629 redirected to /dev/null, the delivery process crashed because it assumed
1630 that a return path would always be set for a "successful" delivery. In this
1631 case, the whole delivery is bypassed as an optimization, and therefore no
1634 21. Internal re-arrangement: the function for sending a challenge and reading
1635 a response while authentication was assuming a zero-terminated challenge
1636 string. It's now changed to take a pointer and a length, to allow for
1637 binary data in such strings.
1639 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1641 23. Exim was not respecting finduser_retries when seeking the login of the
1642 uid under which it was called; it was always trying 10 times. (The default
1643 setting of finduser_retries is zero.) Also, it was sleeping after the final
1644 failure, which is pointless.
1646 24. Implemented tls_on_connect_ports.
1648 25. Implemented acl_smtp_predata.
1650 26. If the domain in control=submission is set empty, Exim assumes that the
1651 authenticated id is a complete email address when it generates From: or
1652 Sender: header lines.
1654 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1655 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1656 chown and chgrp in /bin and hostname in /usr/bin.
1658 28. Exim was keeping the "process log" file open after each use, just as it
1659 does for the main log. This opens the possibility of it remaining open for
1660 long periods when the USR1 signal hits a daemon. Occasional processlog
1661 errors were reported, that could have been caused by this. Anyway, it seems
1662 much more sensible not to leave this file open at all, so that is what now
1665 29. The long-running daemon process does not normally write to the log once it
1666 has entered its main loop, and it closes the log before doing so. This is
1667 so that log files can straightforwardly be renamed and moved. However,
1668 there are a couple of unusual error situations where the daemon does write
1669 log entries, and I had neglected to close the log afterwards.
1671 30. The text of an SMTP error response that was received during a remote
1672 delivery was being truncated at 512 bytes. This is too short for some of
1673 the long messages that one sometimes sees. I've increased the limit to
1676 31. It is now possible to make retry rules that apply only when a message has a
1677 specific sender, in particular, an empty sender.
1679 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1680 it possible to be selective about when SMTP synchronization is enforced.
1682 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1684 32. Implemented hosts_connection_nolog.
1686 33. Added an ACL for QUIT.
1688 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1691 35. Added mailbox_size and mailbox_filecount to appendfile.
1693 36. Added control = no_multiline_responses to ACLs.
1695 37. There was a bug in the logic of the code that waits for the clock to tick
1696 in the case where the clock went backwards by a substantial amount such
1697 that the microsecond fraction of "now" was more than the microsecond
1698 fraction of "then" (but the whole seconds number was less).
1700 38. Added support for the libradius Radius client library this is found on
1701 FreeBSD (previously only the radiusclient library was supported).
1707 1. When certain lookups returned multiple values in the form name=value, the
1708 quoting of the values was not always being done properly. Specifically:
1709 (a) If the value started with a double quote, but contained no whitespace,
1711 (b) If the value contained whitespace other than a space character (i.e.
1712 tabs or newlines or carriage returns) it was not quoted.
1713 This fix has been applied to the mysql and pgsql lookups by writing a
1714 separate quoting function and calling it from the lookup code. The fix
1715 should probably also be applied to nisplus, ibase and oracle lookups, but
1716 since I cannot test any of those, I have not disturbed their existing code.
1718 2. A hit in the callout cache for a specific address caused a log line with no
1719 reason for rejecting RCPT. Now it says "Previous (cached) callout
1720 verification failure".
1722 3. There was an off-by-one bug in the queryprogram router. An over-long
1723 return line was truncated at 256 instead of 255 characters, thereby
1724 overflowing its buffer with the terminating zero. As well as fixing this, I
1725 have increased the buffer size to 1024 (and made a note to document this).
1727 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1728 when Exim is waiting for an SMTP response from a remote server, Exim
1729 restarts its select() call on the socket, thereby resetting its timeout.
1730 This is not a problem when such interrupts are rare. Somebody set up a cron
1731 job to run exiwhat every 2 minutes, which is less than the normal select()
1732 timeout (5 or 10 minutes). This meant that the select() timeout never
1733 kicked in because it was always reset. I have fixed this by comparing the
1734 time when an interrupt arrives with the time at the start of the first call
1735 to select(). If more time than the timeout has elapsed, the interrupt is
1736 treated as a timeout.
1738 5. Some internal re-factoring in preparation for the addition of Sieve
1739 extensions (by MH). In particular, the "personal" test is moved to a
1740 separate function, and given an option for scanning Cc: and Bcc: (which is
1741 not set for Exim filters).
1743 6. When Exim created an email address using the login of the caller as the
1744 local part (e.g. when creating a From: or Sender: header line), it was not
1745 quoting the local part when it contained special characters such as @.
1747 7. Installed new OpenBSD configuration files.
1749 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1750 try to make them clearer.
1752 9. Callout options, other than the timeout value, were being ignored when
1753 verifying sender addresses in header lines. For example, when using
1755 verify = header_sender/callout=no_cache
1757 the cache was (incorrectly) being used.
1759 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1760 only the Cygwin environment.
1762 11. When return_path_on_delivery was set as a log selector, if different remote
1763 addresses in the same message used different return paths and parallel
1764 remote delivery occurred, the wrong values would sometimes be logged.
1765 (Whenever a remote delivery process finished, the return path value from
1766 the most recently started remote delivery process was logged.)
1768 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1769 lines when AUTH and/or TLS are in use. This is the "received protocol"
1770 field. Exim used to use "asmtp" for authenticated SMTP, without any
1771 indication (in the protocol name) for TLS use. Now it follows the RFC and
1772 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1773 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1774 names appear in log lines as well as in Received: header lines.
1776 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1777 extensions, and comparison tests, and to fix some bugs.
1779 14. Changes to the "personal" filter test:
1781 (1) The test was buggy in that it was just doing the equivalent of
1782 "contains" tests on header lines. For example, if a user's address was
1783 anne@some.where, the "personal" test would incorrectly be true for
1785 To: susanne@some.where
1787 This test is now done by extracting each address from the header in turn,
1788 and checking the entire address. Other tests that are part of "personal"
1789 are now done using regular expressions (for example, to check local parts
1790 of addresses in From: header lines).
1792 (2) The list of non-personal local parts in From: addresses has been
1793 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1794 taken from the Sieve specification recommendations.
1796 (3) If the message contains any header line starting with "List-" it is
1797 treated as non-personal.
1799 (4) The test for "circular" in the Subject: header line has been removed
1800 because it now seems ill-conceived.
1802 15. Minor typos in src/EDITME comments corrected.
1804 16. Installed latest exipick from John Jetmore.
1806 17. If headers_add on a router specified a text string that was too long for
1807 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1808 of string_sprintf() is now avoided.
1810 18. $message_body_size was not set (it was always zero) when running the DATA
1811 ACL and the local_scan() function.
1813 19. For the "mail" command in an Exim filter, no default was being set for
1814 the once_repeat time, causing a random time value to be used if "once" was
1815 specified. (If the value happened to be <= 0, no repeat happened.) The
1816 default is now 0s, meaning "never repeat". The "vacation" command was OK
1817 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1818 (I found it when inspecting the code).
1820 20. There is now an overall timeout for performing a callout verification. It
1821 defaults to 4 times the callout timeout, which applies to individual SMTP
1822 commands during the callout. The overall timeout applies when there is more
1823 than one host that can be tried. The timeout is checked before trying the
1824 next host. This prevents very long delays if there are a large number of
1825 hosts and all are timing out (e.g. when the network connections are timing
1826 out). The value of the overall timeout can be changed by specifying an
1827 additional sub-option for "callout", called "maxwait". For example:
1829 verify = sender/callout=5s,maxwait=20s
1831 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1832 to the end before writing, but this should make it even safer).
1834 22. Exim was forgetting that it had advertised PIPELINING for the second and
1835 subsequent messages on an SMTP connection. It was also not resetting its
1836 memory on STARTTLS and an internal HELO.
1838 23. When Exim logs an SMTP synchronization error within a session, it now
1839 records whether PIPELINING has been advertised or not.
1841 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1842 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1843 rather than long int.
1845 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1847 26. Added the never_mail option to autoreply.
1853 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1854 crash if the getsockname() call failed; this can happen if a connection is
1855 closed very soon after it is established. The problem was simply in the
1856 order in which certain operations were done, causing Exim to try to write
1857 to the SMTP stream before it had set up the file descriptor. The bug has
1858 been fixed by making things happen in the correct order.
1864 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1865 before the connection was closed, thus losing the rejection response.
1867 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1868 some early Solaris releases, but causes trouble in current releases where
1869 socklen_t is defined.
1871 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1874 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1877 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1878 that releasing the top of it at the end releases what was used for sub-
1879 expansions (unless the block got too big). However, discard this block if
1880 the first thing is a variable or header, so that we can use its block when
1881 it is dynamic (useful for very large $message_headers, for example).
1883 6. Lookups now cache *every* query, not just the most recent. A new, separate
1884 store pool is used for this. It can be recovered when all lookup caches are
1885 flushed. Lookups now release memory at the end of their result strings.
1886 This has involved some general refactoring of the lookup sources.
1888 7. Some code has been added to the store_xxx() functions to reduce the amount
1889 of flapping under certain conditions.
1891 8. log_incoming_interface used to affect only the <= reception log lines. Now
1892 it causes the local interface and port to be added to several more SMTP log
1893 lines, for example "SMTP connection from", and rejection lines.
1895 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1897 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1899 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1900 could be overwritten at the end of the current message (or the start of a
1901 new message if it was set in a HELO ACL). The value is now preserved for
1902 the duration of the SMTP connection.
1904 12. If a transport had a headers_rewrite setting, and a matching header line
1905 contained an unqualified address, that address was qualified, even if it
1906 did not match any rewriting rules. The underlying bug was that the values
1907 of the flags that permit the existence of unqualified sender and recipient
1908 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1909 non-local messages, and by -bnq for local messages) were not being
1910 preserved with the message after it was received.
1912 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1913 "next input=" as part of the text comprising the host identity instead of
1914 the correct text. The code was using the same buffer for two different
1915 strings. However, depending on which order the printing function evaluated
1916 its arguments, the bug did not always show up. Under Linux, for example, my
1917 test suite worked just fine.
1919 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1920 doesn't work with some older versions of Perl. It has been changed to "my",
1921 which in any case is probably the better facility to use.
1923 15. A really picky compiler found some instances of statements for creating
1924 error messages that either had too many or two few arguments for the format
1927 16. The size of the buffer for calls to the DNS resolver has been increased
1928 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1929 for addresses that have a lot of PTR records. This alleviates a problem; it
1930 does not fully solve it.
1932 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1933 buffer now truncates the list and logs the incident, which is the same
1934 action as happens when Exim is looking up a host name and its aliases.
1935 Previously in this situation something unpredictable would happen;
1936 sometimes it was "internal error: store_reset failed".
1938 18. If a server dropped the connection unexpectedly when an Exim client was
1939 using GnuTLS and trying to read a response, the client delivery process
1940 crashed while trying to generate an error log message.
1942 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1943 string, for example:
1945 warn message = H1: something\nH2: something
1947 the text was added as a single header line from Exim's point of view
1948 though it ended up OK in the delivered message. However, searching for the
1949 second and subsequent header lines using $h_h2: did not work. This has been
1950 fixed. Similarly, if a system filter added multiple headers in this way,
1951 the routers could not see them.
1953 20. Expanded the error message when iplsearch is called with an invalid key to
1954 suggest using net-iplsearch in a host list.
1956 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1957 ACLs are no longer actually imposed (and a message to that effect is
1960 22. If a "gecos" field in a passwd entry contained escaped characters, in
1961 particular, if it contained a \" sequence, Exim got it wrong when building
1962 a From: or a Sender: header from that name. A second bug also caused
1963 incorrect handling when an unquoted " was present following a character
1964 that needed quoting.
1966 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1967 was not being matched caselessly.
1969 24. Arranged for all hyphens in the exim.8 source to be escaped with
1972 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1973 properly. Recipient callout cache records were still being keyed to include
1974 the sender, even when use_sender was set false. This led to far more
1975 callouts that were necessary. The sender is no longer included in the key
1976 when use_sender is false.
1978 26. Added "control = submission" modifier to ACLs.
1980 27. Added the ${base62d: operator to decode base 62 numbers.
1982 28. dnsdb lookups can now access SRV records.
1984 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1985 the configuration file.
1987 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1988 (-v) mode. This makes the output for a verbose queue run more intelligible.
1990 31. Added a use_postmaster feature to recipient callouts.
1992 32. Added the $body_zerocount variable, containing the number of binary zero
1993 bytes in the message body.
1995 33. The time of last modification of the "new" subdirectory is now used as the
1996 "mailbox time last read" when there is a quota error for a maildir
1999 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2001 35. Added +ignore_unknown as a special item in host lists.
2003 36. Code for decoding IPv6 addresses in host lists is now included, even if
2004 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2005 address was recognized as an IP address, but was then not correctly decoded
2006 into binary, causing unexpected and incorrect effects when compared with
2013 1. Very minor rewording of debugging text in manualroute to say "list of
2014 hosts" instead of "hostlist".
2016 2. If verify=header_syntax was set, and a header line with an unqualified
2017 address (no domain) and a large number of spaces between the end of the
2018 name and the colon was received, the reception process suffered a buffer
2019 overflow, and (when I tested it) crashed. This was caused by some obsolete
2020 code that should have been removed. The fix is to remove it!
2022 3. When running in the test harness, delay a bit after writing a bounce
2023 message to get a bit more predictability in the log output.
2025 4. Added a call to search_tidyup() just before forking a reception process. In
2026 theory, someone could use a lookup in the expansion of smtp_accept_max_
2027 per_host which, without the tidyup, could leave open a database connection.
2029 5. Added the variables $recipient_data and $sender_data which get set from a
2030 lookup success in an ACL "recipients" or "senders" condition, or a router
2031 "senders" option, similar to $domain_data and $local_part_data.
2033 6. Moved the writing of debug_print from before to after the "senders" test
2036 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2037 problems for message scanning, either using a data ACL, or using
2038 local_scan() because the Received: header was not generated till after they
2039 were called (in order to set the time as the time of reception completion).
2040 I have revised the way this works. The header is now generated after the
2041 body is received, but before the ACL or local_scan() are called. After they
2042 are run, the timestamp in the header is updated.
2048 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2049 before starting a queue runner without re-exec. This happened only when
2050 deliver_drop_privilege was set or when the Exim user was set to root. The
2051 effect of the bug was that timeouts during subsequent deliveries caused
2052 crashes instead of being properly handled. The handler is now left at its
2053 default (and expected) setting.
2055 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2056 message, again when deliver_drop_privilege is set or Exim is run as root.
2057 The bug described in (1) was not present in this case, but the tidying up
2058 of the other signals was missing. I have made the two cases consistent.
2060 3. The ignore_target_hosts setting on a manualroute router was being ignored
2061 for hosts that were looked up using the /MX notation.
2063 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2066 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2067 operated on the sender address. After changing the $sender_address to <>
2068 for the sender address verify, Exim was re-instated it as the original
2069 (before rewriting) address, but remembering that it had rewritten it, so it
2070 wasn't rewriting it again. This bug also had the effect of breaking the
2071 sender address verification caching when the sender address was rewritten.
2073 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2074 This has been changed so that if the ip literal address matches
2075 ignore_target_hosts, the router declines.
2077 7. Added expansion conditions match_domain, match_address, and match_local_
2078 part (NOT match_host).
2080 8. The placeholder for the Received: header didn't have a length field set.
2082 9. Added code to Exim itself and to exim_lock to test for a specific race
2083 condition that could lead to file corruption when using MBX delivery. The
2084 issue is with the lockfile that is created in /tmp. If this file is removed
2085 after a process has opened it but before that process has acquired a lock,
2086 there is the potential for a second process to recreate the file and also
2087 acquire a lock. This could lead to two Exim processes writing to the file
2088 at the same time. The added code performs the same test as UW imapd; it
2089 checks after acquiring the lock that its file descriptor still refers to
2090 the same named file.
2092 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2093 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2096 11. Added the smtp_active_hostname option. If used, this will typically be made
2097 to depend on the incoming interface address. Because $interface_address is
2098 not set up until the daemon has forked a reception process, error responses
2099 that can happen earlier (such as "too many connections") no longer contain
2102 12. If an expansion in a condition on a "warn" statement fails because a lookup
2103 defers, the "warn" statement is abandoned, and the next ACL statement is
2104 processed. Previously this caused the whole ACL to be aborted.
2106 13. Added the iplsearch lookup type.
2108 14. Added ident_timeout as a log selector.
2110 15. Added tls_certificate_verified as a log selector.
2112 16. Added a global option tls_require_ciphers (compare the smtp transport
2113 option of the same name). This controls incoming TLS connections.
2115 17. I finally figured out how to make tls_require_ciphers do a similar thing
2116 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2117 before starting the TLS session.
2119 18. Tabs are now shown as \t in -bP output.
2121 19. If the log selector return_path_on_delivery was set, Exim crashed when
2122 bouncing a message because it had too many Received: header lines.
2124 20. If two routers both had headers_remove settings, and the first one included
2125 a superfluous trailing colon, the final name in the first list and the
2126 first name in the second list were incorrectly joined into one item (with a
2127 colon in the middle).
2133 1. Added -C and -D options to the exinext utility, mainly to make it easier
2134 to include in the automated testing, but these could be helpful when
2135 multiple configurations are in use.
2137 2. The exinext utility was not formatting the output nicely when there was
2138 an alternate port involved in the retry record key, nor when there was a
2139 message id as well (for retries that were specific to a specific message
2140 and a specific host). It was also confused by IPv6 addresses, because of
2141 the additional colons they contain. I have fixed the IPv4 problem, and
2142 patched it up to do a reasonable job for IPv6.
2144 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2145 delivery, the log line now contains "pipelined" if PIPELINING was used.
2147 4. An SMTP transport process used to panic and die if the bind() call to set
2148 an explicit outgoing interface failed. This has been changed; it is now
2149 treated in the same way as a connect() failure.
2151 5. A reference to $sender_host_name in the part of a conditional expansion
2152 that was being skipped was still causing a DNS lookup. This no longer
2155 6. The def: expansion condition was not recognizing references to header lines
2156 that used bh_ and bheader_.
2158 7. Added the _cache feature to named lists.
2160 8. The code for checking quota_filecount in the appendfile transport was
2161 allowing one more file than it should have been.
2163 9. For compatibility with Sendmail, the command line option
2171 and sets the incoming protocol and host name (for trusted callers). The
2172 host name and its colon can be omitted when only the protocol is to be set.
2173 Note the Exim already has two private options, -pd and -ps, that refer to
2174 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2175 "s", but I don't think that's a major issue.
2177 10. A number of refactoring changes to the code, none of which should affect
2180 (a) The number of logging options was getting close to filling up the
2181 32-bit word that was used as a bit map. I have split them into two classes:
2182 those that are passed in the argument to log_write(), and those that are
2183 only ever tested independently outside of that function. These are now in
2184 separate 32-bit words, so there is plenty of room for expansion again.
2185 There is no change in the user interface or the logging behaviour.
2187 (b) When building, for example, log lines, the code previously used a
2188 macro that called string_cat() twice, in order to add two strings. This is
2189 not really sufficiently general. Furthermore, there was one instance where
2190 it was actually wrong because one of the argument was used twice, and in
2191 one call a function was used. (As it happened, calling the function twice
2192 did not affect the overall behaviour.) The macro has been replaced by a
2193 function that can join an arbitrary number of extra strings onto a growing
2196 (c) The code for expansion conditions now uses a table and a binary chop
2197 instead of a serial search (which was left over from when there were very
2198 few conditions). Also, it now recognizes conditions like "pam" even when
2199 the relevant support is not compiled in: a suitably worded error message is
2200 given if an attempt is made to use such a condition.
2202 11. Added ${time_interval:xxxxx}.
2204 12. A bug was causing one of the ddress fields not to be passed back correctly
2205 from remote delivery subprocesses. The field in question was not being
2206 subsequently used, so this caused to problems in practice.
2208 13. Added new log selectors queue_time and deliver_time.
2210 14. Might have fixed a bug in maildirsizefile handling that threw up
2211 "unexpected character" debug warnings, and recalculated the data
2212 unnecessarily. In any case, I expanded the warning message to give more
2215 15. Added the message "Restricted characters in address" to the statements in
2216 the default ACL that block characters like @ and % in local parts.
2218 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2219 Three changes have been made:
2221 (a) There was a serious bug; a negative response to MAIL caused the whole
2222 recipient domain to be cached as invalid, thereby blocking all messages
2223 to all local parts at the same domain, from all senders. This bug has
2224 been fixed. The domain is no longer cached after a negative response to
2225 MAIL if the sender used is not empty.
2227 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2230 (c) A new callout option, "use_sender" has been added for people who want
2231 the modified behaviour.
2237 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2240 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2241 indeed breaks things for older releases.
2243 3. Added additional logging to the case where there is a problem reading data
2244 from a filter that is running in a subprocess using a pipe, in order to
2245 try to track down a specific problem.
2247 4. Testing facility fudge: when running in the test harness and attempting
2248 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2249 getting "No route to host". Convert this to a timeout.
2251 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2254 6. Some OS don't have socklen_t but use size_t instead. This affects the
2255 fifth argument of getsockopt() amongst other things. This is now
2256 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2257 can be set for individual OS. I have set it for SunOS5, OSF1, and
2258 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2259 some earlier ones do not.
2261 7. Change 4.30/15 was not doing the test caselessly.
2263 8. The standard form for an IPv6 address literal was being rejected by address
2264 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2265 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2266 this, as well as the form without the "IPv6" on the front (but only when
2267 address literals are enabled, of course).
2269 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2271 10. Exim crashed if a message with an empty sender address specified by -f
2272 encountered a router with an errors_to setting. This could be provoked only
2273 by a command such as
2277 where an empty string was supplied; "<>" did not hit this bug.
2279 11. Installed PCRE release 4.5.
2281 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2282 remained set. It is now erased.
2284 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2285 times from message ids (which are base 36 rather than the normal 62).
2287 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2288 were being counted as actual protocol errors, and logged if the log
2289 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2290 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2291 and DATA following a set of rejected RCPTs do not count as protocol errors.
2292 In other words, Exim assumes they were pipelined, though this may not
2293 actually be the case. Of course, in all cases the client gets an
2294 appropriate error code.
2296 15. If a lookup fails in an ACL condition, a message about the failure may
2297 be available; it is used if testing the ACL cannot continue, because most
2298 such messages specify what the cause of the deferral is. However, some
2299 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2300 that caused an old message to be retained and used if a later statement
2301 caused a defer, replacing the real cause of the deferral.
2303 16. If an IP address had so many PTR records that the DNS lookup buffer
2304 was not large enough to hold them, Exim could crash while trying to process
2305 the truncated data. It now detects and logs this case.
2307 17. Further to 4.21/58, another change has been made: if (and only if) the
2308 first line of a message (the first header line) ends with CRLF, a bare LF
2309 in a subsequent header line has a space inserted after it, so as not to
2310 terminate the header.
2312 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2313 unnecessarily, used atoi() instead of strtol(), and didn't check the
2314 termination when getting file sizes from file names by regex.
2316 19. Completely re-implemented the support for maildirsize files, in the light
2317 of a number of problems with the previous contributed implementation
2318 (4.30/29). In particular:
2320 . If the quota is zero, the maildirsize file is maintained, but no quota is
2323 . If the maildir directory does not exist, it is created before any attempt
2324 to write a maildirsize file.
2326 . The quota value in the file is just a cache; if the quota is changed in
2327 the transport, the new value overrides.
2329 . A regular expression is available for excluding directories from the
2332 20. The autoreply transport checks the characters in options that define the
2333 message's headers; it allows continued headers, but it was checking with
2334 isspace() after an embedded newline instead of explicitly looking for a
2337 21. If all the "regular" hosts to which an address was routed had passed their
2338 expiry times, and had not reached their retry times, the address was
2339 bounced, even if fallback hosts were defined. Now Exim should go on to try
2342 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2343 equivalent code in the SMTP transport. Some hosts send humungous responses
2344 to HELO/EHLO, more than 1024 it seems.
2346 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2347 gives compiler warnings in some environments. I've now done it "properly",
2350 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2351 (because of problems with the built-in one) was declared to return uschar *
2352 instead of char *, causing compiler failure.
2354 25. Fixed a file descriptor leak when processing alias/forward files.
2356 26. Fixed a minor format string issue in dbfn.c.
2358 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2360 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2361 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2364 29. When a sender address is verified, it is cached, to save repeating the test
2365 when there is more than one recipient in a message. However, when the
2366 verification involves a callout, it is possible for different callout
2367 options to be set for different recipients. It is too complicated to keep
2368 track of this in the cache, so now Exim always runs a verification when a
2369 callout is required, relying on the callout cache for the optimization.
2370 The overhead is duplication of the address routing, but this should not be
2373 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2374 to be verified with callout=postmaster, and the main callout worked but the
2375 postmaster check failed, the verification correctly failed. However, if a
2376 subsequent RCPT command asked for sender verification *without* the
2377 postmaster check, incorrect caching caused this verification also to fail,
2380 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2381 it was not caching the DNS options (qualify_single, search_parents) that
2382 were used when the lookup failed. A subsequent lookup with different
2383 options therefore always gave the same answer, though there were cases
2384 where it should not have. (Example: a "domains = !$mx_any" option on a
2385 dnslookup router: the "domains" option is always processed without any
2386 widening, but the router might have qualify_single set.) Now Exim uses the
2387 cached value only when the same options are set.
2389 32. Added John Jetmore's "exipick" utility to the distribution.
2391 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2392 than a timeout (e.g. a certificate is required, and is not provided), an
2393 Exim server now closes the connection immediately. Previously it waited for
2394 the client to close - but if the client is SSL, it seems that they each
2395 wait for each other, leading to a delay before one of them times out.
2397 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2398 maintained 0.8.x compatibility because I don't think many are using it, and
2399 it is clearly obsolete.
2401 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2404 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2405 client certificate was expired. A simple patch fixes this, though I don't
2406 understand the full logic of why the verify callback is called multiple
2409 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2410 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2411 which causes problems with some clients (such as the Certicom SSL Plus
2412 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2413 disables the coutermeasure allowing Eudora to connect."
2415 38. Exim was not checking that a write() to a log file succeeded. This could
2416 lead to Bad Things if a log got too big, in particular if it hit a file
2417 size limit. Exim now panics and dies if it cannot write to a log file, just
2418 as it does if it cannot open a log file.
2420 39. Modified OS/Makefile-Linux so that it now contains
2422 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2424 The two -D definitions ensure that Exim is compiled with large file
2425 support, which makes it possible to handle log files that are bigger than
2428 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2429 instance) a domain was checked against a named list that involved a lookup,
2430 causing $domain_data to be set, then another domain was checked against the
2431 same list, then the first domain was re-checked, the value of $domain_data
2432 after the final check could be wrong. In particular, if the second check
2433 failed, it could be set empty. This bug probably also applied to
2436 41. The strip_trailing_dot option was not being applied to the address given
2437 with the -f command-line option.
2439 42. The code for reading a message's header from the spool was incrementing
2440 $received_count, but never initializing it. This meant that the value was
2441 incorrect (doubled) while delivering a message in the same process in which
2442 it was received. In the most common configuration of Exim, this never
2443 happens - a fresh exec is done - but it can happen when
2444 deliver_drop_privilege is set.
2446 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2447 it now includes up to 150 characters of the unexpected data in the log
2450 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2451 and building data strings. The size of both of these buffers was 10 000
2452 bytes - far larger than anybody would *ever* want, thought I. Needless to
2453 say, somebody hit the limit. I have increased the maximum line length to
2454 20 000 and the maximum data length of concatenated lines to 100 000. I have
2455 also fixed two bugs, because there was no checking on these buffers. Tsk,
2456 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2457 buffer is too small.
2459 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2460 lsearch lookups. Now it does.
2462 46. When parsing a route_list item in a manualroute router, a fixed-length
2463 buffer was used for the list of hosts. I made this 1024 bytes long,
2464 thinking that nobody would ever have a list of hosts that long. Wrong.
2465 Somebody had a whole pile of complicated expansion conditions, and the
2466 string was silently truncated, leading to an expansion error. It turns out
2467 that it is easier to change to an unlimited length (owing to other changes
2468 that have happened since this code was originally written) than to build
2469 structure for giving a limitation error. The length of the item that
2470 expands into the list of hosts is now unlimited.
2472 47. The lsearch lookup could not handle data where the length of text line was
2473 more than 4095 characters. Such lines were truncated, leading to shortened
2474 data being returned. It should now handle lines of any length.
2476 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2477 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2480 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2481 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2482 no longer happens in comment lines. A list of replacements is now placed
2483 at the head of all of the source files, except those whose only change is
2484 to replace PERL_COMMAND in the very first #! line.
2486 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2487 messages on the queue, with a bottom-up merge sort, using code contributed
2488 by Michael Haardt. This should make operations like -bp somewhat faster on
2489 large queues. It won't affect queue runners, except when queue_run_in_order
2492 51. Installed eximstats 1.31 in the distribution.
2494 52. Added support for SRV lookups to the dnslookup router.
2496 53. If an ACL referred to $message_body or $message_body_end, the value was not
2497 reset for any messages that followed in the same SMTP session.
2499 54. The store-handling optimization for building very long strings was not
2500 differentiating between the different store pools. I don't think this
2501 actually made any difference in practice, but I've tidied it.
2503 55. While running the routers to verify a sender address, $sender_address
2504 was still set to the sender address. This is wrong, because when routing to
2505 send a bounce to the sender, it would be empty. Therefore, I have changed
2506 it so that, while verifying a sender address, $sender_address is set to <>.
2507 (There is no change to what happens when verifying a recipient address.)
2509 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2510 target A or AAAA records (if not already returned) without resetting the
2511 qualify_single or search_parents options of the DNS resolver. These are
2512 inappropriate in this case because the targets of MX and SRV records must
2513 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2514 target that, when qualified, matched something in the local domain. These
2515 two options are now turned off when doing these lookups.
2517 57. It seems that at least some releases of Reiserfs (which does not have the
2518 concept of a fixed number of inodes) returns zero and not -1 for the
2519 number of available inodes. This interacted badly with check_spool_inodes,
2520 which assumed that -1 was the "no such thing" setting. What I have done is
2521 to check that the total number of inodes is greater than zero before doing
2522 the test of how many are available.
2524 58. When a "warn" ACL statement has a log_message modifier, the message is
2525 remembered, and not repeated. This is to avoid a lot of repetition when a
2526 message has many recipients that cause the same warning to be written.
2527 Howewer, Exim was preserving the list of already written lines for an
2528 entire SMTP session, which doesn't seem right. The memory is now reset if a
2529 new message is started.
2531 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2532 debugging output unless log_rewrite was also set.
2534 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2535 of (int)(handle) when we know that handle contains (void *)(-1).
2537 61. The Exim daemon panic-logs an error return when it closes the incoming
2538 connection. However "connection reset by peer" seems to be common, and
2539 isn't really an error worthy of noting specially, so that particular error
2542 62. When Exim is trying to find all the local interfaces, it used to panic and
2543 die if the ioctl to get the interface flags failed. However, it seems that
2544 on at least one OS (Solaris 9) it is possible to have an interface that is
2545 included in the list of interfaces, but for which you get a failure error
2546 for this call. This happens when the interface is not "plumbed" into a
2547 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2548 failure of the "get flags" call assumes that the interface is down.
2550 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2551 makes life easier for people who are doing arithmetic on fields extracted
2552 from dates, where you often get leading zeros that should not be
2553 interpreted as octal.
2555 64. Added qualify_domain to the redirect router, to override the global
2558 65. If a pathologically long header line contained very many addresses (the
2559 report of this problem mentioned 10 000) and each of them was rewritten,
2560 Exim could use up a very large amount of memory. (It kept on making new
2561 copies of the header line as it rewrote, and never released the old ones.)
2562 At the expense of a bit more processing, the header rewriting function has
2563 been changed so that it no longer eats memory in this way.
2565 66. The generation of the Received: header has been moved from the time that a
2566 message starts to be received, to the time that it finishes. The timestamp
2567 in the Received: header should now be very close to that of the <= log
2568 line. There are two side-effects of this change:
2570 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2571 logged header lines no longer include the local Received: line, because
2572 it has not yet been created. The same applies to a copy of the message
2573 that is returned to a non-SMTP sender when a message is rejected.
2575 (b) When a filter file is tested using -bf, no additional Received: header
2576 is added to the test message. After some thought, I decided that this
2579 This change does not affect the value of $received_for. It is still set
2580 after address rewriting, but before local_scan() is called.
2582 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2584 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2585 gave an unhelpful panic error message, and a defer error. I have managed to
2586 change this behaviour so that it now rejects any supplied certificate,
2587 which seems right, as the list of acceptable certificates is empty.
2589 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2590 gave an unhelpful defer error. I have not managed to make this reject any
2591 supplied certificates, but the error message it gives is "no certificate
2592 supplied", which is not helpful.
2594 70. exigrep's output now also includes lines that are not associated with any
2595 message, but which match the given pattern. Implemented by a patch from
2596 Martin Sluka, which also tidied up the Perl a bit.
2598 71. Recipient callout verification, like sender verification, was using <> in
2599 the MAIL FROM command. This isn't really the right thing, since the actual
2600 sender may affect whether the remote host accepts the recipient or not. I
2601 have changed it to use the actual sender in the callout; this means that
2602 the cache record is now keyed on a recipient/sender pair, not just the
2603 recipient address. There doesn't seem to be a real danger of callout loops,
2604 since a callout by the remote host to check the sender would use <>.
2605 [SEE ABOVE: changed after hitting problems.]
2607 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2608 temporary errors. However, in the case of such a code being given after
2609 the end of a data transmission (i.e. after ".") Exim was failing to write
2610 a retry record for the message. (Yes, there was some broken host that was
2611 actually sending 8xx at this point.)
2613 73. An unknown lookup type in a host list could cause Exim to panic-die when
2614 the list was checked. (An example that provoked this was putting <; in the
2615 middle of a list instead of at the start.) If this happened during a DATA
2616 ACL check, a -D file could be left lying around. This kind of configuration
2617 error no longer causes Exim to die; instead it causes a defer errror. The
2618 incident is still logged to the main and panic logs.
2620 74. Buglet left over from Exim 3 conversion. The message "too many messages
2621 in one connection" was written to the rejectlog but not the mainlog, except
2622 when address rewriting (yes!) was being logged.
2624 75. Added write_rejectlog option.
2626 76. When a system filter was run not as root (that is, when system_filter_user
2627 was set), the values of the $n variables were not being returned to the
2628 main process; thus, they were not subsequently available in the $sn
2631 77. Added +return_path_on_delivery log selector.
2633 78. A connection timeout was being treated differently from recipients deferred
2634 when testing hosts_max_try with a message that was older than the host's
2635 retry timeout. (The host should not be counted, thus allowing all hosts to
2636 be tried at least once before bouncing.) This may have been the cause of an
2637 occasionally reported bug whereby a message would remain on the queue
2638 longer than the retry timeout, but would be bounced if a delivery was
2639 forced. I say "may" because I never totally pinned down the problem;
2640 setting up timeout/retry tests is difficult. See also the next item.
2642 79. The ultimate address timeout was not being applied to errors that involved
2643 a combination of host plus message (for example, a timeout on a MAIL
2644 command). When an address resolved to a number of possible hosts, and they
2645 were not all tried for each delivery (e.g. because of hosts_max_try), a
2646 message could remain on the queue longer than the retry timeout.
2648 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2651 81. Fixed an obscure SMTP outgoing bug which required at least the following
2652 conditions: (a) there was another message waiting for the same server;
2653 (b) the server returned 5xx to all RCPT commands in the first message so
2654 that the message was not completed; (c) the server dropped the connection
2655 or gave a negative response to the RSET that Exim sends to abort the
2656 transaction. The observed case was a dropped connection after DATA that had
2657 been sent in pipelining mode. That is, the server had advertised PIPELINING
2658 but was not implementing it correctly. The effect of the bug was incorrect
2659 behaviour, such as trying another host, and this could lead to a crash.
2665 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2666 and daemon.c were passed as pointers to ints; they should have been
2667 pointers to socklen_t variables (which are typically unsigned ints).
2669 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2672 3. Fixed a really odd bug that affected only the testing scheme; patching a
2673 certain fixed string in the binary changed the value of another string that
2674 happened to be identical to the end of the original first string.
2676 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2677 name", it returns that address as the IP address. On some operating
2678 systems (e.g. Solaris), it also passes back the IP address string as the
2679 "host name". However, on others (e.g. Linux), it passes back an empty
2680 string. Exim wasn't checking for this, and was changing the host name to an
2681 empty string, assuming it had been canonicized.
2683 5. Although rare, it is permitted to have more than one PTR record for a given
2684 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2685 all the names associated with an address, because they do in Solaris.
2686 However, it seems that they do not in Linux for data that comes from the
2687 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2688 I found this out when I moved to a new Linux workstation and tried to run
2689 the Exim test suite.
2691 To get round this problem I have changed the code so that it now does its
2692 own call to the DNS to look up PTR records when searching for a host name.
2693 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2694 addresses that are only in /etc/hosts are still found.
2696 This behaviour is, however, controlled by an option called host_lookup_
2697 order, which defaults to "bydns:byaddr". If people want to use the other
2698 order, or indeed, just use one or the other means of lookup, they can
2699 specify it in this variable.
2701 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2702 some operating systems, this comes back from gethostbyaddr() as an empty
2703 string, and this is what Exim used to test for. However, it seems that in
2704 other systems, "." is yielded. Exim now tests for this case too.
2706 7. The values of check_spool_space and check_log_space are now held internally
2707 as a number of kilobytes instead of an absolute number of bytes. If a
2708 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2709 kilobyte. This means that much larger values can be stored.
2711 8. Exim monitor: an attempt to get the action menu when not actually pointing
2712 at a message produces an empty menu entitled "No message selected". This
2713 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2714 no entries in it ("Shell widget menu has zero width and/or height"). So I
2715 have added a single, blank menu entry in this case.
2717 9. Added ${quote_local_part.
2719 10. MIME decoding is now applied to the contents of Subject: header lines when
2722 11. Now that a reference to $sender_host_address automatically causes a reverse
2723 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2724 host lookup before query-style lookups in lists that might use this
2725 variable. This has therefore been abolished, and the "net-" prefix is no
2726 longer necessary for query-style lookups.
2728 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2729 have been a typo for LFLAGS, so it has been changed.
2731 13. The install script calls Exim with "-C /dev/null" in order to find the
2732 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2733 to be output. Howeve, since Exim outputs its version number before the
2734 error, it didn't break the script. It just looked ugly. I fixed this by
2735 always allowing "-C /dev/null" if the caller is root.
2737 14. Ignore overlarge ACL variable number when reading spool file - insurance
2738 against a later release with more variables having written the file.
2740 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2741 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2742 this, as well as the form without the "IPv6" on the front.
2744 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2745 OS/Makefile-Darwin file.
2747 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2748 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2750 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2751 variable. However, some OS use ints for inodes. I've added cast to long int
2752 to get rid of the compiler warning.
2754 19. I had forgotten to lock out "/../" in configuration file names when
2755 ALT_CONFIG_PREFIX was set.
2757 20. Routers used for verification do not need to specify transports. However,
2758 if such a router generated a host list, and callout was configured, Exim
2759 crashed, because it could not find a port number from the (non-existent)
2760 transport. It now assumes port 25 in this circumstance.
2762 21. Added the -t option to exigrep.
2764 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2765 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2766 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2767 them is set, LOOKUP_LSEARCH is forced.
2769 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2771 24. Added sender and host information to the "rejected by local_scan()" log
2772 line; previously there was no indication of these.
2774 25. Added .include_if_exists.
2776 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2777 Linux. It turns out that not all file systems support this. Apparently some
2778 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2779 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2780 not supported on the file descriptor, is now ignored when Exim is trying to
2781 sync a directory. This applies only to Linux.
2783 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2785 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2786 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2789 29. Added support for maildirsize files from supplied patch (modified a bit).
2791 30. The use of :fail: followed by an empty string could lead Exim to respond to
2792 sender verification failures with (e.g.):
2794 550 Verification failed for <xxx>
2795 550 Sender verify failed
2797 where the first response line was missing the '-' that indicates it is not
2798 the final line of the response.
2800 31. The loop for finding the name of the user that called Exim had a hardwired
2801 limit of 10; it now uses the value of finduser_retries, which is used for
2802 all other user lookups.
2804 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2807 33. Exim was neglecting to zero errno before one call of strtol() when
2808 expanding a string and expecting an integer value. On some systems this
2809 resulted in spurious "integer overflow" errors. Also, it was casting the
2810 result into an int without checking.
2812 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2813 did not work. The code looks as if it has *never* worked, though it appears
2814 to have been documented since at least releast 1.62. I have made it work.
2816 35. The "timeout_DNS" error in retry rules, also documented since at least
2817 1.62, also never worked. As it isn't clear exactly what this means, and
2818 clearly it isn't a major issue, I have abolished the feature by treating it
2819 as "timeout", and writing a warning to the main and panic logs.
2821 36. The display of retry rules for -brt wasn't always showing the error code
2824 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2825 timeout_connect_A, timeout_connect_MX.
2827 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2828 to the empty sender.
2830 39. The daemon was not analysing the content of -oX till after it had closed
2831 stderr and disconnected from the controlling terminal. This meant that any
2832 syntax errors were only noted on the panic log, and the return code from
2833 the command was 0. By re-arranging the code a little, I've made the
2834 decoding happen first, so such errors now appear on stderr, and the return
2835 code is 1. However, the actual setting up of the sockets still happens in
2836 the disconnected process, so errors there are still only recorded on the
2839 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2840 connections (as happens on some IP stacks) was logged at start up time as
2841 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2842 it from "IPv6 and IPv4", which means that two separate sockets are being
2845 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2846 says whether AF_INET or AF_INET6 was passed as an argument.
2848 42. Exiwhat output was messed up when time zones were included in log
2851 43. Exiwhat now gives more information about the daemon's listening ports,
2852 and whether -tls-on-connect was used.
2854 44. The "port" option of the smtp transport is now expanded.
2856 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2857 silently ignored. Now an error message is written to the main and panic
2860 46. There's a new ACL modifier called "logwrite" which writes to a log file
2861 as soon as it is encountered.
2863 47. Added $local_user_uid and $local_user_gid at routing time.
2865 48. Exim crashed when trying to verify a sender address that was being
2868 49. Exim was recognizing only a space character after ".include". It now also
2869 recognizes a tab character.
2871 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2872 extracting the relevant information from the specification. The man page no
2873 longer contains scrambled data for the -d option, and I've added a section
2874 at the front about calling Exim under different names.
2876 51. Added "extra_headers" argument to the "mail" command in filter files.
2878 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2881 53. Installed eximstats 1.29.
2883 54. Added transport_filter_timeout as a generic transport option.
2885 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2886 Cc: header lines. This was required by RFC 822, but it not required by RFC
2889 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2890 incoming messages that did not have them. Now it does so only if the
2891 message originates locally, that is, if there is no associated remote host
2892 address. When Resent- header lines are present, this applies to the Resent-
2893 lines rather than the non-Resent- lines.
2895 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2896 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2898 58. Messages for configuration errors now include the name of the main
2899 configuration file - useful now that there may be more than one file in a
2900 list (.included file names were always shown).
2902 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2903 for those rare installations that do not start the daemon as root or run it
2904 setuid root. I've cut out the call to initgroups() if the daemon is not
2907 60. The Exim user and group can now be bound into the binary as text strings
2908 that are looked up at the start of Exim's processing.
2910 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2912 62. Added $mailstore_basename variable.
2914 63. Installed patch to sieve.c from Michael Haardt.
2916 64. When Exim failed to open the panic log after failing to open the main log,
2917 the original message it was trying to log was written to stderr and debug
2918 output, but if they were not available (the usual case in production), it
2919 was lost. Now it is written to syslog before the two lines that record the
2920 failures to open the logs.
2922 65. Users' Exim filters run in subprocesses under the user's uid. It is
2923 possible for a "deliver" command or an alias in a "personal" command to
2924 provoke an address rewrite. If logging of address rewriting is configured,
2925 this fails because the process is not running as root or exim. There may be
2926 a better way of dealing with this, but for the moment (because 4.30 needs
2927 to be released), I have disabled address rewrite logging when running a
2928 filter in a non-root, non-exim process.
2934 1. The buildconfig auxiliary program wasn't quoting the value set for
2935 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2936 not defined. This bug was masked in 4.22 by the effect that was fixed in
2939 2. Some messages that were rejected after a message id was allocated were
2940 shown as "incomplete" by exigrep. It no longer does this for messages that
2941 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2943 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2944 have allow_domain_literals set, the ID did not get logged in the <= line.
2945 Domain literals are now always recognized in Message-ID: header lines.
2947 4. The first argument for a ${extract expansion item is the key name or field
2948 number. Leading and trailing spaces in this item were not being ignored,
2949 causing some misleading effects.
2951 5. When deliver_drop_privilege was set, single queue runner processes started
2952 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2953 same command in the process it spins off) were not dropping privilege.
2955 6. When the daemon running as "exim" started a queue runner, it always
2956 re-executed Exim in the spun-off process. This is a waste of effort when
2957 deliver_drop_privilege is set. The new process now just calls the
2958 queue-runner function directly.
2964 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2967 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2968 ignored. Though the use of -oP was forcing the writing of a pid file, it
2969 was always written to the default place.
2971 3. If the message "no IP address found for host xxxx" is generated during
2972 incoming verification, it is now followed by identification of the incoming
2973 connection (so you can more easily find what provoked it).
2975 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2977 5. Added some features to "harden" Exim a bit more against certain attacks:
2979 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2980 be put in Local/Makefile. This is like the never_users runtime option,
2981 but it cannot be overridden. The default setting is "root".
2983 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2984 prefix string with which any file named in a -C command line option
2987 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2988 is retained for -C and -D only if the caller of Exim is root. Without
2989 it, the exim user may also use -C and -D and retain privilege.
2991 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2992 command line option is disabled.
2994 6. Macro names set by the -D option must start with an upper case letter, just
2995 like macro names defined in the configuration file.
2997 7. Added "dereference=" facility to LDAP.
2999 8. Two instances of the typo "uknown" in the source files are fixed.
3001 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3002 the Configure-Makefile script screwed up while processing it.
3004 10. Incorporated PCRE 4.4.
3006 11. The SMTP synchronization check was not operating right at the start of an
3007 SMTP session. For example, it could not catch a HELO sent before the client
3008 waited for the greeting. There is now a check for outstanding input at the
3009 point when the greeting is written. Because of the duplex, asynchronous
3010 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3011 way, but not yet received, when the check is performed.
3013 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3014 on TCP/IP sockets, because this apparently causes some broken clients to
3017 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3018 unchanged) from the Cygwin maintainer.
3020 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3021 was supported instead of testing for mbx. Effectively a typo.
3023 15. The spa authenticator server code was not checking that the input it
3024 received was valid base64.
3026 16. The debug output line for the "set" modifier in ACLs was not showing the
3027 name of the variable that was being set.
3029 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3031 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3032 lookup on its own. Something else was needed to trigger the lookup. For
3033 example, a match in host_lookup or the need for a host name in a host list.
3034 Now, if $sender_host_name is referenced and the host name has not yet been
3035 looked up, a lookup is performed. If the lookup fails, the variable remains
3036 empty, and $host_lookup_failed is set to "1".
3038 19. Added "eqi" as a case-independent comparison operator.
3040 20. The saslauthd authentication condition could segfault if neither service
3041 nor realm was specified.
3043 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3044 error message that was logged was misleading, and incoming SMTP
3045 connections were dropped. The message is now more accurate, and temporary
3046 errors are given to SMTP connections.
3048 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3049 (except RSET) with a 421 error, until QUIT is received. However, it was
3050 failing to send a response to QUIT.
3052 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3053 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3054 ACL is now run after the helo_try_verify_hosts code.
3056 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3057 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3058 case-independent, so other case variants are also recognized). Apparently
3059 some systems use these upper case variants.
3061 25. If more than two messages were waiting for the same host, and a transport
3062 filter was specified for the transport, Exim sent two messages over the
3063 same TCP/IP connection, and then failed with "socket operation on non-
3064 socket" when it tried to send the third.
3066 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3068 27. The extern definition of crypt16() in expand.c was not being excluded when
3069 the OS had its own crypt16() function.
3071 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3072 as a preferred synonym for return_size_limit, both as an option and as an
3075 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3077 30. Changed the default configuration ACL to relax the local part checking rule
3078 for addresses that are not in any local domains. For these addresses,
3079 slashes and pipe symbols are allowed within local parts, but the sequence
3080 /../ is explicitly forbidden.
3082 31. SPA server authentication was not clearing the challenge buffer before
3085 32. log_message in a "warn" ACL statement was writing to the reject log as
3086 well as to the main log, which contradicts the documentation and doesn't
3087 seem right (because no rejection is happening). So I have stopped it.
3089 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3090 However, I am unable to do any testing of this.
3092 34. Fixed an infelicity in the appendfile transport. When checking directories
3093 for a mailbox, to see if any needed to be created, it was accidentally
3094 using path names with one or more superfluous leading slashes; tracing
3095 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3097 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3098 contents are added to the log line that is written for every discarded
3099 recipient. (Previously a log_message setting was ignored.)
3101 36. The ${quote: operator now quotes the string if it is empty.
3103 37. The install script runs exim in order to find its version number. If for
3104 some reason other than non-existence or emptiness, which it checks, it
3105 could not run './exim', it was installing it with an empty version number,
3106 i.e. as "exim-". This error state is now caught, and the installation is
3109 38. An argument was missing from the function that creates an error message
3110 when Exim fails to connect to the socket for saslauthd authentication.
3111 This could cause Exim to crash, or give a corrupted message.
3113 39. Added isip, isip4, and isip6 to ${if conditions.
3115 40. The ACL variables $acl_xx are now saved with the message, and can be
3116 accessed later in routers, transports, and filters.
3118 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3119 strings in the file are not string-expanded.
3121 42. If a MAIL command specified a SIZE value that was too large to fit into an
3122 int variable, the check against message_size_limit failed. Such values are
3123 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3124 day this will have to be increased, but I don't think I want to be around
3125 when emails are that large.
3132 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3133 iconv() is not standard in FreeBSD.
3135 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3136 IPv6 enabled. The observed symptom was a segmentation fault on return from
3137 the function os_common_find_running_interfaces() in src/os.c.
3139 3. In the check_special_case() function in daemon.c I had used "errno" as an
3140 argument name, which causes warnings on some systems. This was basically a
3141 typo, since it was named "eno" in the comments!
3143 4. The code that waits for the clock to tick (at a resolution of some fraction
3144 of a second) so as to ensure message-id uniqueness was always waiting for
3145 at least one whole tick, when it could have waited for less. [This is
3146 almost certainly not relevant at current processor speeds, where it is
3147 unlikely to ever wait at all. But we try to future-proof.]
3149 5. The function that sleeps for a time interval that includes fractions of a
3150 second contained a race. It did not block SIGALRM between setting the
3151 timer, and suspending (a couple of lines later). If the interval was short
3152 and the sigsuspend() was delayed until after it had expired, the suspension
3153 never ended. On busy systems this could lead to processes getting stuck for
3156 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3157 process, before it forks any delivery processes. The open lookup caching
3158 mechanism meant that the open file or database connection was passed into
3159 the delivery process. The problem was that delivery processes always tidy
3160 up cached lookup data. This could cause a problem for the next delivery
3161 process started by the queue runner, because the external queue runner
3162 process does not know about the closure. So the next delivery process
3163 still has data in the lookup cache. In the case of a file lookup, there was
3164 no problem because closing a file descriptor in a subprocess doesn't affect
3165 the parent. However, if the lookup was caching a connection to a database,
3166 the connection was closed, and the second delivery process was likely to
3167 see errors such as "PGSQL: query failed: server closed the connection
3168 unexpectedly". The problem has been fixed by closing all cached lookups
3169 in a queue runner before running a delivery process.
3171 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3172 seem to have the "const" qualifier which it has on other OS. I've
3175 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3176 *of the same type* (client or server) with the same public name that an
3177 error should be diagnosed.
3179 9. When Exim looked up a host name for an IP address, but failed to find the
3180 original IP address when looking up the host name (a safety check), it
3181 output the message "<ip address> does not match any IP for NULL", which was
3182 confusing, to say the least. The bug was that the host name should have
3183 appeared instead of "NULL".
3185 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3186 user that is built into the binary, and the -C or -D options is used, root
3187 privilege is dropped before the configuration file is read. In addition,
3188 logging is switched to stderr instead of the normal log files. If the
3189 configuration then re-defines the Exim user, the unprivileged environment
3190 is probably not what is expected, so Exim logs a panic warning message (but
3193 However, if deliver_drop_privilege is set, the unprivileged state may well
3194 be exactly what is intended, so the warning has been cut out in that case,
3195 and Exim is allowed to try to write to its normal log files.
3201 1. smtp_return_error_details was not giving details for temporary sender
3202 or receiver verification errors.
3204 2. Diagnose a configuration error if two authenticators have the same public
3207 3. Exim used not to create the message log file for a message until the first
3208 delivery attempt. This could be confusing when incoming messages were held
3209 for policy or load reasons. The message log file is now created at the time
3210 the message is received, and an initial "Received" line is written to it.
3212 4. The automatically generated man page for command line options had a minor
3213 bug that caused no ill effects; however, a more serious problem was that
3214 the procedure for building the man page automatically didn't always
3215 operate. Consequently, release 4.20 contains an out-of-date version. This
3216 shouldn't happen again.
3218 5. When building Exim with embedded Perl support, the script that builds the
3219 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3220 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3222 6. The freeze_tell option was not being used for messages that were frozen on
3223 arrival, either by an ACL or by local_scan().
3225 7. Added the smtp_incomplete_transaction log selector.
3227 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3228 was accepting AUTH without a new EHLO.
3230 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3231 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3234 10. Logging of TCP/IP connections (when configured) now happens in the main
3235 daemon process instead of the child process, so that the TCP/IP connection
3236 count is more accurate (but it can never be perfect).
3238 11. The use of "drop" in a nested ACL was not being handled correctly in the
3239 outer ACL. Now, if condition failure induced by the nested "drop" causes
3240 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3241 or "require"), the connection is dropped.
3243 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3244 that yield "discard" can now be used with an "accept" or a "discard" verb,
3245 but an error is generated for any others (because I can't see a useful way
3246 to define what should happen).
3248 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3249 are now processed in the same way as lines in the Exim configuration file.
3250 In particular, continuation lines are supported.
3252 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3254 15. Added -ti meaning -t -i.
3256 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3257 domains, and warn by logging if others are found.
3259 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3260 returned from GIFCONF when Exim is trying to find the list of interfaces on
3261 a host. The code in os.c has been modified to copy each ifreq to an aligned
3262 structure in all cases.
3264 Also, in some cases, the returned ifreq's were being copied to a 'struct
3265 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3266 means the last couple of bytes of an IPv6 address could be chopped if the
3267 ifreq contained only a normal sockaddr (14 bytes storage).
3269 18. Named domain lists were not supported in the hosts_treat_as_local option.
3270 An entry such as +xxxx was not recognized, and was treated as a literal
3273 19. Ensure that header lines added by a DATA ACL are included in the reject log
3274 if the ACL subsequently rejects the message.
3276 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3277 MD5 (which is deprecated).
3279 21. When testing a filter file using -bf, Exim was writing a message when it
3280 took the sender from a "From " line in the message, but it was not doing so
3281 when it took $return_path from a Return-Path: header line. It now does.
3283 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3284 with a valid header line field name (a series of printing characters
3285 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3287 23. Changed "disc" in the source to "disk" to conform to the documentation and
3288 the book and for uniformity.
3290 24. Ignore Sendmail's -Ooption=value command line item.
3292 25. When execve() failed while trying to run a command in a pipe transport,
3293 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3294 could be confused with a return value of 69 from the command itself. This
3295 has been changed to 127, the value the shell returns if it is asked to run
3296 a non-existent command. The wording for the related log line suggests a
3297 non-existent command as the problem.
3299 26. If received_header_text expands to an empty string, do not add a Received:
3300 header line to the message. (Well, it adds a token one on the spool, but
3301 marks it "old" so that it doesn't get used or transmitted.)
3303 27. Installed eximstats 1.28 (addition of -nt option).
3305 28. There was no check for failure on the call to getsockname() in the daemon
3306 code. This can fail if there is a shortage of resources on the system, with
3307 ENOMEM, for example. A temporary error is now given on failure.
3309 29. Contrary to the C standard, it seems that in some environments, the
3310 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3311 program. Exim now does this explicitly; it affects the formatting of
3312 timestamps using strftime().
3314 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3315 complaints. I've now initialized all the variables, to avoid this.
3317 32. Header lines added by a system filter were not being "seen" during
3318 transport-time rewrites.
3320 33. The info_callback() function passed to OpenSSL is set up with type void
3321 (*)(SSL *, int, int), as described somewhere. However, when calling the
3322 function (actually a macro) that sets it up, the type void(*)() is
3323 expected. I've put in a cast to prevent warnings from picky compilers.
3325 34. If a DNS black list lookup found a CNAME record, but there were no A
3326 records associated with the domain it pointed at, Exim crashed.
3328 35. If a DNS black list lookup returned more than one A record, Exim ignored
3329 all but the first. It now scans all returned addresses if a particular IP
3330 value is being sought. In this situation, the contents of the
3331 $dnslist_value variable are a list of all the addresses, separated by a
3334 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3335 to accept a host name and all its aliases if the forward lookup for any of
3336 them yielded the IP address of the incoming connection. Now it accepts only
3337 those names whose forward lookup yields the correct IP address. Any other
3338 names are discarded. This closes a loophole whereby a rogue DNS
3339 administrator could create reverse DNS records to break through a
3340 wildcarded host restriction in an ACL.
3342 37. If a user filter or a system filter that ran in a subprocess used any of
3343 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3344 the wrong values were passed to the pipe command ($thisaddress had the
3345 value of $0, $0 had the value of $1, etc). This bug was introduced by
3346 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3348 38. Improved the line breaking for long SMTP error messages from ACLs.
3349 Previously, if there was no break point between 40 and 75 characters, Exim
3350 left the rest of the message alone. Two changes have been made: (a) I've
3351 reduced the minimum length to 35 characters; (b) if it can't find a break
3352 point between 35 and 75 characters, it looks ahead and uses the first one
3353 that it finds. This may give the occasional overlong line, but at least the
3354 remaining text gets split now.
3356 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3357 file descriptors might be low, and that setting 1000 would always raise it.
3358 It turns out that in some environments, the limit is already over 1000 and
3359 that lowering it causes trouble. So now Exim takes care not to decrease it.
3361 40. When delivering a message, the value of $return_path is set to $sender_
3362 address at the start of routing (routers may change the value). By an
3363 oversight, this default was not being set up when an address was tested by
3364 -bt or -bv, which affected the outcome if any router or filter referred to
3367 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3368 log only when "message" or "log_message" are set. However, if one of the
3369 conditions was an address verification, or a call to a nested ACL, the
3370 messages generated by the underlying test were being passed through. This
3371 no longer happens. The underlying message is available in $acl_verify_
3372 message for both "message" and "log_message" expansions, so it can be
3373 passed through if needed.
3375 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3376 new expansion $bh_ to give the encoded byte string without charset
3377 translation. Translation happens only if iconv() is available; HAVE_ICONV
3378 indicates this at build time. HEADERS_CHARSET gives the charset to
3379 translate to; headers_charset can change it in the configuration, and
3380 "headers charset" can change it in an individual filter file.
3382 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3383 that creates RFC 2047 encoded "words" labels them as that charset instead
3384 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3385 expansion operator; (ii) when Exim creates a From: line for a local
3386 message; (iii) when a header line is rewritten to include a "phrase" part.
3388 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3389 buggy, causing it to skip the first lines of messages whose message ID
3390 ended in 'D'. This would not have bitten before Exim release 4.14, because
3391 message IDs were unlikely to end in 'D' before then. The effect was to have
3392 incorrect size information for certain domains.
3394 45. #include "config.h" was missing at the start of the crypt16.c module. This
3395 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3398 46. If there was a timeout during a "random" callout check, Exim treated it as
3399 a failure of the random address, and carried on sending RSET and the real
3400 address. If the delay was just some slowness somewhere, the response to the
3401 original RCPT would be taken as a response to RSET and so on, causing
3402 mayhem of various kinds.
3404 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3405 when I implemented it. It didn't allow for the fact that some option values
3406 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3407 the right test for positive values.
3409 48. Domain names in DNS records are case-independent. Exim always looks them up
3410 in lower case. Some resolvers return domain names in exactly the case they
3411 appear in the zone file, that is, they may contain uppercase letters. Not
3412 all resolvers do this - some return always lower case. Exim was treating a
3413 change of case by a resolver as a change of domain, similar to a widening
3414 of a domain abbreviation. This triggered its re-routing code and so it was
3415 trying to route what was effectively the same domain again. This normally
3416 caused routing to fail (because the router wouldn't handle the domain
3417 twice). Now Exim checks for this case specially, and just changes the
3418 casing of the domain that it ultimately uses when it transmits the message
3421 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3424 50. If a filter generated a file delivery with a non-absolute name (possible if
3425 no home directory exists for the router), the forbid_file option was not
3428 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3429 the existing equality matching.
3431 52. Exim was using ints instead of ino_t variables in some places where it was
3432 dealing with inode numbers.
3434 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3435 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3436 called TMPDIR, and if it finds it is different, it changes its value.
3438 54. The smtp_printf() function is now made available to local_scan() so
3439 additional output lines can be written before returning. There is also an
3440 smtp_fflush() function to enable the detection of a dropped connection.
3441 The variables smtp_input and smtp_batched_input are exported to
3444 55. Changed the default runtime configuration: the message "Unknown user"
3445 has been removed from the ACL, and instead placed on the localuser router,
3446 using the cannot_route_message feature. This means that any verification
3447 failures that generate their own messages won't get overridden. Similarly,
3448 the "Unrouteable address" message that was in the ACL for unverifiable
3449 relay addresses has also been removed.
3451 56. Added hosts_avoid_esmtp to the smtp transport.
3453 57. The exicyclog script was not checking for the esoteric option
3454 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3455 will work only if exicyclog is run under the appropriate euid.
3457 58. Following a discussion on the list, the rules by which Exim recognises line
3458 endings on incoming messages have been changed. The -dropcr and drop_cr
3459 options are now no-ops, retained only for backwards compatibility. The
3460 following line terminators are recognized: LF CRLF CR. However, special
3461 processing applies to CR:
3463 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3464 nor a local message in the state where . is a terminator.
3466 (ii) If a bare CR is encountered in a header line, an extra space is added
3467 after the line terminator so as not to end the header. The reasoning
3468 behind this is that bare CRs in header lines are most likely either
3469 to be mistakes, or people trying to play silly games.
3471 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3472 was being incorrectly given as 18 bytes larger than it should have been.
3473 This is a VOB (very old bug).
3475 60. This may never have affected anything current, but just in case it has:
3476 When the local host is found other than at the start of a list of hosts,
3477 the local host, those with the same MX, and any that follow, are discarded.
3478 When the list in question was part of a longer list of hosts, the following
3479 hosts (not currently being processed) were also being discarded. This no
3480 longer happens. I'm not sure if this situation could ever has previously
3483 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3486 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3487 header. This is something that is recommended in a new Internet Draft, and
3488 is something that is documented as being done by Sendmail. There are two
3489 possible values. For messages generated by the autoreply transport, Exim
3492 Auto-Submitted: auto-replied
3494 whereas for all other generated messages (e.g. bounces) it adds
3496 Auto-Submitted: auto-generated
3498 63. The "personal" condition in filters now includes a test for the
3499 Auto-Submitted: header. If it contains the string "auto-" the message it
3500 not considered personal.
3502 64. Added rcpt_include_affixes as a generic transport option.
3504 65. Added queue_only_override (default true).
3506 66. Added the syslog_duplication option.
3508 67. If what should have been the first header line of a message consisted of
3509 a space followed by a colon, Exim was mis-interpreting it as a header line.
3510 It isn't of course - it is syntactically invalid and should therefore be
3511 treated as the start of the message body. The misbehaviour could have
3512 caused a number of strange effects, including loss of data in subsequent
3513 header lines, and spool format errors.
3515 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3516 client host had authenticated. This control can now be exercised by an ACL
3517 for more flexibility.
3519 69. By default, callouts do not happen when testing with -bh. There is now a
3520 variant, -bhc, which does actually run the callout code, including
3521 consulting and updating the callout cache.
3523 70. Added support for saslauthd authentication, courtesy of Alexander
3526 71. If statvfs() failed on the spool or log directories while checking their
3527 size for availability, Exim confusingly gave the error "space shortage".
3528 Furthermore, in debugging mode it crashed with a floating point exception.
3529 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3530 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3531 serious problem, Exim now writes to the main and panic logs when this
3532 happens, with details of the failure. It then refuses to accept the
3533 incoming message, giving the message "spool directory problem" or "log
3534 directory problem" with a 421 code for SMTP messages.
3536 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3537 0, 1, and 2 exist, because some OS complain for execs without them (see
3538 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3539 descriptors. However, the code omitted to check that the open succeeded,
3540 causing mysterious errors if for some reason the permissions on /dev/null
3541 got screwed. Now Exim writes a message to the main and panic logs, and
3542 bombs out if it can't open /dev/null.
3544 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3545 interact so that it is all more flexible. It is supposed to remain
3546 backwards compatible. Also added extra_local_interfaces.
3548 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3549 to bomb out with an assertion failure - to the client this appears as a
3550 connection drop. This problem occurs in the part of the code that was taken
3551 from the Samba project. Fortunately, the assertion is in a very simple
3552 function, so I have fixed this by reproducing the function inline in the
3553 one place where it is called, and arranging for authentication to fail
3554 instead of killing the process with assert().
3556 75. The SPA client code was not working when the server requested OEM rather
3557 than Unicode encoding.
3559 76. Added code to make require_files with a specific uid setting more usable in
3560 the case where statting the file as root fails - usually a non-root-mounted
3561 NFS file system. When this happens and the failure is EACCES, Exim now
3562 forks a subprocess and does the per-uid checking as the relevant uid.
3564 77. Added process_log_path.
3566 78. If log_file_path was not explicitly set, a setting of check_log_space or
3567 check_log_inodes was ignored.
3569 79. If a space check for the spool or log partitions fails, the incident is now
3570 logged. Of course, in the latter case the data may get lost...
3572 80. Added the %p formatting code to string_format() so that it can be used to
3573 print addresses in debug_print(). Adjusted all the address printing in the
3574 debugging in store.c to use %p rather than %d.
3576 81. There was a concern that a line of code in smtp_in.c could overflow a
3577 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3578 initially expressed, the concern was not well-founded, because trailing
3579 spaces are removed early. However, if the trailing spaces were followed by
3580 a NULL, they did not get removed, so the overflow was possible. Two fixes
3583 (a) I re-wrote the offending code in a cleaner fashion.
3584 (b) If an incoming SMTP command contains a NULL character, it is rejected
3587 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3588 runs initgroups(), so that if the Exim user is in any additional groups,
3589 they will be used during message reception.
3595 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob