1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.268 2005/11/28 10:07:55 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
18 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
37 PH/03 For some reason, the default setting of PATH when running a command from
38 a pipe transport was just "/usr/bin". I have changed it to
41 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
42 anything to be listed in the output from -bV.
44 PH/05 When a filter generated an autoreply, the entire To: header line was
45 quoted in the delivery log line, like this:
47 => >A.N.Other <ano@some.domain> <original@ddress> ...
49 This has been changed so that it extracts the operative address. There
50 may be more than one such address. If so, they are comma-separated, like
53 => >ano@some.domain,ona@other.domain <original@ddress> ...
55 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
56 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
57 not being looked up in the rDNS to get a host name, Exim was showing the
58 IP address twice in Received: lines, even though the IP addresses were
59 identical. For example:
61 Received: from [1.2.3.4] (helo=[1.2.3.4])
63 However, if the real host name was known, it was omitting the HELO data
64 if it matched the actual IP address. This has been tidied up so that it
65 doesn't show the same IP address twice.
67 PH/07 When both +timestamp and +memory debugging was on, the value given by
68 $tod_xxx expansions could be wrong, because the tod_stamp() function was
69 called by the debug printing, thereby overwriting the timestamp buffer.
70 Debugging no longer uses the tod_stamp() function when +timestamp is set.
72 PH/08 When the original message was included in an autoreply transport, it
73 always said "this is a copy of the message, including all the headers",
74 even if body_only or headers_only was set. It now gives an appropriate
77 PH/09 Applied a patch from the Sieve maintainer which:
80 o adds the (disabled) notify extension core
81 o adds some debug output for the result of if/elsif tests
82 o points to the current vacation draft in the documentation
83 and documents the missing references header update
87 o fixes a bug in processing the envelope test (when testing
88 multiple envelope elements, the last element determinted the
91 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
92 Electronic Mail") by including:
94 Auto-submitted: auto-generated
96 in the messages that it generates (bounce messages and others, such as
97 warnings). In the case of bounce messages for non-SMTP mesages, there was
98 also a typo: it was using "Auto_submitted" (underscore instead of
99 hyphen). Since every message generated by Exim is necessarily in response
100 to another message, thes have all been changed to:
102 Auto-Submitted: auto-replied
104 in accordance with these statements in the RFC:
106 The auto-replied keyword:
108 - SHOULD be used on messages sent in direct response to another
109 message by an automatic process,
111 - MUST NOT be used on manually-generated messages,
113 - MAY be used on Delivery Status Notifications (DSNs) and Message
114 Disposition Notifications (MDNs),
116 - MUST NOT be used on messages generated by automatic or periodic
117 processes, except for messages which are automatic responses to
120 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
121 to the default Received: header definition.
123 PH/12 Added log selector acl_warn_skipped (default on).
125 PH/13 After a successful wildlsearch lookup, discard the values of numeric
126 variables because (a) they are in the wrong storage pool and (b) even if
127 they were copied, it wouldn't work properly because of the caching.
129 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
130 checking when decoding. Apparently there are clients that generate
131 overlong encoded strings. Why am I not surprised?
133 PH/15 If the first argument of "${if match_address" was not empty, but did not
134 contain an "@" character, Exim crashed. Now it writes a panic log message
135 and treats the condition as false.
137 PH/16 In autoreply, treat an empty string for "once" the same as unset.
139 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
140 extension "envelope-auth". The code is finished and in agreement with
141 other implementations, but there is no documentation so far and in fact,
142 nobody wrote the draft yet. This extension is currently #undef'ed, thus
143 not changing the active code.
145 Print executed "if" and "elsif" statements when debugging is used. This
146 helps a great deal to understand what a filter does.
148 Document more things not specified clearly in RFC3028. I had all this
149 sorted out, when out of a sudden new issues came to my mind. Oops."
151 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
154 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
155 canonical form (as documented). However, after a host name lookup from
156 the IP address, check_host() was doing a simple string comparison with
157 addresses acquired from the DNS when checking that the found name did
158 have the original IP as one of its addresses. Since any found IPv6
159 addresses are likely to be in abbreviated form, the comparison could
160 fail. Luckily, there already exists a function for doing the comparison
161 by converting both addresses to binary, so now that is used instead of
164 PH/20 There was another similar case to PH/19, when a complete host name was
165 given in a host list; looking up its IP address could give an abbreviated
166 form, whereas the current host's name might or might not be abbreviated.
167 The same fix has been applied.
173 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
174 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
177 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
178 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
180 PH/03 Typo: missing ".o" in src/pcre/Makefile.
182 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
183 header line, restrict the check to what is listed in RFCs 2369 and 2929.
184 Also, for "Auto-Submitted", treat anything other than "no" as
185 non-personal, in accordance with RFC 3834. (Previously it treated
186 anything starting "auto-" as non-personal.)
188 TF/01 The control=submission/name=... option had a problem with syntax
189 errors if the name included a slash character. The /name= option
190 now slurps the rest of the string, so it can include any characters
191 but it must come last in the list of options (after /sender_retain
194 PH/05 Some modifications to the interface to the fake nameserver for the new
197 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
198 into the default Received: header string.
205 TK/01 Added the "success_on_redirect" address verification option. See
206 NewStuff for rationale and an example.
208 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
210 PH/02 Patch to exigrep to allow it to work on syslog lines.
212 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
213 fread() to skip over the body file's header line, because in Cygwin the
214 header line is locked and is inaccessible.
216 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
217 co-exist for some time) to make it clear that it is the Exim ID that is
218 referenced, not the Message-ID: header line.
220 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
221 string_format() function, because snprintf() does not exist on all
224 PH/06 The use of forbid_filter_existstest now also locks out the use of the
225 ${stat: expansion item.
227 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
228 protocol synchronization error", to keep the pedants happy.
230 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
231 well as for IRIX systems, when gcc is being used. See the host.c source
234 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
236 PH/10 Named domain lists were not working if used in a queue_smtp_domains
239 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
240 transport and to the smtp transport in LMTP mode.
242 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
244 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
245 run a filter in a subprocess. This could lead to confusion in subsequent
246 lookups in the parent process. There should also be a search_tidyup() at
247 the end of the subprocess.
249 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
250 only if the host matched helo_try_verify_hosts, which caused the
251 verification to occur when the EHLO/HELO command was issued. The ACL just
252 tested the remembered result. Now, if a previous verification attempt has
253 not happened, "verify = helo" does it there and then.
255 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
257 TK/03 Fix log output including CR from clamd.
259 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
260 exist provoked a memory error which could cause a segfault.
262 PH/15 Installed PCRE 6.2
264 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
266 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
267 of the problem. Specifically, suggested +O2 rather than +O1 for the
270 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
272 PH/20 If a delivery was routed to a non-standard port by means of an SRV
273 record, the port was not correctly logged when the outgoing_port log
274 selector was set (it logged the transort's default port).
276 PH/21 Added support for host-specific ports to manualroute, queryprogram,
277 fallback_hosts, and "hosts" in the smtp transport.
279 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
280 host errors such as "Connection refused".
282 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
283 authentication with radiusclient 0.4.9:
285 - Error returned from rc_read_config was caught wrongly
286 - Username/password not passed on to radius server due to wrong length.
288 The presumption is that some radiusclient API changes for 4.51/PH/17
289 were not taken care of correctly. The code is still untested by me (my
290 Linux distribution still has 0.3.2 of radiusclient), but it was
291 contributed by a Radius user.
293 PH/24 When doing a callout, the value of $domain wasn't set correctly when
294 expanding the "port" option of the smtp transport.
296 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
297 while reading a MIME header. Thanks to Tom Hughes for a patch.
299 PH/24 Include config.h inside local_scan.h so that configuration settings are
302 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
303 that in an ACL for RCPT (for example), you can examine exactly what was
306 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
307 commands, but it was not correctly comparing the address with the actual
308 client host address. Thus, it would show the EHLO address in Received:
309 header lines when this was not necessary.
311 PH/27 Added the % operator to ${eval:}.
313 PH/28 Exim tries to create and chdir to its spool directory when it starts;
314 it should be ignoring failures (because with -C, for example, it has lost
315 privilege). It wasn't ignoring creation failures other than "already
318 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
321 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
322 because an input line was too long, either on its own, or by virtue of
323 too many continuations, the temporary file was not being removed, and the
324 return code was incorrect.
326 PH/31 Missing "BOOL" in function definition in filtertest.c.
328 PH/32 Applied Sieve patches from the maintainer.
330 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
332 PH/33 Added "verify = not_blind".
334 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
335 Local/Makefile (with some defaults set). These are used in built scripts
336 such as exicyclog, but they have never been used in the exim_install
337 script (though there are many overriding facilities there). I have
338 arranged that the exim_install script now takes note of these two
341 PH/35 Installed configuration files for Dragonfly.
343 PH/36 When a locally submitted message by a trusted user did not contain a
344 From: header, and the sender address was obtained from -f or from an SMTP
345 MAIL command, and the trusted user did not use -F to supply a sender
346 name, $originator_name was incorrectly used when constructing a From:
347 header. Furthermore, $originator_name was used for submission mode
348 messages from external hosts without From: headers in a similar way,
349 which is clearly wrong.
351 PH/37 Added control=suppress_local_fixups.
353 PH/38 When log_selector = +received_sender was set, and the addition of the
354 sender made the log line's construction buffer exactly full, or one byte
355 less than full, an overflow happened when the terminating "\n" was
358 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
359 when the result of a list match is failure because a DNS lookup failed.
361 PH/40 RM_COMMAND is now used in the building process.
363 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
364 the "build-* directories that it finds.
366 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
367 address in a domain literal was a prefix of an interface address.
369 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
370 when verifying a sender address, unless rewrite_headers is false.
372 PH/44 Wrote a long comment about why errors_to addresses are verified as
373 recipients, not senders.
375 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
376 the ratelimit ACL was added.
378 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
380 PH/46 Added extra information about PostgreSQL errors to the error string.
382 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
383 avoiding the need to install special zones in a real server. This is
384 backwards compatible; if it can't find the fake resolver, it drops back.
385 Thus, both old and new test suites can be run.
387 TF/02 Added util/ratelimit.pl
389 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
390 clock is set back in time.
392 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
393 Candler <B.Candler@pobox.com>.
395 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
396 OK for addresses that are the result of redirections.
398 PH/48 A number of further additions for the benefit of the new test suite,
399 including a fake gethostbyname() that interfaces to the fake DNS resolver
402 TF/06 The fix for widen_domains has also been applied to qualify_single and
403 search_parents which are the other dnslookup options that can cause
406 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
409 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
411 TF/07 Exim produced the error message "an SRV record indicated no SMTP
412 service" if it encountered an MX record with an empty target hostname.
413 The message is now "an MX or SRV record indicated no SMTP service".
415 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
416 if the DNS of the sending site is misconfigured. This is quite a
417 common situation. This change restores the behaviour of treating a
418 helo verification defer as a failure.
420 PH/51 If self=fail was set on a router, the bounce message did not include the
421 actual error message.
427 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
429 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
430 command itself ended in error, the underlying message about the transport
431 filter timeout was being overwritten with the pipe command error. Now the
432 underlying error message should be appended to the second error message.
434 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
436 PH/02 Reduce the amount of output that "make" produces by default. Full output
437 can still be requested.
439 PH/03 The warning log line about a condition test deferring for a "warn" verb
440 was being output only once per connection, rather than after each
441 occurrence (because it was using the same function as for successful
442 "warn" verbs). This seems wrong, so I have changed it.
444 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
445 it should not have, which might have caused a crash in the right
446 circumstances, but probably never did.
448 PH/04 Installed a modified version of Tony Finch's patch to make submission
449 mode fix the return path as well as the Sender: header line, and to
450 add a /name= option so that you can make the user's friendly name appear
453 TF/03 Added the control = fakedefer ACL modifier.
455 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
456 Mark Lowes for thorough testing.
458 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
460 TK/03 Merged latest SRS patch from Miles Wilton.
462 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
463 with the definition in sysexits.h (which is #included earlier).
464 Fortunately, Exim does not actually use EX_OK. The code used to try to
465 preserve the sysexits.h value, by assumimg that macro definitions were
466 scanned for macro replacements. I have been disabused of this notion,
467 so now the code just undefines EX_OK before #including unistd.h.
469 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
470 in the smtp transport. When a block could not be written in a single
471 write() function, the timeout was being re-applied to each part-write.
472 This seems wrong - if the receiver was accepting one byte at a time it
473 would take for ever. The timeout is now adjusted when this happens. It
474 doesn't have to be particularly precise.
476 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
477 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
479 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
480 without a domain if the check to <postmaster@domain> fails.
482 SC/01 Eximstats: added -xls and the ability to specify output files
483 (patch written by Frank Heydlauf).
485 SC/02 Eximstats: use FileHandles for outputing results.
487 SC/03 Eximstats: allow any combination of xls, txt, and html output.
489 SC/04 Eximstats: fixed display of large numbers with -nvr option
491 SC/05 Eximstats: fixed merging of reports with empty tables.
493 SC/06 Eximstats: added the -include_original_destination flag
495 SC/07 Eximstats: removed tabs and trailing whitespace.
497 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
499 TK/06 MBOX spool code: Add real "From " MBOX separator line
500 so the .eml file is really in mbox format (even though
501 most programs do not really care). Patch from Alex Miller.
503 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
504 The latter is generated from $received_to and is only set if the
505 message has one envelope recipient. SA can use these headers,
506 obviously out-of-the-box. Patch from Alex Miller.
508 PH/08 The ${def test on a variable was returning false if the variable's
509 value was "0", contrary to what the specification has always said!
510 The result should be true unless the variable is empty.
512 PH/09 The syntax error of a character other than { following "${if
513 def:variable_name" (after optional whitespace) was not being diagnosed.
514 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
515 accidental colon was present, for example, could give incorrect results.
517 PH/10 Tidied the code in a number of places where the st_size field of a stat()
518 result is used (not including appendfile, where other changes are about
521 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
522 This involved changing a lot of size variables from int to off_t. It
523 should work with maildirs and everything.
525 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
526 spamd dying while we are connected to it.
528 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
531 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
532 function for BDB 4.3.
534 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
536 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
537 the sources for PCRE have been re-arranged and more files are now
540 PH/15 The code I had for printing potentially long long variables in PH/11
541 above was not the best (it lost precision). The length of off_t variables
542 is now inspected at build time, and an appropriate printing format (%ld
543 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
544 to be "long long int" or "long int". This is needed for the internal
545 formatting function string_vformat().
547 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
548 the configuration file to be ":syslog", then the script "guesses" where
549 the logs files are, rather than using the compiled in default. In our
550 case the guess is not the same as the compiled default, so the script
551 suddenly stopped working when I started to use syslog. The patch checks
552 to see if log_file_path is "". If so, it attempts to read it from exim
553 with no configuration file to get the compiled in version, before it
554 falls back to the previous guessing code."
556 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
557 implementing BATV in an Exim configuration. See NewStuff for the gory
560 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
561 Makefile that are specific to HP-UX.
563 PH/18 If the "use_postmaster" option was set for a recipient callout together
564 with the "random" option, the postmaster address was used as the MAIL
565 FROM address for the random test, but not for the subsequent recipient
566 test. It is now used for both.
568 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
569 patch removes a few documentation additions to RFC 3028, because the
570 latest draft now contains them. It adds the new en;ascii-case comparator
571 and a new error check for 8bit text in MIME parts. Comparator and
572 require names are now matched exactly. I enabled the subaddress
573 extension, but it is not well tested yet (read: it works for me)."
575 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
576 rework some of the code of TK/09 above to avoid the hardwired use of
577 "%lld" and "long long". Replaced the call to snprintf() with a call to
580 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
581 records point to non-existent hosts", "retry timeout exceeded", and
582 "retry time not reached for any host after a long failure period".
584 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
585 experimental DomainKeys support:
587 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
588 (2) On an error such as an illegally used "control", the wrong name for
589 the control was given.
591 These problems did NOT occur unless DomainKeys support was compiled.
593 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
595 PH/24 Added ${if match_ip condition.
597 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
598 databases so that it will be absolutely obvious if a crash occurs in the
599 DB library. This is a regular occurrence (often caused by mis-matched
602 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
603 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
604 picked up on a user's system that detects such things. There doesn't seem
605 to be a gcc warning option for this - only an attribute that has to be
606 put on the function's prototype. It seems that in Fedora Core 4 they have
607 set this on a number of new functions. No doubt there will be more in due
610 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
611 specify a transport. However, if an address that was verified by such a
612 router was the subject of a callout, Exim crashed because it tried to
613 read the rcpt_include_affixes from the non-existent transport. Now it
614 just assumes that the setting of that option is false. This bug was
615 introduced by 4.51/PH/31.
617 PH/28 Changed -d+all to exclude +memory, because that information is very
618 rarely of interest, but it makes the output a lot bigger. People tend to
619 do -d+all out of habit.
621 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
622 code in os-type was giving problems when libc.so lives in lib64, like on
625 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
626 aren't the modern standard, and it seems that some systems' include files
627 don't always have them. Exim was already checking for some of the newer
628 ones like T_AAAA, and defining it itself. I've added checks for all the
629 record types that Exim uses.
631 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
632 not automatically generating a new one, as it is supposed to. This
633 prevented TLS from working. If the file did exist, but contained invalid
634 data, a new version was generated, as expected. It was only the case of a
635 non-existent file that was broken.
637 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
638 with a change in libdomainkeys > 0.64.
640 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
641 from DNS. If the selector record carries the flag, it now has
642 precedence over the domain-wide flag.
644 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
646 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
647 the use of a port name that isn't defined in /etc/services) occurred, the
648 message was deferred as in a normal delivery, and thus remained on the
649 spool, instead of being failed because of the mua_wrapper setting. This
650 is now fixed, and I tidied up some of the mua_wrapper messages at the
653 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
654 the messages in a hash of arrays rather than using individual hashes.
655 This is a bit cleaner and results in dramatic memory savings, albeit
656 at a slight CPU cost.
658 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
659 as requested by Marc Sherman.
661 SC/10 Eximstats: added histograms for user specified patterns as requested
664 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
666 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
667 fopen() in the content-scanning modules that did not already have it.
673 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
674 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
676 TK/02 Fix ACL "control" statement not being available in MIME ACL.
678 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
680 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
681 to test Sieve filters that use "vacation".
683 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
684 that changes the way the GnuTLS parameters are stored in the cache file.
685 The new format can be generated externally. For backward compatibility,
686 if the data in the cache doesn't make sense, Exim assumes it has read an
687 old-format file, and it generates new data and writes a new file. This
688 means that you can't go back to an older release without removing the
691 PH/03 A redirect router that has both "unseen" and "one_time" set does not
692 work if there are any delivery delays because "one_time" forces the
693 parent to be marked "delivered", so its unseen clone is never tried
694 again. For this reason, Exim now forbids the simultaneous setting of
697 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
698 redirected to themselves ("homonym" addresses). Read the long ChangeLog
699 entry if you want to know the details. The fix, however, neglected to
700 consider the case when local delivery batching is involved. The test for
701 "previously delivered" was not happening when checking to see if an
702 address could be batched with a previous (undelivered) one; under
703 certain circumstances this could lead to multiple deliveries to the same
706 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
707 in its include files, and this causes problems building Exim.
709 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
710 header_syntax) but Exim was just ignoring anything given after a slash.
711 In particular, this caused confusion with an attempt to use "verify =
712 reverse_host_lookup/defer_ok". An error is now given when options are
713 supplied for verify items that do not have them. (Maybe reverse_host_
714 lookup should have a defer_ok option, but that's a different point.)
716 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
717 defined by RFC 821) to 2048, because there were problems with some AUTH
718 commands, and RFC 1869 says the size should be increased for extended
719 SMTP commands that take arguments.
721 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
724 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
725 "unknown" error; now it says that the functionality isn't in the binary.
727 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
728 an address' error message when a string expansion fails (syntax or
729 whatever). Otherwise the password may appear in the log. Following change
730 PH/42 below, there is no longer a chance of it appearing in a bounce
733 PH/11 Installed exipick version 20050225.0 from John Jetmore.
735 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
736 of its addresses was ever tried. (Bugzilla bug #2.)
738 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
739 the result incorrectly in the debug output. (It correctly added a newline
740 to what was transported.)
742 TF/01 Added $received_time.
744 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
745 commented out examples of how to interface to a virus scanner and to
746 SpamAssassin. Also added commented examples of av_scanner and
747 spamd_address settings.
749 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
750 and controls are allowed in which ACLs. There were a couple of minor
751 errors. Some of the entries in the conditions table (which is a table of
752 where they are NOT allowed) were getting very unwieldy; rewrote them as a
753 negation of where the condition IS allowed.
755 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
757 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
758 header file does not have a version number, so I've had to invent a new
759 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
760 API. The code is untested by me (my Linux distribution still has 0.3.2 of
761 radiusclient), but it was contributed by a Radius user.
763 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
764 files or directories, for OpenSSL.
766 PH/19 When an Exim process that is running as root has to create an Exim log
767 file, it does so in a subprocess that runs as exim:exim so as to get the
768 ownership right at creation (otherwise, other Exim processes might see
769 the file with the wrong ownership). There was no test for failure of this
770 fork() call, which would lead to the process getting stuck as it waited
771 for a non-existent subprocess. Forks do occasionally fail when resources
772 run out. I reviewed all the other calls to fork(); they all seem to check
775 PH/20 When checking for unexpected SMTP input at connect time (before writing
776 the banner), Exim was not dealing correctly with a non-positive return
777 from the read() function. If the client had disconnected by this time,
778 the result was a log entry for a synchronization error with an empty
779 string after "input=" when read() returned zero. If read() returned -1
780 (an event I could not check), uninitialized data bytes were printed.
781 There were reports of junk text (parts of files, etc) appearing after
784 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
786 PH/22 Added support for macro redefinition, and (re)definition in between
787 driver and ACL definitions.
789 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
790 forgetting to use the resulting value; it was using the unexpanded value.
792 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
793 hadn't been configured. The fix is from Juergen Kreileder, who
794 understands it better than I do:
796 "Here's what I see happening with three configured cyrus_sasl
797 authenticators configured (plain, login, cram-md5):
799 On startup auth_cyrus_sasl_init() gets called for each of these.
800 This means three calls to sasl_listmech() without a specified mech_list.
801 => SASL tests which mechs of all available mechs actually work
802 => three warnings about OTP not working
803 => the returned list contains: plain, login, cram-md5, digest-md5, ...
805 With the patch, sasl_listmech() also gets called three times. But now
806 SASL's mech_list option is set to the server_mech specified in the the
807 authenticator. Or in other words, the answer from sasl_listmech()
808 gets limited to just the mech you're testing for (which is different
810 => the return list contains just 'plain' or 'login', 'cram-md5' or
811 nothing depending on the value of ob->server_mech.
813 I've just tested the patch: Authentication still works fine,
814 unavailable mechs specified in the exim configuration are still
815 caught, and the auth.log warnings about OTP are gone."
817 PH/25 When debugging is enabled, the contents of the command line are added
818 to the debugging output, even when log_selector=+arguments is not
821 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
822 answer is "GNU", and only if the return is "GNU/something" is the answer
825 PH/27 $acl_verify_message is now set immediately after the failure of a
826 verification in an ACL, and so is available in subsequent modifiers. In
827 particular, the message can be preserved by coding like this:
829 warn !verify = sender
830 set acl_m0 = $acl_verify_message
832 Previously, $acl_verify_message was set only while expanding "message"
833 and "log_message" when a very denied access.
835 PH/28 Modified OS/os.c-Linux with
837 -#ifndef OS_LOAD_AVERAGE
838 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
840 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
841 nomenclature these days.)
843 PH/29 Installed patch from the Sieve maintainer that adds the options
844 sieve_useraddress and sieve_subaddress to the redirect router.
846 PH/30 In these circumstances:
847 . Two addresses routed to the same list of hosts;
848 . First host does not offer TLS;
849 . First host accepts first address;
850 . First host gives temporary error to second address;
851 . Second host offers TLS and a TLS session is established;
852 . Second host accepts second address.
853 Exim incorrectly logged both deliveries with the TLS parameters (cipher
854 and peerdn, if requested) that were in fact used only for the second
857 PH/31 When doing a callout as part of verifying an address, Exim was not paying
858 attention to any local part prefix or suffix that was matched by the
859 router that accepted the address. It now behaves in the same way as it
860 does for delivery: the affixes are removed from the local part unless
861 rcpt_include_affixes is set on the transport.
863 PH/32 Add the sender address, as F=<...>, to the log line when logging a
864 timeout during the DATA phase of an incoming message.
866 PH/33 Sieve envelope tests were broken for match types other than :is. I have
867 applied a patch sanctioned by the Sieve maintainer.
869 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
870 the uid or gid is negative. A case of a negative gid caused this to be
871 noticed. The fix allows for either to be negative.
873 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
874 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
877 PH/36 The change PH/12 above was broken. Fixed it.
879 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
880 the grounds that routing the same address twice would always produce the
881 same answer. This might have been true once, but it is certainly no
882 longer true now. Routing a child address may depend on the previous
883 routing that produced that child. Some complicated redirection strategies
884 went wrong when messages had multiple recipients, and made Exim's
885 behaviour dependent on the order in which the addresses were given.
887 I have moved the duplicate checking until after the routing is complete.
888 Exim scans the addresses that are assigned to local and remote
889 transports, and removes any duplicates. This means that more work will be
890 done, as duplicates will always all be routed, but duplicates are
891 presumably rare, so I don't expect this is of any significance.
893 For deliveries to pipes, files, and autoreplies, the duplicate checking
894 still happens during the routing process, since they are not going to be
897 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
898 It corrects a timeout issue with spamd. This is Ian's comment: "The
899 background is that sometimes spamd either never reads data from a
900 connection it has accepted, or it never writes response data. The exiscan
901 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
902 blindly assumes that writes won't block so it may never time out."
904 PH/39 Allow G after quota size as well as K and M.
906 PH/40 The value set for $authenticated_id in an authenticator may not contain
907 binary zeroes or newlines because the value is written to log lines and
908 to spool files. There was no check on this. Now the value is run through
909 the string_printing() function so that such characters are converted to
910 printable escape sequences.
912 PH/41 $message_linecount is a new variable that contains the total number of
913 lines in the message. Compare $body_linecount, which is the count for the
916 PH/42 Exim no longer gives details of delivery errors for specific addresses in
917 bounce and delay warning messages, except in certain special cases, which
920 (a) An SMTP error message from a remote host;
921 (b) A message specified in a :fail: redirection;
922 (c) A message specified in a "fail" command in a system filter;
923 (d) A message specified in a FAIL return from the queryprogram router;
924 (e) A message specified by the cannot_route_message router option.
926 In these cases only, Exim does include the error details in bounce and
927 warning messages. There are also a few cases where bland messages such
928 as "unrouteable address" or "local delivery error" are given.
930 PH/43 $value is now also set for the "else" part of a ${run expansion.
932 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
933 being worked on, but at least Exim now implements the latest version to
936 PH/45 In a pipe transport, although a timeout while waiting for the pipe
937 process to complete was treated as a delivery failure, a timeout while
938 writing the message to the pipe was logged, but erroneously treated as a
939 successful delivery. Such timeouts include transport filter timeouts. For
940 consistency with the overall process timeout, these timeouts are now
941 treated as errors, giving rise to delivery failures by default. However,
942 there is now a new Boolean option for the pipe transport called
943 timeout_defer, which, if set TRUE, converts the failures into defers for
944 both kinds of timeout. A transport filter timeout is now identified in
947 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
948 systems where "make" and "gmake" are different, calling "gmake" at top
949 level broke things. I've arranged for the value of $(MAKE) to be passed
950 from the Makefile to this script so that it can call the same version of
954 A note about Exim versions 4.44 and 4.50
955 ----------------------------------------
957 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
958 changes of various kinds. As a consequence, a big documentation update was
959 needed. This delayed the release for rather longer than seemed good, especially
960 in the light of a couple of (minor) security issues. Therefore, the changes
961 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
962 release. So 4.44 and 4.50 are in effect two different branches that both start
965 I have left the 4.50 change log unchanged; it contains all the changes since
966 4.43. The change log for 4.44 is below; many of its items are identical to
967 those for 4.50. This seems to be the most sensible way to preserve the
968 historical information.
974 1. Minor wording change to the doc/README.SIEVE file.
976 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
977 computation of the current number of files was incorrect.
979 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
980 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
983 4. Give more explanation in the error message when the command for a transport
984 filter fails to execute.
986 5. There are several places where Exim runs a non-Exim command in a
987 subprocess. The SIGUSR1 signal should be disabled for these processes. This
988 was being done only for the command run by the queryprogram router. It is
989 now done for all such subprocesses. The other cases are: ${run, transport
990 filters, and the commands run by the lmtp and pipe transports.
992 6. Added CONFIGURE_GROUP build-time option.
994 7. Some older OS have a limit of 256 on the maximum number of file
995 descriptors. Exim was using setrlimit() to set 1000 as a large value
996 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
997 systems. I've change it so that if it can't get 1000, it tries for 256.
999 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1000 was an oversight, and furthermore, ever since the addition of extra
1001 controls (e.g. 4.43/32), the checks on when to allow different forms of
1002 "control" were broken. There should now be diagnostics for all cases when a
1003 control that does not make sense is encountered.
1005 9. Added the /retain_sender option to "control=submission".
1007 10. $recipients is now available in the predata ACL (oversight).
1009 11. Tidy the search cache before the fork to do a delivery from a message
1010 received from the command line. Otherwise the child will trigger a lookup
1011 failure and thereby defer the delivery if it tries to use (for example) a
1012 cached ldap connection that the parent has called unbind on.
1014 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1015 of $address_data from the recipient verification was clobbered by the
1016 sender verification.
1018 13. The value of address_data from a sender verification is now available in
1019 $sender_address_data in subsequent conditions in the ACL statement.
1021 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1023 15. Added a new option "connect=<time>" to callout options, to set a different
1026 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1027 was its contents. (It was OK if the option was not defined at all.)
1029 17. A "Completed" log line is now written for messages that are removed from
1030 the spool by the -Mrm option.
1032 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1033 information about exactly what failed.
1035 19. Added -dd to debug only the daemon process.
1037 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1038 handles timeouts, both on the server side and network timeouts. Renamed the
1039 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1041 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1042 It is now set to "smtps".
1044 22. $host_address is now set to the target address during the checking of
1045 ignore_target_hosts.
1047 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1048 being passed; this would have caused $sender_host_name to have been used if
1049 matching the list had actually called for a host name (not very likely,
1050 since this list is usually IP addresses). A host name is now passed as
1053 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1054 code that specifies a non-restarting handler (typically sigaction() in
1055 modern systems) in an attempt to fix a rare and obscure crash bug.
1057 25. Narrowed the window for a race in the daemon that could cause it to ignore
1058 SIGCHLD signals. This is not a major problem, because they are used only to
1059 wake it up if nothing else does.
1061 26. A malformed maildirsize file could cause Exim to calculate negative values
1062 for the mailbox size or file count. Odd effects could occur as a result.
1063 The maildirsize information is now recalculated if the size or filecount
1066 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1067 support for a long time. Removed HAVE_SYS_VFS_H.
1069 28. Installed the latest version of exipick from John Jetmore.
1071 29. In an address list, if the pattern was not a regular expression, an empty
1072 subject address (from a bounce message) matched only if the pattern was an
1073 empty string. Non-empty patterns were not even tested. This was the wrong
1074 because it is perfectly reasonable to use an empty address as part of a
1075 database query. An empty address is now tested by patterns that are
1076 lookups. However, all the other forms of pattern expect the subject to
1077 contain a local part and a domain, and therefore, for them, an empty
1078 address still always fails if the pattern is not itself empty.
1080 30. Exim went into a mad DNS loop when attempting to do a callout where the
1081 host was specified on an smtp transport, and looking it up yielded more
1082 than one IP address.
1084 31. Re-factored the code for checking spool and log partition space into a
1085 function that finds that data and another that does the check. The former
1086 is then used to implement four new variables: $spool_space, $log_space,
1087 $spool_inodes, and $log_inodes.
1089 32. The RFC2047 encoding function was originally intended for short strings
1090 such as real names; it was not keeping to the 75-character limit for
1091 encoded words that the RFC imposes. It now respects the limit, and
1092 generates multiple encoded words if necessary. To be on the safe side, I
1093 have increased the buffer size for the ${rfc2047: expansion operator from
1096 33. It is now permitted to omit both strings after an "if" condition; if the
1097 condition is true, the result is "true". As before, when the second string
1098 is omitted, a false condition yields an empty string. This makes it less
1099 cumbersome to write custom ACL and router conditions.
1101 34. Failure to deliver a bounce message always caused it to be frozen, even if
1102 there was an errors_to setting on the router. The errors_to setting is now
1105 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1106 canonical form (fully expanded) before being placed in
1107 $sender_host_address.
1109 36. The table in the code that translates DNS record types into text (T_A to
1110 "A" for instance) was missing entries for NS and CNAME. It is just possible
1111 that this could have caused confusion if both these types were looked up
1112 for the same domain, because the text type is used as part of Exim's
1113 per-process caching. But the chance of anyone hitting this buglet seems
1116 37. The dnsdb lookup has been extended in a number of ways.
1118 (1) There is a new type, "zns", which walks up the domain tree until it
1119 finds some nameserver records. It should be used with care.
1121 (2) There is a new type, "mxh", which is like "mx" except that it returns
1122 just the host names, not the priorities.
1124 (3) It is now possible to give a list of domains (or IP addresses) to be
1125 looked up. The behaviour when one of the lookups defers can be
1126 controlled by a keyword.
1128 (4) It is now possible to specify the separator character for use when
1129 multiple records are returned.
1131 38. The dnslists ACL condition has been extended: it is now possible to supply
1132 a list of IP addresses and/or domains to be looked up in a particular DNS
1135 39. Added log_selector=+queue_time_overall.
1137 40. When running the queue in the test harness, wait just a tad after forking a
1138 delivery process, to get repeatability of debugging output.
1140 41. Include certificate and key file names in error message when GnuTLS fails
1141 to set them up, because the GnuTLS error message doesn't include the name
1142 of the failing file when there is a problem reading it.
1144 42. Allow both -bf and -bF in the same test run.
1146 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1148 44. The "Exiscan patch" is now merged into the mainline Exim source.
1150 45. Sometimes the final signoff response after QUIT could fail to get
1151 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1152 < 0 before doing a fflush(). This bug looks as though it goes back to the
1153 introduction of TLS in release 3.20, but "sometimes" must have been rare
1154 because the tests only now provoked it.
1156 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1157 (this can affect the format of dates).
1159 47. exim_tidydb, when checking for the continued existence of a message for
1160 which it has found a message-specific retry record, was not finding
1161 messages that were in split spool directories. Consequently, it was
1162 deleting retry records that should have stayed in existence.
1164 48. Steve fixed some bugs in eximstats.
1166 49. The SPA authentication driver was not abandoning authentication and moving
1167 on to the next authenticator when an expansion was forced to fail,
1168 contradicting the general specification for all authenticators. Instead it
1169 was generating a temporary error. It now behaves as specified.
1171 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1172 (the order specifies the preference for clients). The order is now AES256,
1173 AES128, 3DES, ARCFOUR128.
1175 51. Small patch to Sieve code - explicitly set From: when generating an
1178 52. Exim crashed if a remote delivery caused a very long error message to be
1179 recorded - for instance if somebody sent an entire SpamAssassin report back
1180 as a large number of 550 error lines. This bug was coincidentally fixed by
1181 increasing the size of one of Exim's internal buffers (big_buffer) that
1182 happened as part of the Exiscan merge. However, to be on the safe side, I
1183 have made the code more robust (and fixed the comments that describe what
1186 53. Now that there can be additional text after "Completed" in log lines (if
1187 the queue_time_overall log selector is set), a one-byte patch to exigrep
1188 was needed to allow it to recognize "Completed" as not the last thing in
1191 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1192 patch that reportedly fixes this has been added. I am not expert enough to
1193 create a test for it. This is what the patch creator wrote:
1195 "I found a little strange behaviour of ldap code when working with
1196 Windows 2003 AD Domain, where users was placed in more than one
1197 Organization Units. When I tried to give exim partial DN, the exit code
1198 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1199 But simultaneously result of request was absolutely normal ldap result,
1200 so I produce this patch..."
1202 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1203 REFERENCE, so I have modified the code to exclude the patch when that macro
1206 55. Some experimental protocols are using DNS PTR records for new purposes. The
1207 keys for these records are domain names, not reversed IP addresses. The
1208 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1209 leaves it alone. Component reversal etc. now happens only for IP addresses.
1212 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1214 57. Double the size of the debug message buffer (to 2048) so that more of very
1215 long debug lines gets shown.
1217 58. The exicyclog utility now does better if the number of log files to keep
1218 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1220 59. Two changes related to the smtp_active_hostname option:
1222 (1) $smtp_active_hostname is now available as a variable.
1223 (2) The default for smtp_banner uses $smtp_active_hostname instead
1224 of $primary_hostname.
1226 60. The host_aton() function is supposed to be passed a string that is known
1227 to be a valid IP address. However, in the case of IPv6 addresses, it was
1228 not checking this. This is a hostage to fortune. Exim now panics and dies
1229 if the condition is not met. A case was found where this could be provoked
1230 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1231 components; fortuitously, this particular loophole had already been fixed
1232 by change 4.50/55 above.
1234 If there are any other similar loopholes, the new check in host_aton()
1235 itself should stop them being exploited. The report I received stated that
1236 data on the command line could provoke the exploit when Exim was running as
1237 exim, but did not say which command line option was involved. All I could
1238 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1239 running as the user.
1242 61. There was a buffer overflow vulnerability in the SPA authentication code
1243 (which came originally from the Samba project). I have added a test to the
1244 spa_base64_to_bits() function which I hope fixes it.
1247 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1248 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1250 63. The daemon start-up calls getloadavg() while still root for those OS that
1251 need the first call to be done as root, but it missed one case: when
1252 deliver_queue_load_max is set with deliver_drop_privilege. This is
1253 necessary for the benefit of the queue runner, because there is no re-exec
1254 when deliver_drop_privilege is set.
1256 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1257 This has been fixed.
1259 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1260 was in use, was not putting the data itself into the right store pool;
1261 consequently, it could be overwritten for a subsequent message in the same
1262 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1265 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1267 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1268 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1269 the function were treating the return as a boolean value, which happened to
1270 work because 0=false and not-0=true, but is not correct code.
1272 68. The host_aton() function was not handling scoped IPv6 addresses (those
1273 with, for example, "%eth0" on the end) correctly.
1275 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1276 negated items (that is, ~something) in unsigned ints. Some compilers
1277 apparently mutter when there is no cast.
1279 70. If an address verification called from an ACL failed, and did not produce a
1280 user-specific message (i.e. there was only a "system" message), nothing was
1281 put in $acl_verify_message. In this situation, it now puts the system
1284 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1285 session; change 4.31/43 added the unwanted input to the log line - except
1286 that it did not do this in the start of session case. It now does.
1288 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1289 This is wrong and can cause the other end to generate a synchronization
1290 error if it is another Exim or anything else that does the synchronization
1291 check. A QUIT command is no longer sent after a timeout.
1293 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1294 during host lookups.
1296 74. The defer_ok option of callout verification was not working if it was used
1297 when verifying addresses in header lines, that is, for this case:
1299 verify = header_sender/callout=defer_ok
1301 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1302 those file descriptors could be used for SMTP connections. If anything
1303 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1304 could be sent to the SMTP client, causing chaos. The daemon now opens
1305 stdin, stdout, and stderr to /dev/null when it puts itself into the
1308 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1309 log by default. The user can override this with suitable Perl magic.
1311 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1312 the log message when discard triggers, was not working for the DATA ACL or
1313 for the non-SMTP ACL.
1315 78. Error message wording change in sieve.c.
1317 79. If smtp_accept_max_per_host was set, the number of connections could be
1318 restricted to fewer than expected, because the daemon was trying to set up
1319 a new connection before checking whether the processes handling previous
1320 connections had finished. The check for completed processes is now done
1321 earlier. On busy systems, this bug wouldn't be noticed because something
1322 else would have woken the daemon, and it would have reaped the completed
1325 80. If a message was submitted locally by a user whose login name contained one
1326 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1327 It caused a spool format error. I have fixed the spool reading code. A
1328 related problem was that the "from" clause in the Received: line became
1329 illegal because of the space(s). It is now covered by ${quote_local_part.
1331 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1334 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1336 83. Patch from Sieve maintainer to fix unterminated string problem in
1337 "vacation" handling.
1339 84. Some minor changes to the Linux configuration files to help with other
1340 OS variants using glibc.
1342 85. One more patch for Sieve to update vacation handling to latest spec.
1345 ----------------------------------------------------
1346 See the note above about the 4.44 and 4.50 releases.
1347 ----------------------------------------------------
1353 1. Change 4.43/35 introduced a bug that caused file counts to be
1354 incorrectly computed when quota_filecount was set in an appendfile
1357 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1358 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1361 3. Give more explanation in the error message when the command for a transport
1362 filter fails to execute.
1364 4. There are several places where Exim runs a non-Exim command in a
1365 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1366 was being done only for the command run by the queryprogram router. It is
1367 now done for all such subprocesses. The other cases are: ${run, transport
1368 filters, and the commands run by the lmtp and pipe transports.
1370 5. Some older OS have a limit of 256 on the maximum number of file
1371 descriptors. Exim was using setrlimit() to set 1000 as a large value
1372 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1373 systems. I've change it so that if it can't get 1000, it tries for 256.
1375 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1376 was an oversight, and furthermore, ever since the addition of extra
1377 controls (e.g. 4.43/32), the checks on when to allow different forms of
1378 "control" were broken. There should now be diagnostics for all cases when a
1379 control that does not make sense is encountered.
1381 7. $recipients is now available in the predata ACL (oversight).
1383 8. Tidy the search cache before the fork to do a delivery from a message
1384 received from the command line. Otherwise the child will trigger a lookup
1385 failure and thereby defer the delivery if it tries to use (for example) a
1386 cached ldap connection that the parent has called unbind on.
1388 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1389 of $address_data from the recipient verification was clobbered by the
1390 sender verification.
1392 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1393 was its contents. (It was OK if the option was not defined at all.)
1395 11. A "Completed" log line is now written for messages that are removed from
1396 the spool by the -Mrm option.
1398 12. $host_address is now set to the target address during the checking of
1399 ignore_target_hosts.
1401 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1402 being passed; this would have caused $sender_host_name to have been used if
1403 matching the list had actually called for a host name (not very likely,
1404 since this list is usually IP addresses). A host name is now passed as
1407 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1408 code that specifies a non-restarting handler (typically sigaction() in
1409 modern systems) in an attempt to fix a rare and obscure crash bug.
1411 15. Narrowed the window for a race in the daemon that could cause it to ignore
1412 SIGCHLD signals. This is not a major problem, because they are used only to
1413 wake it up if nothing else does.
1415 16. A malformed maildirsize file could cause Exim to calculate negative values
1416 for the mailbox size or file count. Odd effects could occur as a result.
1417 The maildirsize information is now recalculated if the size or filecount
1420 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1421 support for a long time. Removed HAVE_SYS_VFS_H.
1423 18. Updated exipick to current release from John Jetmore.
1425 19. Allow an empty sender to be matched against a lookup in an address list.
1426 Previously the only cases considered were a regular expression, or an
1429 20. Exim went into a mad DNS lookup loop when doing a callout where the
1430 host was specified on the transport, if the DNS lookup yielded more than
1433 21. The RFC2047 encoding function was originally intended for short strings
1434 such as real names; it was not keeping to the 75-character limit for
1435 encoded words that the RFC imposes. It now respects the limit, and
1436 generates multiple encoded words if necessary. To be on the safe side, I
1437 have increased the buffer size for the ${rfc2047: expansion operator from
1440 22. Failure to deliver a bounce message always caused it to be frozen, even if
1441 there was an errors_to setting on the router. The errors_to setting is now
1444 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1445 canonical form (fully expanded) before being placed in
1446 $sender_host_address.
1448 24. Updated eximstats to version 1.33
1450 25. Include certificate and key file names in error message when GnuTLS fails
1451 to set them up, because the GnuTLS error message doesn't include the name
1452 of the failing file when there is a problem reading it.
1454 26. Expand error message when OpenSSL has problems setting up cert/key files.
1457 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1458 (this can affect the format of dates).
1460 28. exim_tidydb, when checking for the continued existence of a message for
1461 which it has found a message-specific retry record, was not finding
1462 messages that were in split spool directories. Consequently, it was
1463 deleting retry records that should have stayed in existence.
1465 29. eximstats updated to version 1.35
1466 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1467 1.35 - bugfix such that pie charts by volume are generated correctly
1469 30. The SPA authentication driver was not abandoning authentication and moving
1470 on to the next authenticator when an expansion was forced to fail,
1471 contradicting the general specification for all authenticators. Instead it
1472 was generating a temporary error. It now behaves as specified.
1474 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1475 (the order specifies the preference for clients). The order is now AES256,
1476 AES128, 3DES, ARCFOUR128.
1478 31. Small patch to Sieve code - explicitly set From: when generating an
1481 32. Exim crashed if a remote delivery caused a very long error message to be
1482 recorded - for instance if somebody sent an entire SpamAssassin report back
1483 as a large number of 550 error lines. This bug was coincidentally fixed by
1484 increasing the size of one of Exim's internal buffers (big_buffer) that
1485 happened as part of the Exiscan merge. However, to be on the safe side, I
1486 have made the code more robust (and fixed the comments that describe what
1489 33. Some experimental protocols are using DNS PTR records for new purposes. The
1490 keys for these records are domain names, not reversed IP addresses. The
1491 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1492 leaves it alone. Component reversal etc. now happens only for IP addresses.
1495 34. The host_aton() function is supposed to be passed a string that is known
1496 to be a valid IP address. However, in the case of IPv6 addresses, it was
1497 not checking this. This is a hostage to fortune. Exim now panics and dies
1498 if the condition is not met. A case was found where this could be provoked
1499 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1500 components; fortuitously, this particular loophole had already been fixed
1501 by change 4.50/55 or 4.44/33 above.
1503 If there are any other similar loopholes, the new check in host_aton()
1504 itself should stop them being exploited. The report I received stated that
1505 data on the command line could provoke the exploit when Exim was running as
1506 exim, but did not say which command line option was involved. All I could
1507 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1508 running as the user.
1511 35. There was a buffer overflow vulnerability in the SPA authentication code
1512 (which came originally from the Samba project). I have added a test to the
1513 spa_base64_to_bits() function which I hope fixes it.
1516 36. The daemon start-up calls getloadavg() while still root for those OS that
1517 need the first call to be done as root, but it missed one case: when
1518 deliver_queue_load_max is set with deliver_drop_privilege. This is
1519 necessary for the benefit of the queue runner, because there is no re-exec
1520 when deliver_drop_privilege is set.
1522 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1523 was in use, was not putting the data itself into the right store pool;
1524 consequently, it could be overwritten for a subsequent message in the same
1525 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1528 38. Sometimes the final signoff response after QUIT could fail to get
1529 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1530 < 0 before doing a fflush(). This bug looks as though it goes back to the
1531 introduction of TLS in release 3.20, but "sometimes" must have been rare
1532 because the tests only now provoked it.
1538 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1539 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1540 Now it returns an integer. A number of calls were still expecting a T/F
1541 return. Fortuitously, in all cases, the tests worked in OK situations,
1542 which is the norm. However, things would have gone wrong on any write
1543 failures on the smtp file descriptor. This function is used when sending
1544 messages over SMTP and also when doing verify callouts.
1546 2. When Exim is called to do synchronous delivery of a locally submitted
1547 message (the -odf or -odi options), it no longer closes stderr before doing
1550 3. Implemented the mua_wrapper option.
1552 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1554 5. Implemented the functions header_remove(), header_testname(),
1555 header_add_at_position(), and receive_remove_recipient(), and exported them
1558 6. If an ACL "warn" statement specified the addition of headers, Exim already
1559 inserted X-ACL-Warn: at the start if there was no header name. However, it
1560 was not making this test for the second and subsequent header lines if
1561 there were newlines in the string. This meant that an invalid header could
1562 be inserted if Exim was badly configured.
1564 7. Allow an ACL "warn" statement to add header lines at the start or after all
1565 the Received: headers, as well as at the end.
1567 8. Added the rcpt_4xx retry error code.
1569 9. Added postmaster_mailfrom=xxx to callout verification option.
1571 10. Added mailfrom=xxxx to the callout verification option, for verify=
1574 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1575 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1577 12. Inserted some casts to stop certain compilers warning when using pointer
1578 differences as field lengths or precisions in printf-type calls (mostly
1579 affecting debugging statements).
1581 13. Added optional readline() support for -be (dynamically loaded).
1583 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1584 same clock tick as a message's arrival, so that its received time was the
1585 same as the "first fail" time on the retry record, and that message
1586 remained on the queue past the ultimate address timeout, every queue runner
1587 would try a delivery (because it was past the ultimate address timeout) but
1588 after another failure, the ultimate address timeout, which should have then
1589 bounced the address, did not kick in. This was a "< instead of <=" error;
1590 in most cases the first failure would have been in the next clock tick
1591 after the received time, and all would be well.
1593 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1594 being recognized when the domain list was tested by the match_domain
1595 condition in an expansion string.
1597 16. Added the ${str2b64: operator.
1599 17. Exim was always calling setrlimit() to set a large limit for the number of
1600 processes, without checking whether the existing limit was already
1601 adequate. (It did check for the limit on file descriptors.) Furthermore,
1602 errors from getrlimit() and setrlimit() were being ignored. Now they are
1603 logged to the main and panic logs, but Exim does carry on, to try to do its
1604 job under whatever limits there are.
1606 18. Imported PCRE 5.0.
1608 19. Trivial typo in log message " temporarily refused connection" (the leading
1611 20. If the log selector return_path_on_delivery was set and an address was
1612 redirected to /dev/null, the delivery process crashed because it assumed
1613 that a return path would always be set for a "successful" delivery. In this
1614 case, the whole delivery is bypassed as an optimization, and therefore no
1617 21. Internal re-arrangement: the function for sending a challenge and reading
1618 a response while authentication was assuming a zero-terminated challenge
1619 string. It's now changed to take a pointer and a length, to allow for
1620 binary data in such strings.
1622 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1624 23. Exim was not respecting finduser_retries when seeking the login of the
1625 uid under which it was called; it was always trying 10 times. (The default
1626 setting of finduser_retries is zero.) Also, it was sleeping after the final
1627 failure, which is pointless.
1629 24. Implemented tls_on_connect_ports.
1631 25. Implemented acl_smtp_predata.
1633 26. If the domain in control=submission is set empty, Exim assumes that the
1634 authenticated id is a complete email address when it generates From: or
1635 Sender: header lines.
1637 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1638 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1639 chown and chgrp in /bin and hostname in /usr/bin.
1641 28. Exim was keeping the "process log" file open after each use, just as it
1642 does for the main log. This opens the possibility of it remaining open for
1643 long periods when the USR1 signal hits a daemon. Occasional processlog
1644 errors were reported, that could have been caused by this. Anyway, it seems
1645 much more sensible not to leave this file open at all, so that is what now
1648 29. The long-running daemon process does not normally write to the log once it
1649 has entered its main loop, and it closes the log before doing so. This is
1650 so that log files can straightforwardly be renamed and moved. However,
1651 there are a couple of unusual error situations where the daemon does write
1652 log entries, and I had neglected to close the log afterwards.
1654 30. The text of an SMTP error response that was received during a remote
1655 delivery was being truncated at 512 bytes. This is too short for some of
1656 the long messages that one sometimes sees. I've increased the limit to
1659 31. It is now possible to make retry rules that apply only when a message has a
1660 specific sender, in particular, an empty sender.
1662 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1663 it possible to be selective about when SMTP synchronization is enforced.
1665 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1667 32. Implemented hosts_connection_nolog.
1669 33. Added an ACL for QUIT.
1671 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1674 35. Added mailbox_size and mailbox_filecount to appendfile.
1676 36. Added control = no_multiline_responses to ACLs.
1678 37. There was a bug in the logic of the code that waits for the clock to tick
1679 in the case where the clock went backwards by a substantial amount such
1680 that the microsecond fraction of "now" was more than the microsecond
1681 fraction of "then" (but the whole seconds number was less).
1683 38. Added support for the libradius Radius client library this is found on
1684 FreeBSD (previously only the radiusclient library was supported).
1690 1. When certain lookups returned multiple values in the form name=value, the
1691 quoting of the values was not always being done properly. Specifically:
1692 (a) If the value started with a double quote, but contained no whitespace,
1694 (b) If the value contained whitespace other than a space character (i.e.
1695 tabs or newlines or carriage returns) it was not quoted.
1696 This fix has been applied to the mysql and pgsql lookups by writing a
1697 separate quoting function and calling it from the lookup code. The fix
1698 should probably also be applied to nisplus, ibase and oracle lookups, but
1699 since I cannot test any of those, I have not disturbed their existing code.
1701 2. A hit in the callout cache for a specific address caused a log line with no
1702 reason for rejecting RCPT. Now it says "Previous (cached) callout
1703 verification failure".
1705 3. There was an off-by-one bug in the queryprogram router. An over-long
1706 return line was truncated at 256 instead of 255 characters, thereby
1707 overflowing its buffer with the terminating zero. As well as fixing this, I
1708 have increased the buffer size to 1024 (and made a note to document this).
1710 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1711 when Exim is waiting for an SMTP response from a remote server, Exim
1712 restarts its select() call on the socket, thereby resetting its timeout.
1713 This is not a problem when such interrupts are rare. Somebody set up a cron
1714 job to run exiwhat every 2 minutes, which is less than the normal select()
1715 timeout (5 or 10 minutes). This meant that the select() timeout never
1716 kicked in because it was always reset. I have fixed this by comparing the
1717 time when an interrupt arrives with the time at the start of the first call
1718 to select(). If more time than the timeout has elapsed, the interrupt is
1719 treated as a timeout.
1721 5. Some internal re-factoring in preparation for the addition of Sieve
1722 extensions (by MH). In particular, the "personal" test is moved to a
1723 separate function, and given an option for scanning Cc: and Bcc: (which is
1724 not set for Exim filters).
1726 6. When Exim created an email address using the login of the caller as the
1727 local part (e.g. when creating a From: or Sender: header line), it was not
1728 quoting the local part when it contained special characters such as @.
1730 7. Installed new OpenBSD configuration files.
1732 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1733 try to make them clearer.
1735 9. Callout options, other than the timeout value, were being ignored when
1736 verifying sender addresses in header lines. For example, when using
1738 verify = header_sender/callout=no_cache
1740 the cache was (incorrectly) being used.
1742 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1743 only the Cygwin environment.
1745 11. When return_path_on_delivery was set as a log selector, if different remote
1746 addresses in the same message used different return paths and parallel
1747 remote delivery occurred, the wrong values would sometimes be logged.
1748 (Whenever a remote delivery process finished, the return path value from
1749 the most recently started remote delivery process was logged.)
1751 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1752 lines when AUTH and/or TLS are in use. This is the "received protocol"
1753 field. Exim used to use "asmtp" for authenticated SMTP, without any
1754 indication (in the protocol name) for TLS use. Now it follows the RFC and
1755 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1756 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1757 names appear in log lines as well as in Received: header lines.
1759 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1760 extensions, and comparison tests, and to fix some bugs.
1762 14. Changes to the "personal" filter test:
1764 (1) The test was buggy in that it was just doing the equivalent of
1765 "contains" tests on header lines. For example, if a user's address was
1766 anne@some.where, the "personal" test would incorrectly be true for
1768 To: susanne@some.where
1770 This test is now done by extracting each address from the header in turn,
1771 and checking the entire address. Other tests that are part of "personal"
1772 are now done using regular expressions (for example, to check local parts
1773 of addresses in From: header lines).
1775 (2) The list of non-personal local parts in From: addresses has been
1776 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1777 taken from the Sieve specification recommendations.
1779 (3) If the message contains any header line starting with "List-" it is
1780 treated as non-personal.
1782 (4) The test for "circular" in the Subject: header line has been removed
1783 because it now seems ill-conceived.
1785 15. Minor typos in src/EDITME comments corrected.
1787 16. Installed latest exipick from John Jetmore.
1789 17. If headers_add on a router specified a text string that was too long for
1790 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1791 of string_sprintf() is now avoided.
1793 18. $message_body_size was not set (it was always zero) when running the DATA
1794 ACL and the local_scan() function.
1796 19. For the "mail" command in an Exim filter, no default was being set for
1797 the once_repeat time, causing a random time value to be used if "once" was
1798 specified. (If the value happened to be <= 0, no repeat happened.) The
1799 default is now 0s, meaning "never repeat". The "vacation" command was OK
1800 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1801 (I found it when inspecting the code).
1803 20. There is now an overall timeout for performing a callout verification. It
1804 defaults to 4 times the callout timeout, which applies to individual SMTP
1805 commands during the callout. The overall timeout applies when there is more
1806 than one host that can be tried. The timeout is checked before trying the
1807 next host. This prevents very long delays if there are a large number of
1808 hosts and all are timing out (e.g. when the network connections are timing
1809 out). The value of the overall timeout can be changed by specifying an
1810 additional sub-option for "callout", called "maxwait". For example:
1812 verify = sender/callout=5s,maxwait=20s
1814 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1815 to the end before writing, but this should make it even safer).
1817 22. Exim was forgetting that it had advertised PIPELINING for the second and
1818 subsequent messages on an SMTP connection. It was also not resetting its
1819 memory on STARTTLS and an internal HELO.
1821 23. When Exim logs an SMTP synchronization error within a session, it now
1822 records whether PIPELINING has been advertised or not.
1824 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1825 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1826 rather than long int.
1828 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1830 26. Added the never_mail option to autoreply.
1836 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1837 crash if the getsockname() call failed; this can happen if a connection is
1838 closed very soon after it is established. The problem was simply in the
1839 order in which certain operations were done, causing Exim to try to write
1840 to the SMTP stream before it had set up the file descriptor. The bug has
1841 been fixed by making things happen in the correct order.
1847 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1848 before the connection was closed, thus losing the rejection response.
1850 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1851 some early Solaris releases, but causes trouble in current releases where
1852 socklen_t is defined.
1854 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1857 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1860 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1861 that releasing the top of it at the end releases what was used for sub-
1862 expansions (unless the block got too big). However, discard this block if
1863 the first thing is a variable or header, so that we can use its block when
1864 it is dynamic (useful for very large $message_headers, for example).
1866 6. Lookups now cache *every* query, not just the most recent. A new, separate
1867 store pool is used for this. It can be recovered when all lookup caches are
1868 flushed. Lookups now release memory at the end of their result strings.
1869 This has involved some general refactoring of the lookup sources.
1871 7. Some code has been added to the store_xxx() functions to reduce the amount
1872 of flapping under certain conditions.
1874 8. log_incoming_interface used to affect only the <= reception log lines. Now
1875 it causes the local interface and port to be added to several more SMTP log
1876 lines, for example "SMTP connection from", and rejection lines.
1878 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1880 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1882 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1883 could be overwritten at the end of the current message (or the start of a
1884 new message if it was set in a HELO ACL). The value is now preserved for
1885 the duration of the SMTP connection.
1887 12. If a transport had a headers_rewrite setting, and a matching header line
1888 contained an unqualified address, that address was qualified, even if it
1889 did not match any rewriting rules. The underlying bug was that the values
1890 of the flags that permit the existence of unqualified sender and recipient
1891 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1892 non-local messages, and by -bnq for local messages) were not being
1893 preserved with the message after it was received.
1895 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1896 "next input=" as part of the text comprising the host identity instead of
1897 the correct text. The code was using the same buffer for two different
1898 strings. However, depending on which order the printing function evaluated
1899 its arguments, the bug did not always show up. Under Linux, for example, my
1900 test suite worked just fine.
1902 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1903 doesn't work with some older versions of Perl. It has been changed to "my",
1904 which in any case is probably the better facility to use.
1906 15. A really picky compiler found some instances of statements for creating
1907 error messages that either had too many or two few arguments for the format
1910 16. The size of the buffer for calls to the DNS resolver has been increased
1911 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1912 for addresses that have a lot of PTR records. This alleviates a problem; it
1913 does not fully solve it.
1915 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1916 buffer now truncates the list and logs the incident, which is the same
1917 action as happens when Exim is looking up a host name and its aliases.
1918 Previously in this situation something unpredictable would happen;
1919 sometimes it was "internal error: store_reset failed".
1921 18. If a server dropped the connection unexpectedly when an Exim client was
1922 using GnuTLS and trying to read a response, the client delivery process
1923 crashed while trying to generate an error log message.
1925 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1926 string, for example:
1928 warn message = H1: something\nH2: something
1930 the text was added as a single header line from Exim's point of view
1931 though it ended up OK in the delivered message. However, searching for the
1932 second and subsequent header lines using $h_h2: did not work. This has been
1933 fixed. Similarly, if a system filter added multiple headers in this way,
1934 the routers could not see them.
1936 20. Expanded the error message when iplsearch is called with an invalid key to
1937 suggest using net-iplsearch in a host list.
1939 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1940 ACLs are no longer actually imposed (and a message to that effect is
1943 22. If a "gecos" field in a passwd entry contained escaped characters, in
1944 particular, if it contained a \" sequence, Exim got it wrong when building
1945 a From: or a Sender: header from that name. A second bug also caused
1946 incorrect handling when an unquoted " was present following a character
1947 that needed quoting.
1949 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1950 was not being matched caselessly.
1952 24. Arranged for all hyphens in the exim.8 source to be escaped with
1955 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1956 properly. Recipient callout cache records were still being keyed to include
1957 the sender, even when use_sender was set false. This led to far more
1958 callouts that were necessary. The sender is no longer included in the key
1959 when use_sender is false.
1961 26. Added "control = submission" modifier to ACLs.
1963 27. Added the ${base62d: operator to decode base 62 numbers.
1965 28. dnsdb lookups can now access SRV records.
1967 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1968 the configuration file.
1970 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1971 (-v) mode. This makes the output for a verbose queue run more intelligible.
1973 31. Added a use_postmaster feature to recipient callouts.
1975 32. Added the $body_zerocount variable, containing the number of binary zero
1976 bytes in the message body.
1978 33. The time of last modification of the "new" subdirectory is now used as the
1979 "mailbox time last read" when there is a quota error for a maildir
1982 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1984 35. Added +ignore_unknown as a special item in host lists.
1986 36. Code for decoding IPv6 addresses in host lists is now included, even if
1987 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1988 address was recognized as an IP address, but was then not correctly decoded
1989 into binary, causing unexpected and incorrect effects when compared with
1996 1. Very minor rewording of debugging text in manualroute to say "list of
1997 hosts" instead of "hostlist".
1999 2. If verify=header_syntax was set, and a header line with an unqualified
2000 address (no domain) and a large number of spaces between the end of the
2001 name and the colon was received, the reception process suffered a buffer
2002 overflow, and (when I tested it) crashed. This was caused by some obsolete
2003 code that should have been removed. The fix is to remove it!
2005 3. When running in the test harness, delay a bit after writing a bounce
2006 message to get a bit more predictability in the log output.
2008 4. Added a call to search_tidyup() just before forking a reception process. In
2009 theory, someone could use a lookup in the expansion of smtp_accept_max_
2010 per_host which, without the tidyup, could leave open a database connection.
2012 5. Added the variables $recipient_data and $sender_data which get set from a
2013 lookup success in an ACL "recipients" or "senders" condition, or a router
2014 "senders" option, similar to $domain_data and $local_part_data.
2016 6. Moved the writing of debug_print from before to after the "senders" test
2019 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2020 problems for message scanning, either using a data ACL, or using
2021 local_scan() because the Received: header was not generated till after they
2022 were called (in order to set the time as the time of reception completion).
2023 I have revised the way this works. The header is now generated after the
2024 body is received, but before the ACL or local_scan() are called. After they
2025 are run, the timestamp in the header is updated.
2031 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2032 before starting a queue runner without re-exec. This happened only when
2033 deliver_drop_privilege was set or when the Exim user was set to root. The
2034 effect of the bug was that timeouts during subsequent deliveries caused
2035 crashes instead of being properly handled. The handler is now left at its
2036 default (and expected) setting.
2038 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2039 message, again when deliver_drop_privilege is set or Exim is run as root.
2040 The bug described in (1) was not present in this case, but the tidying up
2041 of the other signals was missing. I have made the two cases consistent.
2043 3. The ignore_target_hosts setting on a manualroute router was being ignored
2044 for hosts that were looked up using the /MX notation.
2046 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2049 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2050 operated on the sender address. After changing the $sender_address to <>
2051 for the sender address verify, Exim was re-instated it as the original
2052 (before rewriting) address, but remembering that it had rewritten it, so it
2053 wasn't rewriting it again. This bug also had the effect of breaking the
2054 sender address verification caching when the sender address was rewritten.
2056 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2057 This has been changed so that if the ip literal address matches
2058 ignore_target_hosts, the router declines.
2060 7. Added expansion conditions match_domain, match_address, and match_local_
2061 part (NOT match_host).
2063 8. The placeholder for the Received: header didn't have a length field set.
2065 9. Added code to Exim itself and to exim_lock to test for a specific race
2066 condition that could lead to file corruption when using MBX delivery. The
2067 issue is with the lockfile that is created in /tmp. If this file is removed
2068 after a process has opened it but before that process has acquired a lock,
2069 there is the potential for a second process to recreate the file and also
2070 acquire a lock. This could lead to two Exim processes writing to the file
2071 at the same time. The added code performs the same test as UW imapd; it
2072 checks after acquiring the lock that its file descriptor still refers to
2073 the same named file.
2075 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2076 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2079 11. Added the smtp_active_hostname option. If used, this will typically be made
2080 to depend on the incoming interface address. Because $interface_address is
2081 not set up until the daemon has forked a reception process, error responses
2082 that can happen earlier (such as "too many connections") no longer contain
2085 12. If an expansion in a condition on a "warn" statement fails because a lookup
2086 defers, the "warn" statement is abandoned, and the next ACL statement is
2087 processed. Previously this caused the whole ACL to be aborted.
2089 13. Added the iplsearch lookup type.
2091 14. Added ident_timeout as a log selector.
2093 15. Added tls_certificate_verified as a log selector.
2095 16. Added a global option tls_require_ciphers (compare the smtp transport
2096 option of the same name). This controls incoming TLS connections.
2098 17. I finally figured out how to make tls_require_ciphers do a similar thing
2099 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2100 before starting the TLS session.
2102 18. Tabs are now shown as \t in -bP output.
2104 19. If the log selector return_path_on_delivery was set, Exim crashed when
2105 bouncing a message because it had too many Received: header lines.
2107 20. If two routers both had headers_remove settings, and the first one included
2108 a superfluous trailing colon, the final name in the first list and the
2109 first name in the second list were incorrectly joined into one item (with a
2110 colon in the middle).
2116 1. Added -C and -D options to the exinext utility, mainly to make it easier
2117 to include in the automated testing, but these could be helpful when
2118 multiple configurations are in use.
2120 2. The exinext utility was not formatting the output nicely when there was
2121 an alternate port involved in the retry record key, nor when there was a
2122 message id as well (for retries that were specific to a specific message
2123 and a specific host). It was also confused by IPv6 addresses, because of
2124 the additional colons they contain. I have fixed the IPv4 problem, and
2125 patched it up to do a reasonable job for IPv6.
2127 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2128 delivery, the log line now contains "pipelined" if PIPELINING was used.
2130 4. An SMTP transport process used to panic and die if the bind() call to set
2131 an explicit outgoing interface failed. This has been changed; it is now
2132 treated in the same way as a connect() failure.
2134 5. A reference to $sender_host_name in the part of a conditional expansion
2135 that was being skipped was still causing a DNS lookup. This no longer
2138 6. The def: expansion condition was not recognizing references to header lines
2139 that used bh_ and bheader_.
2141 7. Added the _cache feature to named lists.
2143 8. The code for checking quota_filecount in the appendfile transport was
2144 allowing one more file than it should have been.
2146 9. For compatibility with Sendmail, the command line option
2154 and sets the incoming protocol and host name (for trusted callers). The
2155 host name and its colon can be omitted when only the protocol is to be set.
2156 Note the Exim already has two private options, -pd and -ps, that refer to
2157 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2158 "s", but I don't think that's a major issue.
2160 10. A number of refactoring changes to the code, none of which should affect
2163 (a) The number of logging options was getting close to filling up the
2164 32-bit word that was used as a bit map. I have split them into two classes:
2165 those that are passed in the argument to log_write(), and those that are
2166 only ever tested independently outside of that function. These are now in
2167 separate 32-bit words, so there is plenty of room for expansion again.
2168 There is no change in the user interface or the logging behaviour.
2170 (b) When building, for example, log lines, the code previously used a
2171 macro that called string_cat() twice, in order to add two strings. This is
2172 not really sufficiently general. Furthermore, there was one instance where
2173 it was actually wrong because one of the argument was used twice, and in
2174 one call a function was used. (As it happened, calling the function twice
2175 did not affect the overall behaviour.) The macro has been replaced by a
2176 function that can join an arbitrary number of extra strings onto a growing
2179 (c) The code for expansion conditions now uses a table and a binary chop
2180 instead of a serial search (which was left over from when there were very
2181 few conditions). Also, it now recognizes conditions like "pam" even when
2182 the relevant support is not compiled in: a suitably worded error message is
2183 given if an attempt is made to use such a condition.
2185 11. Added ${time_interval:xxxxx}.
2187 12. A bug was causing one of the ddress fields not to be passed back correctly
2188 from remote delivery subprocesses. The field in question was not being
2189 subsequently used, so this caused to problems in practice.
2191 13. Added new log selectors queue_time and deliver_time.
2193 14. Might have fixed a bug in maildirsizefile handling that threw up
2194 "unexpected character" debug warnings, and recalculated the data
2195 unnecessarily. In any case, I expanded the warning message to give more
2198 15. Added the message "Restricted characters in address" to the statements in
2199 the default ACL that block characters like @ and % in local parts.
2201 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2202 Three changes have been made:
2204 (a) There was a serious bug; a negative response to MAIL caused the whole
2205 recipient domain to be cached as invalid, thereby blocking all messages
2206 to all local parts at the same domain, from all senders. This bug has
2207 been fixed. The domain is no longer cached after a negative response to
2208 MAIL if the sender used is not empty.
2210 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2213 (c) A new callout option, "use_sender" has been added for people who want
2214 the modified behaviour.
2220 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2223 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2224 indeed breaks things for older releases.
2226 3. Added additional logging to the case where there is a problem reading data
2227 from a filter that is running in a subprocess using a pipe, in order to
2228 try to track down a specific problem.
2230 4. Testing facility fudge: when running in the test harness and attempting
2231 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2232 getting "No route to host". Convert this to a timeout.
2234 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2237 6. Some OS don't have socklen_t but use size_t instead. This affects the
2238 fifth argument of getsockopt() amongst other things. This is now
2239 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2240 can be set for individual OS. I have set it for SunOS5, OSF1, and
2241 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2242 some earlier ones do not.
2244 7. Change 4.30/15 was not doing the test caselessly.
2246 8. The standard form for an IPv6 address literal was being rejected by address
2247 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2248 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2249 this, as well as the form without the "IPv6" on the front (but only when
2250 address literals are enabled, of course).
2252 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2254 10. Exim crashed if a message with an empty sender address specified by -f
2255 encountered a router with an errors_to setting. This could be provoked only
2256 by a command such as
2260 where an empty string was supplied; "<>" did not hit this bug.
2262 11. Installed PCRE release 4.5.
2264 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2265 remained set. It is now erased.
2267 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2268 times from message ids (which are base 36 rather than the normal 62).
2270 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2271 were being counted as actual protocol errors, and logged if the log
2272 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2273 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2274 and DATA following a set of rejected RCPTs do not count as protocol errors.
2275 In other words, Exim assumes they were pipelined, though this may not
2276 actually be the case. Of course, in all cases the client gets an
2277 appropriate error code.
2279 15. If a lookup fails in an ACL condition, a message about the failure may
2280 be available; it is used if testing the ACL cannot continue, because most
2281 such messages specify what the cause of the deferral is. However, some
2282 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2283 that caused an old message to be retained and used if a later statement
2284 caused a defer, replacing the real cause of the deferral.
2286 16. If an IP address had so many PTR records that the DNS lookup buffer
2287 was not large enough to hold them, Exim could crash while trying to process
2288 the truncated data. It now detects and logs this case.
2290 17. Further to 4.21/58, another change has been made: if (and only if) the
2291 first line of a message (the first header line) ends with CRLF, a bare LF
2292 in a subsequent header line has a space inserted after it, so as not to
2293 terminate the header.
2295 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2296 unnecessarily, used atoi() instead of strtol(), and didn't check the
2297 termination when getting file sizes from file names by regex.
2299 19. Completely re-implemented the support for maildirsize files, in the light
2300 of a number of problems with the previous contributed implementation
2301 (4.30/29). In particular:
2303 . If the quota is zero, the maildirsize file is maintained, but no quota is
2306 . If the maildir directory does not exist, it is created before any attempt
2307 to write a maildirsize file.
2309 . The quota value in the file is just a cache; if the quota is changed in
2310 the transport, the new value overrides.
2312 . A regular expression is available for excluding directories from the
2315 20. The autoreply transport checks the characters in options that define the
2316 message's headers; it allows continued headers, but it was checking with
2317 isspace() after an embedded newline instead of explicitly looking for a
2320 21. If all the "regular" hosts to which an address was routed had passed their
2321 expiry times, and had not reached their retry times, the address was
2322 bounced, even if fallback hosts were defined. Now Exim should go on to try
2325 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2326 equivalent code in the SMTP transport. Some hosts send humungous responses
2327 to HELO/EHLO, more than 1024 it seems.
2329 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2330 gives compiler warnings in some environments. I've now done it "properly",
2333 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2334 (because of problems with the built-in one) was declared to return uschar *
2335 instead of char *, causing compiler failure.
2337 25. Fixed a file descriptor leak when processing alias/forward files.
2339 26. Fixed a minor format string issue in dbfn.c.
2341 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2343 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2344 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2347 29. When a sender address is verified, it is cached, to save repeating the test
2348 when there is more than one recipient in a message. However, when the
2349 verification involves a callout, it is possible for different callout
2350 options to be set for different recipients. It is too complicated to keep
2351 track of this in the cache, so now Exim always runs a verification when a
2352 callout is required, relying on the callout cache for the optimization.
2353 The overhead is duplication of the address routing, but this should not be
2356 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2357 to be verified with callout=postmaster, and the main callout worked but the
2358 postmaster check failed, the verification correctly failed. However, if a
2359 subsequent RCPT command asked for sender verification *without* the
2360 postmaster check, incorrect caching caused this verification also to fail,
2363 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2364 it was not caching the DNS options (qualify_single, search_parents) that
2365 were used when the lookup failed. A subsequent lookup with different
2366 options therefore always gave the same answer, though there were cases
2367 where it should not have. (Example: a "domains = !$mx_any" option on a
2368 dnslookup router: the "domains" option is always processed without any
2369 widening, but the router might have qualify_single set.) Now Exim uses the
2370 cached value only when the same options are set.
2372 32. Added John Jetmore's "exipick" utility to the distribution.
2374 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2375 than a timeout (e.g. a certificate is required, and is not provided), an
2376 Exim server now closes the connection immediately. Previously it waited for
2377 the client to close - but if the client is SSL, it seems that they each
2378 wait for each other, leading to a delay before one of them times out.
2380 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2381 maintained 0.8.x compatibility because I don't think many are using it, and
2382 it is clearly obsolete.
2384 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2387 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2388 client certificate was expired. A simple patch fixes this, though I don't
2389 understand the full logic of why the verify callback is called multiple
2392 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2393 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2394 which causes problems with some clients (such as the Certicom SSL Plus
2395 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2396 disables the coutermeasure allowing Eudora to connect."
2398 38. Exim was not checking that a write() to a log file succeeded. This could
2399 lead to Bad Things if a log got too big, in particular if it hit a file
2400 size limit. Exim now panics and dies if it cannot write to a log file, just
2401 as it does if it cannot open a log file.
2403 39. Modified OS/Makefile-Linux so that it now contains
2405 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2407 The two -D definitions ensure that Exim is compiled with large file
2408 support, which makes it possible to handle log files that are bigger than
2411 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2412 instance) a domain was checked against a named list that involved a lookup,
2413 causing $domain_data to be set, then another domain was checked against the
2414 same list, then the first domain was re-checked, the value of $domain_data
2415 after the final check could be wrong. In particular, if the second check
2416 failed, it could be set empty. This bug probably also applied to
2419 41. The strip_trailing_dot option was not being applied to the address given
2420 with the -f command-line option.
2422 42. The code for reading a message's header from the spool was incrementing
2423 $received_count, but never initializing it. This meant that the value was
2424 incorrect (doubled) while delivering a message in the same process in which
2425 it was received. In the most common configuration of Exim, this never
2426 happens - a fresh exec is done - but it can happen when
2427 deliver_drop_privilege is set.
2429 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2430 it now includes up to 150 characters of the unexpected data in the log
2433 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2434 and building data strings. The size of both of these buffers was 10 000
2435 bytes - far larger than anybody would *ever* want, thought I. Needless to
2436 say, somebody hit the limit. I have increased the maximum line length to
2437 20 000 and the maximum data length of concatenated lines to 100 000. I have
2438 also fixed two bugs, because there was no checking on these buffers. Tsk,
2439 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2440 buffer is too small.
2442 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2443 lsearch lookups. Now it does.
2445 46. When parsing a route_list item in a manualroute router, a fixed-length
2446 buffer was used for the list of hosts. I made this 1024 bytes long,
2447 thinking that nobody would ever have a list of hosts that long. Wrong.
2448 Somebody had a whole pile of complicated expansion conditions, and the
2449 string was silently truncated, leading to an expansion error. It turns out
2450 that it is easier to change to an unlimited length (owing to other changes
2451 that have happened since this code was originally written) than to build
2452 structure for giving a limitation error. The length of the item that
2453 expands into the list of hosts is now unlimited.
2455 47. The lsearch lookup could not handle data where the length of text line was
2456 more than 4095 characters. Such lines were truncated, leading to shortened
2457 data being returned. It should now handle lines of any length.
2459 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2460 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2463 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2464 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2465 no longer happens in comment lines. A list of replacements is now placed
2466 at the head of all of the source files, except those whose only change is
2467 to replace PERL_COMMAND in the very first #! line.
2469 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2470 messages on the queue, with a bottom-up merge sort, using code contributed
2471 by Michael Haardt. This should make operations like -bp somewhat faster on
2472 large queues. It won't affect queue runners, except when queue_run_in_order
2475 51. Installed eximstats 1.31 in the distribution.
2477 52. Added support for SRV lookups to the dnslookup router.
2479 53. If an ACL referred to $message_body or $message_body_end, the value was not
2480 reset for any messages that followed in the same SMTP session.
2482 54. The store-handling optimization for building very long strings was not
2483 differentiating between the different store pools. I don't think this
2484 actually made any difference in practice, but I've tidied it.
2486 55. While running the routers to verify a sender address, $sender_address
2487 was still set to the sender address. This is wrong, because when routing to
2488 send a bounce to the sender, it would be empty. Therefore, I have changed
2489 it so that, while verifying a sender address, $sender_address is set to <>.
2490 (There is no change to what happens when verifying a recipient address.)
2492 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2493 target A or AAAA records (if not already returned) without resetting the
2494 qualify_single or search_parents options of the DNS resolver. These are
2495 inappropriate in this case because the targets of MX and SRV records must
2496 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2497 target that, when qualified, matched something in the local domain. These
2498 two options are now turned off when doing these lookups.
2500 57. It seems that at least some releases of Reiserfs (which does not have the
2501 concept of a fixed number of inodes) returns zero and not -1 for the
2502 number of available inodes. This interacted badly with check_spool_inodes,
2503 which assumed that -1 was the "no such thing" setting. What I have done is
2504 to check that the total number of inodes is greater than zero before doing
2505 the test of how many are available.
2507 58. When a "warn" ACL statement has a log_message modifier, the message is
2508 remembered, and not repeated. This is to avoid a lot of repetition when a
2509 message has many recipients that cause the same warning to be written.
2510 Howewer, Exim was preserving the list of already written lines for an
2511 entire SMTP session, which doesn't seem right. The memory is now reset if a
2512 new message is started.
2514 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2515 debugging output unless log_rewrite was also set.
2517 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2518 of (int)(handle) when we know that handle contains (void *)(-1).
2520 61. The Exim daemon panic-logs an error return when it closes the incoming
2521 connection. However "connection reset by peer" seems to be common, and
2522 isn't really an error worthy of noting specially, so that particular error
2525 62. When Exim is trying to find all the local interfaces, it used to panic and
2526 die if the ioctl to get the interface flags failed. However, it seems that
2527 on at least one OS (Solaris 9) it is possible to have an interface that is
2528 included in the list of interfaces, but for which you get a failure error
2529 for this call. This happens when the interface is not "plumbed" into a
2530 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2531 failure of the "get flags" call assumes that the interface is down.
2533 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2534 makes life easier for people who are doing arithmetic on fields extracted
2535 from dates, where you often get leading zeros that should not be
2536 interpreted as octal.
2538 64. Added qualify_domain to the redirect router, to override the global
2541 65. If a pathologically long header line contained very many addresses (the
2542 report of this problem mentioned 10 000) and each of them was rewritten,
2543 Exim could use up a very large amount of memory. (It kept on making new
2544 copies of the header line as it rewrote, and never released the old ones.)
2545 At the expense of a bit more processing, the header rewriting function has
2546 been changed so that it no longer eats memory in this way.
2548 66. The generation of the Received: header has been moved from the time that a
2549 message starts to be received, to the time that it finishes. The timestamp
2550 in the Received: header should now be very close to that of the <= log
2551 line. There are two side-effects of this change:
2553 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2554 logged header lines no longer include the local Received: line, because
2555 it has not yet been created. The same applies to a copy of the message
2556 that is returned to a non-SMTP sender when a message is rejected.
2558 (b) When a filter file is tested using -bf, no additional Received: header
2559 is added to the test message. After some thought, I decided that this
2562 This change does not affect the value of $received_for. It is still set
2563 after address rewriting, but before local_scan() is called.
2565 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2567 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2568 gave an unhelpful panic error message, and a defer error. I have managed to
2569 change this behaviour so that it now rejects any supplied certificate,
2570 which seems right, as the list of acceptable certificates is empty.
2572 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2573 gave an unhelpful defer error. I have not managed to make this reject any
2574 supplied certificates, but the error message it gives is "no certificate
2575 supplied", which is not helpful.
2577 70. exigrep's output now also includes lines that are not associated with any
2578 message, but which match the given pattern. Implemented by a patch from
2579 Martin Sluka, which also tidied up the Perl a bit.
2581 71. Recipient callout verification, like sender verification, was using <> in
2582 the MAIL FROM command. This isn't really the right thing, since the actual
2583 sender may affect whether the remote host accepts the recipient or not. I
2584 have changed it to use the actual sender in the callout; this means that
2585 the cache record is now keyed on a recipient/sender pair, not just the
2586 recipient address. There doesn't seem to be a real danger of callout loops,
2587 since a callout by the remote host to check the sender would use <>.
2588 [SEE ABOVE: changed after hitting problems.]
2590 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2591 temporary errors. However, in the case of such a code being given after
2592 the end of a data transmission (i.e. after ".") Exim was failing to write
2593 a retry record for the message. (Yes, there was some broken host that was
2594 actually sending 8xx at this point.)
2596 73. An unknown lookup type in a host list could cause Exim to panic-die when
2597 the list was checked. (An example that provoked this was putting <; in the
2598 middle of a list instead of at the start.) If this happened during a DATA
2599 ACL check, a -D file could be left lying around. This kind of configuration
2600 error no longer causes Exim to die; instead it causes a defer errror. The
2601 incident is still logged to the main and panic logs.
2603 74. Buglet left over from Exim 3 conversion. The message "too many messages
2604 in one connection" was written to the rejectlog but not the mainlog, except
2605 when address rewriting (yes!) was being logged.
2607 75. Added write_rejectlog option.
2609 76. When a system filter was run not as root (that is, when system_filter_user
2610 was set), the values of the $n variables were not being returned to the
2611 main process; thus, they were not subsequently available in the $sn
2614 77. Added +return_path_on_delivery log selector.
2616 78. A connection timeout was being treated differently from recipients deferred
2617 when testing hosts_max_try with a message that was older than the host's
2618 retry timeout. (The host should not be counted, thus allowing all hosts to
2619 be tried at least once before bouncing.) This may have been the cause of an
2620 occasionally reported bug whereby a message would remain on the queue
2621 longer than the retry timeout, but would be bounced if a delivery was
2622 forced. I say "may" because I never totally pinned down the problem;
2623 setting up timeout/retry tests is difficult. See also the next item.
2625 79. The ultimate address timeout was not being applied to errors that involved
2626 a combination of host plus message (for example, a timeout on a MAIL
2627 command). When an address resolved to a number of possible hosts, and they
2628 were not all tried for each delivery (e.g. because of hosts_max_try), a
2629 message could remain on the queue longer than the retry timeout.
2631 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2634 81. Fixed an obscure SMTP outgoing bug which required at least the following
2635 conditions: (a) there was another message waiting for the same server;
2636 (b) the server returned 5xx to all RCPT commands in the first message so
2637 that the message was not completed; (c) the server dropped the connection
2638 or gave a negative response to the RSET that Exim sends to abort the
2639 transaction. The observed case was a dropped connection after DATA that had
2640 been sent in pipelining mode. That is, the server had advertised PIPELINING
2641 but was not implementing it correctly. The effect of the bug was incorrect
2642 behaviour, such as trying another host, and this could lead to a crash.
2648 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2649 and daemon.c were passed as pointers to ints; they should have been
2650 pointers to socklen_t variables (which are typically unsigned ints).
2652 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2655 3. Fixed a really odd bug that affected only the testing scheme; patching a
2656 certain fixed string in the binary changed the value of another string that
2657 happened to be identical to the end of the original first string.
2659 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2660 name", it returns that address as the IP address. On some operating
2661 systems (e.g. Solaris), it also passes back the IP address string as the
2662 "host name". However, on others (e.g. Linux), it passes back an empty
2663 string. Exim wasn't checking for this, and was changing the host name to an
2664 empty string, assuming it had been canonicized.
2666 5. Although rare, it is permitted to have more than one PTR record for a given
2667 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2668 all the names associated with an address, because they do in Solaris.
2669 However, it seems that they do not in Linux for data that comes from the
2670 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2671 I found this out when I moved to a new Linux workstation and tried to run
2672 the Exim test suite.
2674 To get round this problem I have changed the code so that it now does its
2675 own call to the DNS to look up PTR records when searching for a host name.
2676 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2677 addresses that are only in /etc/hosts are still found.
2679 This behaviour is, however, controlled by an option called host_lookup_
2680 order, which defaults to "bydns:byaddr". If people want to use the other
2681 order, or indeed, just use one or the other means of lookup, they can
2682 specify it in this variable.
2684 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2685 some operating systems, this comes back from gethostbyaddr() as an empty
2686 string, and this is what Exim used to test for. However, it seems that in
2687 other systems, "." is yielded. Exim now tests for this case too.
2689 7. The values of check_spool_space and check_log_space are now held internally
2690 as a number of kilobytes instead of an absolute number of bytes. If a
2691 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2692 kilobyte. This means that much larger values can be stored.
2694 8. Exim monitor: an attempt to get the action menu when not actually pointing
2695 at a message produces an empty menu entitled "No message selected". This
2696 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2697 no entries in it ("Shell widget menu has zero width and/or height"). So I
2698 have added a single, blank menu entry in this case.
2700 9. Added ${quote_local_part.
2702 10. MIME decoding is now applied to the contents of Subject: header lines when
2705 11. Now that a reference to $sender_host_address automatically causes a reverse
2706 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2707 host lookup before query-style lookups in lists that might use this
2708 variable. This has therefore been abolished, and the "net-" prefix is no
2709 longer necessary for query-style lookups.
2711 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2712 have been a typo for LFLAGS, so it has been changed.
2714 13. The install script calls Exim with "-C /dev/null" in order to find the
2715 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2716 to be output. Howeve, since Exim outputs its version number before the
2717 error, it didn't break the script. It just looked ugly. I fixed this by
2718 always allowing "-C /dev/null" if the caller is root.
2720 14. Ignore overlarge ACL variable number when reading spool file - insurance
2721 against a later release with more variables having written the file.
2723 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2724 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2725 this, as well as the form without the "IPv6" on the front.
2727 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2728 OS/Makefile-Darwin file.
2730 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2731 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2733 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2734 variable. However, some OS use ints for inodes. I've added cast to long int
2735 to get rid of the compiler warning.
2737 19. I had forgotten to lock out "/../" in configuration file names when
2738 ALT_CONFIG_PREFIX was set.
2740 20. Routers used for verification do not need to specify transports. However,
2741 if such a router generated a host list, and callout was configured, Exim
2742 crashed, because it could not find a port number from the (non-existent)
2743 transport. It now assumes port 25 in this circumstance.
2745 21. Added the -t option to exigrep.
2747 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2748 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2749 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2750 them is set, LOOKUP_LSEARCH is forced.
2752 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2754 24. Added sender and host information to the "rejected by local_scan()" log
2755 line; previously there was no indication of these.
2757 25. Added .include_if_exists.
2759 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2760 Linux. It turns out that not all file systems support this. Apparently some
2761 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2762 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2763 not supported on the file descriptor, is now ignored when Exim is trying to
2764 sync a directory. This applies only to Linux.
2766 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2768 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2769 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2772 29. Added support for maildirsize files from supplied patch (modified a bit).
2774 30. The use of :fail: followed by an empty string could lead Exim to respond to
2775 sender verification failures with (e.g.):
2777 550 Verification failed for <xxx>
2778 550 Sender verify failed
2780 where the first response line was missing the '-' that indicates it is not
2781 the final line of the response.
2783 31. The loop for finding the name of the user that called Exim had a hardwired
2784 limit of 10; it now uses the value of finduser_retries, which is used for
2785 all other user lookups.
2787 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2790 33. Exim was neglecting to zero errno before one call of strtol() when
2791 expanding a string and expecting an integer value. On some systems this
2792 resulted in spurious "integer overflow" errors. Also, it was casting the
2793 result into an int without checking.
2795 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2796 did not work. The code looks as if it has *never* worked, though it appears
2797 to have been documented since at least releast 1.62. I have made it work.
2799 35. The "timeout_DNS" error in retry rules, also documented since at least
2800 1.62, also never worked. As it isn't clear exactly what this means, and
2801 clearly it isn't a major issue, I have abolished the feature by treating it
2802 as "timeout", and writing a warning to the main and panic logs.
2804 36. The display of retry rules for -brt wasn't always showing the error code
2807 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2808 timeout_connect_A, timeout_connect_MX.
2810 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2811 to the empty sender.
2813 39. The daemon was not analysing the content of -oX till after it had closed
2814 stderr and disconnected from the controlling terminal. This meant that any
2815 syntax errors were only noted on the panic log, and the return code from
2816 the command was 0. By re-arranging the code a little, I've made the
2817 decoding happen first, so such errors now appear on stderr, and the return
2818 code is 1. However, the actual setting up of the sockets still happens in
2819 the disconnected process, so errors there are still only recorded on the
2822 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2823 connections (as happens on some IP stacks) was logged at start up time as
2824 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2825 it from "IPv6 and IPv4", which means that two separate sockets are being
2828 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2829 says whether AF_INET or AF_INET6 was passed as an argument.
2831 42. Exiwhat output was messed up when time zones were included in log
2834 43. Exiwhat now gives more information about the daemon's listening ports,
2835 and whether -tls-on-connect was used.
2837 44. The "port" option of the smtp transport is now expanded.
2839 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2840 silently ignored. Now an error message is written to the main and panic
2843 46. There's a new ACL modifier called "logwrite" which writes to a log file
2844 as soon as it is encountered.
2846 47. Added $local_user_uid and $local_user_gid at routing time.
2848 48. Exim crashed when trying to verify a sender address that was being
2851 49. Exim was recognizing only a space character after ".include". It now also
2852 recognizes a tab character.
2854 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2855 extracting the relevant information from the specification. The man page no
2856 longer contains scrambled data for the -d option, and I've added a section
2857 at the front about calling Exim under different names.
2859 51. Added "extra_headers" argument to the "mail" command in filter files.
2861 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2864 53. Installed eximstats 1.29.
2866 54. Added transport_filter_timeout as a generic transport option.
2868 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2869 Cc: header lines. This was required by RFC 822, but it not required by RFC
2872 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2873 incoming messages that did not have them. Now it does so only if the
2874 message originates locally, that is, if there is no associated remote host
2875 address. When Resent- header lines are present, this applies to the Resent-
2876 lines rather than the non-Resent- lines.
2878 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2879 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2881 58. Messages for configuration errors now include the name of the main
2882 configuration file - useful now that there may be more than one file in a
2883 list (.included file names were always shown).
2885 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2886 for those rare installations that do not start the daemon as root or run it
2887 setuid root. I've cut out the call to initgroups() if the daemon is not
2890 60. The Exim user and group can now be bound into the binary as text strings
2891 that are looked up at the start of Exim's processing.
2893 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2895 62. Added $mailstore_basename variable.
2897 63. Installed patch to sieve.c from Michael Haardt.
2899 64. When Exim failed to open the panic log after failing to open the main log,
2900 the original message it was trying to log was written to stderr and debug
2901 output, but if they were not available (the usual case in production), it
2902 was lost. Now it is written to syslog before the two lines that record the
2903 failures to open the logs.
2905 65. Users' Exim filters run in subprocesses under the user's uid. It is
2906 possible for a "deliver" command or an alias in a "personal" command to
2907 provoke an address rewrite. If logging of address rewriting is configured,
2908 this fails because the process is not running as root or exim. There may be
2909 a better way of dealing with this, but for the moment (because 4.30 needs
2910 to be released), I have disabled address rewrite logging when running a
2911 filter in a non-root, non-exim process.
2917 1. The buildconfig auxiliary program wasn't quoting the value set for
2918 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2919 not defined. This bug was masked in 4.22 by the effect that was fixed in
2922 2. Some messages that were rejected after a message id was allocated were
2923 shown as "incomplete" by exigrep. It no longer does this for messages that
2924 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2926 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2927 have allow_domain_literals set, the ID did not get logged in the <= line.
2928 Domain literals are now always recognized in Message-ID: header lines.
2930 4. The first argument for a ${extract expansion item is the key name or field
2931 number. Leading and trailing spaces in this item were not being ignored,
2932 causing some misleading effects.
2934 5. When deliver_drop_privilege was set, single queue runner processes started
2935 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2936 same command in the process it spins off) were not dropping privilege.
2938 6. When the daemon running as "exim" started a queue runner, it always
2939 re-executed Exim in the spun-off process. This is a waste of effort when
2940 deliver_drop_privilege is set. The new process now just calls the
2941 queue-runner function directly.
2947 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2950 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2951 ignored. Though the use of -oP was forcing the writing of a pid file, it
2952 was always written to the default place.
2954 3. If the message "no IP address found for host xxxx" is generated during
2955 incoming verification, it is now followed by identification of the incoming
2956 connection (so you can more easily find what provoked it).
2958 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2960 5. Added some features to "harden" Exim a bit more against certain attacks:
2962 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2963 be put in Local/Makefile. This is like the never_users runtime option,
2964 but it cannot be overridden. The default setting is "root".
2966 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2967 prefix string with which any file named in a -C command line option
2970 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2971 is retained for -C and -D only if the caller of Exim is root. Without
2972 it, the exim user may also use -C and -D and retain privilege.
2974 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2975 command line option is disabled.
2977 6. Macro names set by the -D option must start with an upper case letter, just
2978 like macro names defined in the configuration file.
2980 7. Added "dereference=" facility to LDAP.
2982 8. Two instances of the typo "uknown" in the source files are fixed.
2984 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2985 the Configure-Makefile script screwed up while processing it.
2987 10. Incorporated PCRE 4.4.
2989 11. The SMTP synchronization check was not operating right at the start of an
2990 SMTP session. For example, it could not catch a HELO sent before the client
2991 waited for the greeting. There is now a check for outstanding input at the
2992 point when the greeting is written. Because of the duplex, asynchronous
2993 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2994 way, but not yet received, when the check is performed.
2996 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2997 on TCP/IP sockets, because this apparently causes some broken clients to
3000 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3001 unchanged) from the Cygwin maintainer.
3003 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3004 was supported instead of testing for mbx. Effectively a typo.
3006 15. The spa authenticator server code was not checking that the input it
3007 received was valid base64.
3009 16. The debug output line for the "set" modifier in ACLs was not showing the
3010 name of the variable that was being set.
3012 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3014 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3015 lookup on its own. Something else was needed to trigger the lookup. For
3016 example, a match in host_lookup or the need for a host name in a host list.
3017 Now, if $sender_host_name is referenced and the host name has not yet been
3018 looked up, a lookup is performed. If the lookup fails, the variable remains
3019 empty, and $host_lookup_failed is set to "1".
3021 19. Added "eqi" as a case-independent comparison operator.
3023 20. The saslauthd authentication condition could segfault if neither service
3024 nor realm was specified.
3026 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3027 error message that was logged was misleading, and incoming SMTP
3028 connections were dropped. The message is now more accurate, and temporary
3029 errors are given to SMTP connections.
3031 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3032 (except RSET) with a 421 error, until QUIT is received. However, it was
3033 failing to send a response to QUIT.
3035 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3036 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3037 ACL is now run after the helo_try_verify_hosts code.
3039 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3040 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3041 case-independent, so other case variants are also recognized). Apparently
3042 some systems use these upper case variants.
3044 25. If more than two messages were waiting for the same host, and a transport
3045 filter was specified for the transport, Exim sent two messages over the
3046 same TCP/IP connection, and then failed with "socket operation on non-
3047 socket" when it tried to send the third.
3049 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3051 27. The extern definition of crypt16() in expand.c was not being excluded when
3052 the OS had its own crypt16() function.
3054 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3055 as a preferred synonym for return_size_limit, both as an option and as an
3058 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3060 30. Changed the default configuration ACL to relax the local part checking rule
3061 for addresses that are not in any local domains. For these addresses,
3062 slashes and pipe symbols are allowed within local parts, but the sequence
3063 /../ is explicitly forbidden.
3065 31. SPA server authentication was not clearing the challenge buffer before
3068 32. log_message in a "warn" ACL statement was writing to the reject log as
3069 well as to the main log, which contradicts the documentation and doesn't
3070 seem right (because no rejection is happening). So I have stopped it.
3072 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3073 However, I am unable to do any testing of this.
3075 34. Fixed an infelicity in the appendfile transport. When checking directories
3076 for a mailbox, to see if any needed to be created, it was accidentally
3077 using path names with one or more superfluous leading slashes; tracing
3078 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3080 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3081 contents are added to the log line that is written for every discarded
3082 recipient. (Previously a log_message setting was ignored.)
3084 36. The ${quote: operator now quotes the string if it is empty.
3086 37. The install script runs exim in order to find its version number. If for
3087 some reason other than non-existence or emptiness, which it checks, it
3088 could not run './exim', it was installing it with an empty version number,
3089 i.e. as "exim-". This error state is now caught, and the installation is
3092 38. An argument was missing from the function that creates an error message
3093 when Exim fails to connect to the socket for saslauthd authentication.
3094 This could cause Exim to crash, or give a corrupted message.
3096 39. Added isip, isip4, and isip6 to ${if conditions.
3098 40. The ACL variables $acl_xx are now saved with the message, and can be
3099 accessed later in routers, transports, and filters.
3101 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3102 strings in the file are not string-expanded.
3104 42. If a MAIL command specified a SIZE value that was too large to fit into an
3105 int variable, the check against message_size_limit failed. Such values are
3106 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3107 day this will have to be increased, but I don't think I want to be around
3108 when emails are that large.
3115 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3116 iconv() is not standard in FreeBSD.
3118 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3119 IPv6 enabled. The observed symptom was a segmentation fault on return from
3120 the function os_common_find_running_interfaces() in src/os.c.
3122 3. In the check_special_case() function in daemon.c I had used "errno" as an
3123 argument name, which causes warnings on some systems. This was basically a
3124 typo, since it was named "eno" in the comments!
3126 4. The code that waits for the clock to tick (at a resolution of some fraction
3127 of a second) so as to ensure message-id uniqueness was always waiting for
3128 at least one whole tick, when it could have waited for less. [This is
3129 almost certainly not relevant at current processor speeds, where it is
3130 unlikely to ever wait at all. But we try to future-proof.]
3132 5. The function that sleeps for a time interval that includes fractions of a
3133 second contained a race. It did not block SIGALRM between setting the
3134 timer, and suspending (a couple of lines later). If the interval was short
3135 and the sigsuspend() was delayed until after it had expired, the suspension
3136 never ended. On busy systems this could lead to processes getting stuck for
3139 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3140 process, before it forks any delivery processes. The open lookup caching
3141 mechanism meant that the open file or database connection was passed into
3142 the delivery process. The problem was that delivery processes always tidy
3143 up cached lookup data. This could cause a problem for the next delivery
3144 process started by the queue runner, because the external queue runner
3145 process does not know about the closure. So the next delivery process
3146 still has data in the lookup cache. In the case of a file lookup, there was
3147 no problem because closing a file descriptor in a subprocess doesn't affect
3148 the parent. However, if the lookup was caching a connection to a database,
3149 the connection was closed, and the second delivery process was likely to
3150 see errors such as "PGSQL: query failed: server closed the connection
3151 unexpectedly". The problem has been fixed by closing all cached lookups
3152 in a queue runner before running a delivery process.
3154 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3155 seem to have the "const" qualifier which it has on other OS. I've
3158 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3159 *of the same type* (client or server) with the same public name that an
3160 error should be diagnosed.
3162 9. When Exim looked up a host name for an IP address, but failed to find the
3163 original IP address when looking up the host name (a safety check), it
3164 output the message "<ip address> does not match any IP for NULL", which was
3165 confusing, to say the least. The bug was that the host name should have
3166 appeared instead of "NULL".
3168 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3169 user that is built into the binary, and the -C or -D options is used, root
3170 privilege is dropped before the configuration file is read. In addition,
3171 logging is switched to stderr instead of the normal log files. If the
3172 configuration then re-defines the Exim user, the unprivileged environment
3173 is probably not what is expected, so Exim logs a panic warning message (but
3176 However, if deliver_drop_privilege is set, the unprivileged state may well
3177 be exactly what is intended, so the warning has been cut out in that case,
3178 and Exim is allowed to try to write to its normal log files.
3184 1. smtp_return_error_details was not giving details for temporary sender
3185 or receiver verification errors.
3187 2. Diagnose a configuration error if two authenticators have the same public
3190 3. Exim used not to create the message log file for a message until the first
3191 delivery attempt. This could be confusing when incoming messages were held
3192 for policy or load reasons. The message log file is now created at the time
3193 the message is received, and an initial "Received" line is written to it.
3195 4. The automatically generated man page for command line options had a minor
3196 bug that caused no ill effects; however, a more serious problem was that
3197 the procedure for building the man page automatically didn't always
3198 operate. Consequently, release 4.20 contains an out-of-date version. This
3199 shouldn't happen again.
3201 5. When building Exim with embedded Perl support, the script that builds the
3202 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3203 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3205 6. The freeze_tell option was not being used for messages that were frozen on
3206 arrival, either by an ACL or by local_scan().
3208 7. Added the smtp_incomplete_transaction log selector.
3210 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3211 was accepting AUTH without a new EHLO.
3213 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3214 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3217 10. Logging of TCP/IP connections (when configured) now happens in the main
3218 daemon process instead of the child process, so that the TCP/IP connection
3219 count is more accurate (but it can never be perfect).
3221 11. The use of "drop" in a nested ACL was not being handled correctly in the
3222 outer ACL. Now, if condition failure induced by the nested "drop" causes
3223 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3224 or "require"), the connection is dropped.
3226 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3227 that yield "discard" can now be used with an "accept" or a "discard" verb,
3228 but an error is generated for any others (because I can't see a useful way
3229 to define what should happen).
3231 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3232 are now processed in the same way as lines in the Exim configuration file.
3233 In particular, continuation lines are supported.
3235 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3237 15. Added -ti meaning -t -i.
3239 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3240 domains, and warn by logging if others are found.
3242 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3243 returned from GIFCONF when Exim is trying to find the list of interfaces on
3244 a host. The code in os.c has been modified to copy each ifreq to an aligned
3245 structure in all cases.
3247 Also, in some cases, the returned ifreq's were being copied to a 'struct
3248 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3249 means the last couple of bytes of an IPv6 address could be chopped if the
3250 ifreq contained only a normal sockaddr (14 bytes storage).
3252 18. Named domain lists were not supported in the hosts_treat_as_local option.
3253 An entry such as +xxxx was not recognized, and was treated as a literal
3256 19. Ensure that header lines added by a DATA ACL are included in the reject log
3257 if the ACL subsequently rejects the message.
3259 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3260 MD5 (which is deprecated).
3262 21. When testing a filter file using -bf, Exim was writing a message when it
3263 took the sender from a "From " line in the message, but it was not doing so
3264 when it took $return_path from a Return-Path: header line. It now does.
3266 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3267 with a valid header line field name (a series of printing characters
3268 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3270 23. Changed "disc" in the source to "disk" to conform to the documentation and
3271 the book and for uniformity.
3273 24. Ignore Sendmail's -Ooption=value command line item.
3275 25. When execve() failed while trying to run a command in a pipe transport,
3276 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3277 could be confused with a return value of 69 from the command itself. This
3278 has been changed to 127, the value the shell returns if it is asked to run
3279 a non-existent command. The wording for the related log line suggests a
3280 non-existent command as the problem.
3282 26. If received_header_text expands to an empty string, do not add a Received:
3283 header line to the message. (Well, it adds a token one on the spool, but
3284 marks it "old" so that it doesn't get used or transmitted.)
3286 27. Installed eximstats 1.28 (addition of -nt option).
3288 28. There was no check for failure on the call to getsockname() in the daemon
3289 code. This can fail if there is a shortage of resources on the system, with
3290 ENOMEM, for example. A temporary error is now given on failure.
3292 29. Contrary to the C standard, it seems that in some environments, the
3293 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3294 program. Exim now does this explicitly; it affects the formatting of
3295 timestamps using strftime().
3297 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3298 complaints. I've now initialized all the variables, to avoid this.
3300 32. Header lines added by a system filter were not being "seen" during
3301 transport-time rewrites.
3303 33. The info_callback() function passed to OpenSSL is set up with type void
3304 (*)(SSL *, int, int), as described somewhere. However, when calling the
3305 function (actually a macro) that sets it up, the type void(*)() is
3306 expected. I've put in a cast to prevent warnings from picky compilers.
3308 34. If a DNS black list lookup found a CNAME record, but there were no A
3309 records associated with the domain it pointed at, Exim crashed.
3311 35. If a DNS black list lookup returned more than one A record, Exim ignored
3312 all but the first. It now scans all returned addresses if a particular IP
3313 value is being sought. In this situation, the contents of the
3314 $dnslist_value variable are a list of all the addresses, separated by a
3317 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3318 to accept a host name and all its aliases if the forward lookup for any of
3319 them yielded the IP address of the incoming connection. Now it accepts only
3320 those names whose forward lookup yields the correct IP address. Any other
3321 names are discarded. This closes a loophole whereby a rogue DNS
3322 administrator could create reverse DNS records to break through a
3323 wildcarded host restriction in an ACL.
3325 37. If a user filter or a system filter that ran in a subprocess used any of
3326 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3327 the wrong values were passed to the pipe command ($thisaddress had the
3328 value of $0, $0 had the value of $1, etc). This bug was introduced by
3329 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3331 38. Improved the line breaking for long SMTP error messages from ACLs.
3332 Previously, if there was no break point between 40 and 75 characters, Exim
3333 left the rest of the message alone. Two changes have been made: (a) I've
3334 reduced the minimum length to 35 characters; (b) if it can't find a break
3335 point between 35 and 75 characters, it looks ahead and uses the first one
3336 that it finds. This may give the occasional overlong line, but at least the
3337 remaining text gets split now.
3339 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3340 file descriptors might be low, and that setting 1000 would always raise it.
3341 It turns out that in some environments, the limit is already over 1000 and
3342 that lowering it causes trouble. So now Exim takes care not to decrease it.
3344 40. When delivering a message, the value of $return_path is set to $sender_
3345 address at the start of routing (routers may change the value). By an
3346 oversight, this default was not being set up when an address was tested by
3347 -bt or -bv, which affected the outcome if any router or filter referred to
3350 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3351 log only when "message" or "log_message" are set. However, if one of the
3352 conditions was an address verification, or a call to a nested ACL, the
3353 messages generated by the underlying test were being passed through. This
3354 no longer happens. The underlying message is available in $acl_verify_
3355 message for both "message" and "log_message" expansions, so it can be
3356 passed through if needed.
3358 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3359 new expansion $bh_ to give the encoded byte string without charset
3360 translation. Translation happens only if iconv() is available; HAVE_ICONV
3361 indicates this at build time. HEADERS_CHARSET gives the charset to
3362 translate to; headers_charset can change it in the configuration, and
3363 "headers charset" can change it in an individual filter file.
3365 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3366 that creates RFC 2047 encoded "words" labels them as that charset instead
3367 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3368 expansion operator; (ii) when Exim creates a From: line for a local
3369 message; (iii) when a header line is rewritten to include a "phrase" part.
3371 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3372 buggy, causing it to skip the first lines of messages whose message ID
3373 ended in 'D'. This would not have bitten before Exim release 4.14, because
3374 message IDs were unlikely to end in 'D' before then. The effect was to have
3375 incorrect size information for certain domains.
3377 45. #include "config.h" was missing at the start of the crypt16.c module. This
3378 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3381 46. If there was a timeout during a "random" callout check, Exim treated it as
3382 a failure of the random address, and carried on sending RSET and the real
3383 address. If the delay was just some slowness somewhere, the response to the
3384 original RCPT would be taken as a response to RSET and so on, causing
3385 mayhem of various kinds.
3387 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3388 when I implemented it. It didn't allow for the fact that some option values
3389 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3390 the right test for positive values.
3392 48. Domain names in DNS records are case-independent. Exim always looks them up
3393 in lower case. Some resolvers return domain names in exactly the case they
3394 appear in the zone file, that is, they may contain uppercase letters. Not
3395 all resolvers do this - some return always lower case. Exim was treating a
3396 change of case by a resolver as a change of domain, similar to a widening
3397 of a domain abbreviation. This triggered its re-routing code and so it was
3398 trying to route what was effectively the same domain again. This normally
3399 caused routing to fail (because the router wouldn't handle the domain
3400 twice). Now Exim checks for this case specially, and just changes the
3401 casing of the domain that it ultimately uses when it transmits the message
3404 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3407 50. If a filter generated a file delivery with a non-absolute name (possible if
3408 no home directory exists for the router), the forbid_file option was not
3411 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3412 the existing equality matching.
3414 52. Exim was using ints instead of ino_t variables in some places where it was
3415 dealing with inode numbers.
3417 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3418 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3419 called TMPDIR, and if it finds it is different, it changes its value.
3421 54. The smtp_printf() function is now made available to local_scan() so
3422 additional output lines can be written before returning. There is also an
3423 smtp_fflush() function to enable the detection of a dropped connection.
3424 The variables smtp_input and smtp_batched_input are exported to
3427 55. Changed the default runtime configuration: the message "Unknown user"
3428 has been removed from the ACL, and instead placed on the localuser router,
3429 using the cannot_route_message feature. This means that any verification
3430 failures that generate their own messages won't get overridden. Similarly,
3431 the "Unrouteable address" message that was in the ACL for unverifiable
3432 relay addresses has also been removed.
3434 56. Added hosts_avoid_esmtp to the smtp transport.
3436 57. The exicyclog script was not checking for the esoteric option
3437 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3438 will work only if exicyclog is run under the appropriate euid.
3440 58. Following a discussion on the list, the rules by which Exim recognises line
3441 endings on incoming messages have been changed. The -dropcr and drop_cr
3442 options are now no-ops, retained only for backwards compatibility. The
3443 following line terminators are recognized: LF CRLF CR. However, special
3444 processing applies to CR:
3446 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3447 nor a local message in the state where . is a terminator.
3449 (ii) If a bare CR is encountered in a header line, an extra space is added
3450 after the line terminator so as not to end the header. The reasoning
3451 behind this is that bare CRs in header lines are most likely either
3452 to be mistakes, or people trying to play silly games.
3454 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3455 was being incorrectly given as 18 bytes larger than it should have been.
3456 This is a VOB (very old bug).
3458 60. This may never have affected anything current, but just in case it has:
3459 When the local host is found other than at the start of a list of hosts,
3460 the local host, those with the same MX, and any that follow, are discarded.
3461 When the list in question was part of a longer list of hosts, the following
3462 hosts (not currently being processed) were also being discarded. This no
3463 longer happens. I'm not sure if this situation could ever has previously
3466 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3469 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3470 header. This is something that is recommended in a new Internet Draft, and
3471 is something that is documented as being done by Sendmail. There are two
3472 possible values. For messages generated by the autoreply transport, Exim
3475 Auto-Submitted: auto-replied
3477 whereas for all other generated messages (e.g. bounces) it adds
3479 Auto-Submitted: auto-generated
3481 63. The "personal" condition in filters now includes a test for the
3482 Auto-Submitted: header. If it contains the string "auto-" the message it
3483 not considered personal.
3485 64. Added rcpt_include_affixes as a generic transport option.
3487 65. Added queue_only_override (default true).
3489 66. Added the syslog_duplication option.
3491 67. If what should have been the first header line of a message consisted of
3492 a space followed by a colon, Exim was mis-interpreting it as a header line.
3493 It isn't of course - it is syntactically invalid and should therefore be
3494 treated as the start of the message body. The misbehaviour could have
3495 caused a number of strange effects, including loss of data in subsequent
3496 header lines, and spool format errors.
3498 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3499 client host had authenticated. This control can now be exercised by an ACL
3500 for more flexibility.
3502 69. By default, callouts do not happen when testing with -bh. There is now a
3503 variant, -bhc, which does actually run the callout code, including
3504 consulting and updating the callout cache.
3506 70. Added support for saslauthd authentication, courtesy of Alexander
3509 71. If statvfs() failed on the spool or log directories while checking their
3510 size for availability, Exim confusingly gave the error "space shortage".
3511 Furthermore, in debugging mode it crashed with a floating point exception.
3512 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3513 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3514 serious problem, Exim now writes to the main and panic logs when this
3515 happens, with details of the failure. It then refuses to accept the
3516 incoming message, giving the message "spool directory problem" or "log
3517 directory problem" with a 421 code for SMTP messages.
3519 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3520 0, 1, and 2 exist, because some OS complain for execs without them (see
3521 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3522 descriptors. However, the code omitted to check that the open succeeded,
3523 causing mysterious errors if for some reason the permissions on /dev/null
3524 got screwed. Now Exim writes a message to the main and panic logs, and
3525 bombs out if it can't open /dev/null.
3527 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3528 interact so that it is all more flexible. It is supposed to remain
3529 backwards compatible. Also added extra_local_interfaces.
3531 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3532 to bomb out with an assertion failure - to the client this appears as a
3533 connection drop. This problem occurs in the part of the code that was taken
3534 from the Samba project. Fortunately, the assertion is in a very simple
3535 function, so I have fixed this by reproducing the function inline in the
3536 one place where it is called, and arranging for authentication to fail
3537 instead of killing the process with assert().
3539 75. The SPA client code was not working when the server requested OEM rather
3540 than Unicode encoding.
3542 76. Added code to make require_files with a specific uid setting more usable in
3543 the case where statting the file as root fails - usually a non-root-mounted
3544 NFS file system. When this happens and the failure is EACCES, Exim now
3545 forks a subprocess and does the per-uid checking as the relevant uid.
3547 77. Added process_log_path.
3549 78. If log_file_path was not explicitly set, a setting of check_log_space or
3550 check_log_inodes was ignored.
3552 79. If a space check for the spool or log partitions fails, the incident is now
3553 logged. Of course, in the latter case the data may get lost...
3555 80. Added the %p formatting code to string_format() so that it can be used to
3556 print addresses in debug_print(). Adjusted all the address printing in the
3557 debugging in store.c to use %p rather than %d.
3559 81. There was a concern that a line of code in smtp_in.c could overflow a
3560 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3561 initially expressed, the concern was not well-founded, because trailing
3562 spaces are removed early. However, if the trailing spaces were followed by
3563 a NULL, they did not get removed, so the overflow was possible. Two fixes
3566 (a) I re-wrote the offending code in a cleaner fashion.
3567 (b) If an incoming SMTP command contains a NULL character, it is rejected
3570 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3571 runs initgroups(), so that if the Exim user is in any additional groups,
3572 they will be used during message reception.
3578 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob