1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.257 2005/11/14 11:41:23 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
18 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
37 PH/03 For some reason, the default setting of PATH when running a command from
38 a pipe transport was just "/usr/bin". I have changed it to
41 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
42 anything to be listed in the output from -bV.
44 PH/05 When a filter generated an autoreply, the entire To: header line was
45 quoted in the delivery log line, like this:
47 => >A.N.Other <ano@some.domain> <original@ddress> ...
49 This has been changed so that it extracts the operative address. There
50 may be more than one such address. If so, they are comma-separated, like
53 => >ano@some.domain,ona@other.domain <original@ddress> ...
55 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
56 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
57 not being looked up in the rDNS to get a host name, Exim was showing the
58 IP address twice in Received: lines, even though the IP addresses were
59 identical. For example:
61 Received: from [1.2.3.4] (helo=[1.2.3.4])
63 However, if the real host name was known, it was omitting the HELO data
64 if it matched the actual IP address. This has been tidied up so that it
65 doesn't show the same IP address twice.
67 PH/07 When both +timestamp and +memory debugging was on, the value given by
68 $tod_xxx expansions could be wrong, because the tod_stamp() function was
69 called by the debug printing, thereby overwriting the timestamp buffer.
70 Debugging no longer uses the tod_stamp() function when +timestamp is set.
72 PH/08 When the original message was included in an autoreply transport, it
73 always said "this is a copy of the message, including all the headers",
74 even if body_only or headers_only was set. It now gives an appropriate
77 PH/09 Applied a patch from the Sieve maintainer which:
80 o adds the (disabled) notify extension core
81 o adds some debug output for the result of if/elsif tests
82 o points to the current vacation draft in the documentation
83 and documents the missing references header update
87 o fixes a bug in processing the envelope test (when testing
88 multiple envelope elements, the last element determinted the
95 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
96 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
99 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
100 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
102 PH/03 Typo: missing ".o" in src/pcre/Makefile.
104 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
105 header line, restrict the check to what is listed in RFCs 2369 and 2929.
106 Also, for "Auto-Submitted", treat anything other than "no" as
107 non-personal, in accordance with RFC 3834. (Previously it treated
108 anything starting "auto-" as non-personal.)
110 TF/01 The control=submission/name=... option had a problem with syntax
111 errors if the name included a slash character. The /name= option
112 now slurps the rest of the string, so it can include any characters
113 but it must come last in the list of options (after /sender_retain
116 PH/05 Some modifications to the interface to the fake nameserver for the new
123 TK/01 Added the "success_on_redirect" address verification option. See
124 NewStuff for rationale and an example.
126 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
128 PH/02 Patch to exigrep to allow it to work on syslog lines.
130 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
131 fread() to skip over the body file's header line, because in Cygwin the
132 header line is locked and is inaccessible.
134 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
135 co-exist for some time) to make it clear that it is the Exim ID that is
136 referenced, not the Message-ID: header line.
138 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
139 string_format() function, because snprintf() does not exist on all
142 PH/06 The use of forbid_filter_existstest now also locks out the use of the
143 ${stat: expansion item.
145 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
146 protocol synchronization error", to keep the pedants happy.
148 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
149 well as for IRIX systems, when gcc is being used. See the host.c source
152 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
154 PH/10 Named domain lists were not working if used in a queue_smtp_domains
157 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
158 transport and to the smtp transport in LMTP mode.
160 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
162 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
163 run a filter in a subprocess. This could lead to confusion in subsequent
164 lookups in the parent process. There should also be a search_tidyup() at
165 the end of the subprocess.
167 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
168 only if the host matched helo_try_verify_hosts, which caused the
169 verification to occur when the EHLO/HELO command was issued. The ACL just
170 tested the remembered result. Now, if a previous verification attempt has
171 not happened, "verify = helo" does it there and then.
173 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
175 TK/03 Fix log output including CR from clamd.
177 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
178 exist provoked a memory error which could cause a segfault.
180 PH/15 Installed PCRE 6.2
182 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
184 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
185 of the problem. Specifically, suggested +O2 rather than +O1 for the
188 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
190 PH/20 If a delivery was routed to a non-standard port by means of an SRV
191 record, the port was not correctly logged when the outgoing_port log
192 selector was set (it logged the transort's default port).
194 PH/21 Added support for host-specific ports to manualroute, queryprogram,
195 fallback_hosts, and "hosts" in the smtp transport.
197 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
198 host errors such as "Connection refused".
200 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
201 authentication with radiusclient 0.4.9:
203 - Error returned from rc_read_config was caught wrongly
204 - Username/password not passed on to radius server due to wrong length.
206 The presumption is that some radiusclient API changes for 4.51/PH/17
207 were not taken care of correctly. The code is still untested by me (my
208 Linux distribution still has 0.3.2 of radiusclient), but it was
209 contributed by a Radius user.
211 PH/24 When doing a callout, the value of $domain wasn't set correctly when
212 expanding the "port" option of the smtp transport.
214 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
215 while reading a MIME header. Thanks to Tom Hughes for a patch.
217 PH/24 Include config.h inside local_scan.h so that configuration settings are
220 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
221 that in an ACL for RCPT (for example), you can examine exactly what was
224 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
225 commands, but it was not correctly comparing the address with the actual
226 client host address. Thus, it would show the EHLO address in Received:
227 header lines when this was not necessary.
229 PH/27 Added the % operator to ${eval:}.
231 PH/28 Exim tries to create and chdir to its spool directory when it starts;
232 it should be ignoring failures (because with -C, for example, it has lost
233 privilege). It wasn't ignoring creation failures other than "already
236 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
239 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
240 because an input line was too long, either on its own, or by virtue of
241 too many continuations, the temporary file was not being removed, and the
242 return code was incorrect.
244 PH/31 Missing "BOOL" in function definition in filtertest.c.
246 PH/32 Applied Sieve patches from the maintainer.
248 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
250 PH/33 Added "verify = not_blind".
252 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
253 Local/Makefile (with some defaults set). These are used in built scripts
254 such as exicyclog, but they have never been used in the exim_install
255 script (though there are many overriding facilities there). I have
256 arranged that the exim_install script now takes note of these two
259 PH/35 Installed configuration files for Dragonfly.
261 PH/36 When a locally submitted message by a trusted user did not contain a
262 From: header, and the sender address was obtained from -f or from an SMTP
263 MAIL command, and the trusted user did not use -F to supply a sender
264 name, $originator_name was incorrectly used when constructing a From:
265 header. Furthermore, $originator_name was used for submission mode
266 messages from external hosts without From: headers in a similar way,
267 which is clearly wrong.
269 PH/37 Added control=suppress_local_fixups.
271 PH/38 When log_selector = +received_sender was set, and the addition of the
272 sender made the log line's construction buffer exactly full, or one byte
273 less than full, an overflow happened when the terminating "\n" was
276 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
277 when the result of a list match is failure because a DNS lookup failed.
279 PH/40 RM_COMMAND is now used in the building process.
281 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
282 the "build-* directories that it finds.
284 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
285 address in a domain literal was a prefix of an interface address.
287 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
288 when verifying a sender address, unless rewrite_headers is false.
290 PH/44 Wrote a long comment about why errors_to addresses are verified as
291 recipients, not senders.
293 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
294 the ratelimit ACL was added.
296 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
298 PH/46 Added extra information about PostgreSQL errors to the error string.
300 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
301 avoiding the need to install special zones in a real server. This is
302 backwards compatible; if it can't find the fake resolver, it drops back.
303 Thus, both old and new test suites can be run.
305 TF/02 Added util/ratelimit.pl
307 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
308 clock is set back in time.
310 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
311 Candler <B.Candler@pobox.com>.
313 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
314 OK for addresses that are the result of redirections.
316 PH/48 A number of further additions for the benefit of the new test suite,
317 including a fake gethostbyname() that interfaces to the fake DNS resolver
320 TF/06 The fix for widen_domains has also been applied to qualify_single and
321 search_parents which are the other dnslookup options that can cause
324 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
327 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
329 TF/07 Exim produced the error message "an SRV record indicated no SMTP
330 service" if it encountered an MX record with an empty target hostname.
331 The message is now "an MX or SRV record indicated no SMTP service".
333 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
334 if the DNS of the sending site is misconfigured. This is quite a
335 common situation. This change restores the behaviour of treating a
336 helo verification defer as a failure.
338 PH/51 If self=fail was set on a router, the bounce message did not include the
339 actual error message.
345 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
347 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
348 command itself ended in error, the underlying message about the transport
349 filter timeout was being overwritten with the pipe command error. Now the
350 underlying error message should be appended to the second error message.
352 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
354 PH/02 Reduce the amount of output that "make" produces by default. Full output
355 can still be requested.
357 PH/03 The warning log line about a condition test deferring for a "warn" verb
358 was being output only once per connection, rather than after each
359 occurrence (because it was using the same function as for successful
360 "warn" verbs). This seems wrong, so I have changed it.
362 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
363 it should not have, which might have caused a crash in the right
364 circumstances, but probably never did.
366 PH/04 Installed a modified version of Tony Finch's patch to make submission
367 mode fix the return path as well as the Sender: header line, and to
368 add a /name= option so that you can make the user's friendly name appear
371 TF/03 Added the control = fakedefer ACL modifier.
373 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
374 Mark Lowes for thorough testing.
376 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
378 TK/03 Merged latest SRS patch from Miles Wilton.
380 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
381 with the definition in sysexits.h (which is #included earlier).
382 Fortunately, Exim does not actually use EX_OK. The code used to try to
383 preserve the sysexits.h value, by assumimg that macro definitions were
384 scanned for macro replacements. I have been disabused of this notion,
385 so now the code just undefines EX_OK before #including unistd.h.
387 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
388 in the smtp transport. When a block could not be written in a single
389 write() function, the timeout was being re-applied to each part-write.
390 This seems wrong - if the receiver was accepting one byte at a time it
391 would take for ever. The timeout is now adjusted when this happens. It
392 doesn't have to be particularly precise.
394 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
395 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
397 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
398 without a domain if the check to <postmaster@domain> fails.
400 SC/01 Eximstats: added -xls and the ability to specify output files
401 (patch written by Frank Heydlauf).
403 SC/02 Eximstats: use FileHandles for outputing results.
405 SC/03 Eximstats: allow any combination of xls, txt, and html output.
407 SC/04 Eximstats: fixed display of large numbers with -nvr option
409 SC/05 Eximstats: fixed merging of reports with empty tables.
411 SC/06 Eximstats: added the -include_original_destination flag
413 SC/07 Eximstats: removed tabs and trailing whitespace.
415 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
417 TK/06 MBOX spool code: Add real "From " MBOX separator line
418 so the .eml file is really in mbox format (even though
419 most programs do not really care). Patch from Alex Miller.
421 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
422 The latter is generated from $received_to and is only set if the
423 message has one envelope recipient. SA can use these headers,
424 obviously out-of-the-box. Patch from Alex Miller.
426 PH/08 The ${def test on a variable was returning false if the variable's
427 value was "0", contrary to what the specification has always said!
428 The result should be true unless the variable is empty.
430 PH/09 The syntax error of a character other than { following "${if
431 def:variable_name" (after optional whitespace) was not being diagnosed.
432 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
433 accidental colon was present, for example, could give incorrect results.
435 PH/10 Tidied the code in a number of places where the st_size field of a stat()
436 result is used (not including appendfile, where other changes are about
439 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
440 This involved changing a lot of size variables from int to off_t. It
441 should work with maildirs and everything.
443 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
444 spamd dying while we are connected to it.
446 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
449 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
450 function for BDB 4.3.
452 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
454 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
455 the sources for PCRE have been re-arranged and more files are now
458 PH/15 The code I had for printing potentially long long variables in PH/11
459 above was not the best (it lost precision). The length of off_t variables
460 is now inspected at build time, and an appropriate printing format (%ld
461 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
462 to be "long long int" or "long int". This is needed for the internal
463 formatting function string_vformat().
465 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
466 the configuration file to be ":syslog", then the script "guesses" where
467 the logs files are, rather than using the compiled in default. In our
468 case the guess is not the same as the compiled default, so the script
469 suddenly stopped working when I started to use syslog. The patch checks
470 to see if log_file_path is "". If so, it attempts to read it from exim
471 with no configuration file to get the compiled in version, before it
472 falls back to the previous guessing code."
474 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
475 implementing BATV in an Exim configuration. See NewStuff for the gory
478 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
479 Makefile that are specific to HP-UX.
481 PH/18 If the "use_postmaster" option was set for a recipient callout together
482 with the "random" option, the postmaster address was used as the MAIL
483 FROM address for the random test, but not for the subsequent recipient
484 test. It is now used for both.
486 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
487 patch removes a few documentation additions to RFC 3028, because the
488 latest draft now contains them. It adds the new en;ascii-case comparator
489 and a new error check for 8bit text in MIME parts. Comparator and
490 require names are now matched exactly. I enabled the subaddress
491 extension, but it is not well tested yet (read: it works for me)."
493 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
494 rework some of the code of TK/09 above to avoid the hardwired use of
495 "%lld" and "long long". Replaced the call to snprintf() with a call to
498 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
499 records point to non-existent hosts", "retry timeout exceeded", and
500 "retry time not reached for any host after a long failure period".
502 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
503 experimental DomainKeys support:
505 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
506 (2) On an error such as an illegally used "control", the wrong name for
507 the control was given.
509 These problems did NOT occur unless DomainKeys support was compiled.
511 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
513 PH/24 Added ${if match_ip condition.
515 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
516 databases so that it will be absolutely obvious if a crash occurs in the
517 DB library. This is a regular occurrence (often caused by mis-matched
520 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
521 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
522 picked up on a user's system that detects such things. There doesn't seem
523 to be a gcc warning option for this - only an attribute that has to be
524 put on the function's prototype. It seems that in Fedora Core 4 they have
525 set this on a number of new functions. No doubt there will be more in due
528 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
529 specify a transport. However, if an address that was verified by such a
530 router was the subject of a callout, Exim crashed because it tried to
531 read the rcpt_include_affixes from the non-existent transport. Now it
532 just assumes that the setting of that option is false. This bug was
533 introduced by 4.51/PH/31.
535 PH/28 Changed -d+all to exclude +memory, because that information is very
536 rarely of interest, but it makes the output a lot bigger. People tend to
537 do -d+all out of habit.
539 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
540 code in os-type was giving problems when libc.so lives in lib64, like on
543 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
544 aren't the modern standard, and it seems that some systems' include files
545 don't always have them. Exim was already checking for some of the newer
546 ones like T_AAAA, and defining it itself. I've added checks for all the
547 record types that Exim uses.
549 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
550 not automatically generating a new one, as it is supposed to. This
551 prevented TLS from working. If the file did exist, but contained invalid
552 data, a new version was generated, as expected. It was only the case of a
553 non-existent file that was broken.
555 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
556 with a change in libdomainkeys > 0.64.
558 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
559 from DNS. If the selector record carries the flag, it now has
560 precedence over the domain-wide flag.
562 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
564 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
565 the use of a port name that isn't defined in /etc/services) occurred, the
566 message was deferred as in a normal delivery, and thus remained on the
567 spool, instead of being failed because of the mua_wrapper setting. This
568 is now fixed, and I tidied up some of the mua_wrapper messages at the
571 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
572 the messages in a hash of arrays rather than using individual hashes.
573 This is a bit cleaner and results in dramatic memory savings, albeit
574 at a slight CPU cost.
576 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
577 as requested by Marc Sherman.
579 SC/10 Eximstats: added histograms for user specified patterns as requested
582 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
584 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
585 fopen() in the content-scanning modules that did not already have it.
591 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
592 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
594 TK/02 Fix ACL "control" statement not being available in MIME ACL.
596 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
598 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
599 to test Sieve filters that use "vacation".
601 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
602 that changes the way the GnuTLS parameters are stored in the cache file.
603 The new format can be generated externally. For backward compatibility,
604 if the data in the cache doesn't make sense, Exim assumes it has read an
605 old-format file, and it generates new data and writes a new file. This
606 means that you can't go back to an older release without removing the
609 PH/03 A redirect router that has both "unseen" and "one_time" set does not
610 work if there are any delivery delays because "one_time" forces the
611 parent to be marked "delivered", so its unseen clone is never tried
612 again. For this reason, Exim now forbids the simultaneous setting of
615 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
616 redirected to themselves ("homonym" addresses). Read the long ChangeLog
617 entry if you want to know the details. The fix, however, neglected to
618 consider the case when local delivery batching is involved. The test for
619 "previously delivered" was not happening when checking to see if an
620 address could be batched with a previous (undelivered) one; under
621 certain circumstances this could lead to multiple deliveries to the same
624 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
625 in its include files, and this causes problems building Exim.
627 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
628 header_syntax) but Exim was just ignoring anything given after a slash.
629 In particular, this caused confusion with an attempt to use "verify =
630 reverse_host_lookup/defer_ok". An error is now given when options are
631 supplied for verify items that do not have them. (Maybe reverse_host_
632 lookup should have a defer_ok option, but that's a different point.)
634 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
635 defined by RFC 821) to 2048, because there were problems with some AUTH
636 commands, and RFC 1869 says the size should be increased for extended
637 SMTP commands that take arguments.
639 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
642 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
643 "unknown" error; now it says that the functionality isn't in the binary.
645 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
646 an address' error message when a string expansion fails (syntax or
647 whatever). Otherwise the password may appear in the log. Following change
648 PH/42 below, there is no longer a chance of it appearing in a bounce
651 PH/11 Installed exipick version 20050225.0 from John Jetmore.
653 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
654 of its addresses was ever tried. (Bugzilla bug #2.)
656 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
657 the result incorrectly in the debug output. (It correctly added a newline
658 to what was transported.)
660 TF/01 Added $received_time.
662 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
663 commented out examples of how to interface to a virus scanner and to
664 SpamAssassin. Also added commented examples of av_scanner and
665 spamd_address settings.
667 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
668 and controls are allowed in which ACLs. There were a couple of minor
669 errors. Some of the entries in the conditions table (which is a table of
670 where they are NOT allowed) were getting very unwieldy; rewrote them as a
671 negation of where the condition IS allowed.
673 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
675 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
676 header file does not have a version number, so I've had to invent a new
677 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
678 API. The code is untested by me (my Linux distribution still has 0.3.2 of
679 radiusclient), but it was contributed by a Radius user.
681 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
682 files or directories, for OpenSSL.
684 PH/19 When an Exim process that is running as root has to create an Exim log
685 file, it does so in a subprocess that runs as exim:exim so as to get the
686 ownership right at creation (otherwise, other Exim processes might see
687 the file with the wrong ownership). There was no test for failure of this
688 fork() call, which would lead to the process getting stuck as it waited
689 for a non-existent subprocess. Forks do occasionally fail when resources
690 run out. I reviewed all the other calls to fork(); they all seem to check
693 PH/20 When checking for unexpected SMTP input at connect time (before writing
694 the banner), Exim was not dealing correctly with a non-positive return
695 from the read() function. If the client had disconnected by this time,
696 the result was a log entry for a synchronization error with an empty
697 string after "input=" when read() returned zero. If read() returned -1
698 (an event I could not check), uninitialized data bytes were printed.
699 There were reports of junk text (parts of files, etc) appearing after
702 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
704 PH/22 Added support for macro redefinition, and (re)definition in between
705 driver and ACL definitions.
707 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
708 forgetting to use the resulting value; it was using the unexpanded value.
710 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
711 hadn't been configured. The fix is from Juergen Kreileder, who
712 understands it better than I do:
714 "Here's what I see happening with three configured cyrus_sasl
715 authenticators configured (plain, login, cram-md5):
717 On startup auth_cyrus_sasl_init() gets called for each of these.
718 This means three calls to sasl_listmech() without a specified mech_list.
719 => SASL tests which mechs of all available mechs actually work
720 => three warnings about OTP not working
721 => the returned list contains: plain, login, cram-md5, digest-md5, ...
723 With the patch, sasl_listmech() also gets called three times. But now
724 SASL's mech_list option is set to the server_mech specified in the the
725 authenticator. Or in other words, the answer from sasl_listmech()
726 gets limited to just the mech you're testing for (which is different
728 => the return list contains just 'plain' or 'login', 'cram-md5' or
729 nothing depending on the value of ob->server_mech.
731 I've just tested the patch: Authentication still works fine,
732 unavailable mechs specified in the exim configuration are still
733 caught, and the auth.log warnings about OTP are gone."
735 PH/25 When debugging is enabled, the contents of the command line are added
736 to the debugging output, even when log_selector=+arguments is not
739 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
740 answer is "GNU", and only if the return is "GNU/something" is the answer
743 PH/27 $acl_verify_message is now set immediately after the failure of a
744 verification in an ACL, and so is available in subsequent modifiers. In
745 particular, the message can be preserved by coding like this:
747 warn !verify = sender
748 set acl_m0 = $acl_verify_message
750 Previously, $acl_verify_message was set only while expanding "message"
751 and "log_message" when a very denied access.
753 PH/28 Modified OS/os.c-Linux with
755 -#ifndef OS_LOAD_AVERAGE
756 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
758 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
759 nomenclature these days.)
761 PH/29 Installed patch from the Sieve maintainer that adds the options
762 sieve_useraddress and sieve_subaddress to the redirect router.
764 PH/30 In these circumstances:
765 . Two addresses routed to the same list of hosts;
766 . First host does not offer TLS;
767 . First host accepts first address;
768 . First host gives temporary error to second address;
769 . Second host offers TLS and a TLS session is established;
770 . Second host accepts second address.
771 Exim incorrectly logged both deliveries with the TLS parameters (cipher
772 and peerdn, if requested) that were in fact used only for the second
775 PH/31 When doing a callout as part of verifying an address, Exim was not paying
776 attention to any local part prefix or suffix that was matched by the
777 router that accepted the address. It now behaves in the same way as it
778 does for delivery: the affixes are removed from the local part unless
779 rcpt_include_affixes is set on the transport.
781 PH/32 Add the sender address, as F=<...>, to the log line when logging a
782 timeout during the DATA phase of an incoming message.
784 PH/33 Sieve envelope tests were broken for match types other than :is. I have
785 applied a patch sanctioned by the Sieve maintainer.
787 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
788 the uid or gid is negative. A case of a negative gid caused this to be
789 noticed. The fix allows for either to be negative.
791 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
792 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
795 PH/36 The change PH/12 above was broken. Fixed it.
797 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
798 the grounds that routing the same address twice would always produce the
799 same answer. This might have been true once, but it is certainly no
800 longer true now. Routing a child address may depend on the previous
801 routing that produced that child. Some complicated redirection strategies
802 went wrong when messages had multiple recipients, and made Exim's
803 behaviour dependent on the order in which the addresses were given.
805 I have moved the duplicate checking until after the routing is complete.
806 Exim scans the addresses that are assigned to local and remote
807 transports, and removes any duplicates. This means that more work will be
808 done, as duplicates will always all be routed, but duplicates are
809 presumably rare, so I don't expect this is of any significance.
811 For deliveries to pipes, files, and autoreplies, the duplicate checking
812 still happens during the routing process, since they are not going to be
815 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
816 It corrects a timeout issue with spamd. This is Ian's comment: "The
817 background is that sometimes spamd either never reads data from a
818 connection it has accepted, or it never writes response data. The exiscan
819 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
820 blindly assumes that writes won't block so it may never time out."
822 PH/39 Allow G after quota size as well as K and M.
824 PH/40 The value set for $authenticated_id in an authenticator may not contain
825 binary zeroes or newlines because the value is written to log lines and
826 to spool files. There was no check on this. Now the value is run through
827 the string_printing() function so that such characters are converted to
828 printable escape sequences.
830 PH/41 $message_linecount is a new variable that contains the total number of
831 lines in the message. Compare $body_linecount, which is the count for the
834 PH/42 Exim no longer gives details of delivery errors for specific addresses in
835 bounce and delay warning messages, except in certain special cases, which
838 (a) An SMTP error message from a remote host;
839 (b) A message specified in a :fail: redirection;
840 (c) A message specified in a "fail" command in a system filter;
841 (d) A message specified in a FAIL return from the queryprogram router;
842 (e) A message specified by the cannot_route_message router option.
844 In these cases only, Exim does include the error details in bounce and
845 warning messages. There are also a few cases where bland messages such
846 as "unrouteable address" or "local delivery error" are given.
848 PH/43 $value is now also set for the "else" part of a ${run expansion.
850 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
851 being worked on, but at least Exim now implements the latest version to
854 PH/45 In a pipe transport, although a timeout while waiting for the pipe
855 process to complete was treated as a delivery failure, a timeout while
856 writing the message to the pipe was logged, but erroneously treated as a
857 successful delivery. Such timeouts include transport filter timeouts. For
858 consistency with the overall process timeout, these timeouts are now
859 treated as errors, giving rise to delivery failures by default. However,
860 there is now a new Boolean option for the pipe transport called
861 timeout_defer, which, if set TRUE, converts the failures into defers for
862 both kinds of timeout. A transport filter timeout is now identified in
865 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
866 systems where "make" and "gmake" are different, calling "gmake" at top
867 level broke things. I've arranged for the value of $(MAKE) to be passed
868 from the Makefile to this script so that it can call the same version of
872 A note about Exim versions 4.44 and 4.50
873 ----------------------------------------
875 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
876 changes of various kinds. As a consequence, a big documentation update was
877 needed. This delayed the release for rather longer than seemed good, especially
878 in the light of a couple of (minor) security issues. Therefore, the changes
879 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
880 release. So 4.44 and 4.50 are in effect two different branches that both start
883 I have left the 4.50 change log unchanged; it contains all the changes since
884 4.43. The change log for 4.44 is below; many of its items are identical to
885 those for 4.50. This seems to be the most sensible way to preserve the
886 historical information.
892 1. Minor wording change to the doc/README.SIEVE file.
894 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
895 computation of the current number of files was incorrect.
897 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
898 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
901 4. Give more explanation in the error message when the command for a transport
902 filter fails to execute.
904 5. There are several places where Exim runs a non-Exim command in a
905 subprocess. The SIGUSR1 signal should be disabled for these processes. This
906 was being done only for the command run by the queryprogram router. It is
907 now done for all such subprocesses. The other cases are: ${run, transport
908 filters, and the commands run by the lmtp and pipe transports.
910 6. Added CONFIGURE_GROUP build-time option.
912 7. Some older OS have a limit of 256 on the maximum number of file
913 descriptors. Exim was using setrlimit() to set 1000 as a large value
914 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
915 systems. I've change it so that if it can't get 1000, it tries for 256.
917 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
918 was an oversight, and furthermore, ever since the addition of extra
919 controls (e.g. 4.43/32), the checks on when to allow different forms of
920 "control" were broken. There should now be diagnostics for all cases when a
921 control that does not make sense is encountered.
923 9. Added the /retain_sender option to "control=submission".
925 10. $recipients is now available in the predata ACL (oversight).
927 11. Tidy the search cache before the fork to do a delivery from a message
928 received from the command line. Otherwise the child will trigger a lookup
929 failure and thereby defer the delivery if it tries to use (for example) a
930 cached ldap connection that the parent has called unbind on.
932 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
933 of $address_data from the recipient verification was clobbered by the
936 13. The value of address_data from a sender verification is now available in
937 $sender_address_data in subsequent conditions in the ACL statement.
939 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
941 15. Added a new option "connect=<time>" to callout options, to set a different
944 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
945 was its contents. (It was OK if the option was not defined at all.)
947 17. A "Completed" log line is now written for messages that are removed from
948 the spool by the -Mrm option.
950 18. New variables $sender_verify_failure and $recipient_verify_failure contain
951 information about exactly what failed.
953 19. Added -dd to debug only the daemon process.
955 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
956 handles timeouts, both on the server side and network timeouts. Renamed the
957 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
959 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
960 It is now set to "smtps".
962 22. $host_address is now set to the target address during the checking of
965 23. When checking ignore_target_hosts for an ipliteral router, no host name was
966 being passed; this would have caused $sender_host_name to have been used if
967 matching the list had actually called for a host name (not very likely,
968 since this list is usually IP addresses). A host name is now passed as
971 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
972 code that specifies a non-restarting handler (typically sigaction() in
973 modern systems) in an attempt to fix a rare and obscure crash bug.
975 25. Narrowed the window for a race in the daemon that could cause it to ignore
976 SIGCHLD signals. This is not a major problem, because they are used only to
977 wake it up if nothing else does.
979 26. A malformed maildirsize file could cause Exim to calculate negative values
980 for the mailbox size or file count. Odd effects could occur as a result.
981 The maildirsize information is now recalculated if the size or filecount
984 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
985 support for a long time. Removed HAVE_SYS_VFS_H.
987 28. Installed the latest version of exipick from John Jetmore.
989 29. In an address list, if the pattern was not a regular expression, an empty
990 subject address (from a bounce message) matched only if the pattern was an
991 empty string. Non-empty patterns were not even tested. This was the wrong
992 because it is perfectly reasonable to use an empty address as part of a
993 database query. An empty address is now tested by patterns that are
994 lookups. However, all the other forms of pattern expect the subject to
995 contain a local part and a domain, and therefore, for them, an empty
996 address still always fails if the pattern is not itself empty.
998 30. Exim went into a mad DNS loop when attempting to do a callout where the
999 host was specified on an smtp transport, and looking it up yielded more
1000 than one IP address.
1002 31. Re-factored the code for checking spool and log partition space into a
1003 function that finds that data and another that does the check. The former
1004 is then used to implement four new variables: $spool_space, $log_space,
1005 $spool_inodes, and $log_inodes.
1007 32. The RFC2047 encoding function was originally intended for short strings
1008 such as real names; it was not keeping to the 75-character limit for
1009 encoded words that the RFC imposes. It now respects the limit, and
1010 generates multiple encoded words if necessary. To be on the safe side, I
1011 have increased the buffer size for the ${rfc2047: expansion operator from
1014 33. It is now permitted to omit both strings after an "if" condition; if the
1015 condition is true, the result is "true". As before, when the second string
1016 is omitted, a false condition yields an empty string. This makes it less
1017 cumbersome to write custom ACL and router conditions.
1019 34. Failure to deliver a bounce message always caused it to be frozen, even if
1020 there was an errors_to setting on the router. The errors_to setting is now
1023 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1024 canonical form (fully expanded) before being placed in
1025 $sender_host_address.
1027 36. The table in the code that translates DNS record types into text (T_A to
1028 "A" for instance) was missing entries for NS and CNAME. It is just possible
1029 that this could have caused confusion if both these types were looked up
1030 for the same domain, because the text type is used as part of Exim's
1031 per-process caching. But the chance of anyone hitting this buglet seems
1034 37. The dnsdb lookup has been extended in a number of ways.
1036 (1) There is a new type, "zns", which walks up the domain tree until it
1037 finds some nameserver records. It should be used with care.
1039 (2) There is a new type, "mxh", which is like "mx" except that it returns
1040 just the host names, not the priorities.
1042 (3) It is now possible to give a list of domains (or IP addresses) to be
1043 looked up. The behaviour when one of the lookups defers can be
1044 controlled by a keyword.
1046 (4) It is now possible to specify the separator character for use when
1047 multiple records are returned.
1049 38. The dnslists ACL condition has been extended: it is now possible to supply
1050 a list of IP addresses and/or domains to be looked up in a particular DNS
1053 39. Added log_selector=+queue_time_overall.
1055 40. When running the queue in the test harness, wait just a tad after forking a
1056 delivery process, to get repeatability of debugging output.
1058 41. Include certificate and key file names in error message when GnuTLS fails
1059 to set them up, because the GnuTLS error message doesn't include the name
1060 of the failing file when there is a problem reading it.
1062 42. Allow both -bf and -bF in the same test run.
1064 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1066 44. The "Exiscan patch" is now merged into the mainline Exim source.
1068 45. Sometimes the final signoff response after QUIT could fail to get
1069 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1070 < 0 before doing a fflush(). This bug looks as though it goes back to the
1071 introduction of TLS in release 3.20, but "sometimes" must have been rare
1072 because the tests only now provoked it.
1074 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1075 (this can affect the format of dates).
1077 47. exim_tidydb, when checking for the continued existence of a message for
1078 which it has found a message-specific retry record, was not finding
1079 messages that were in split spool directories. Consequently, it was
1080 deleting retry records that should have stayed in existence.
1082 48. Steve fixed some bugs in eximstats.
1084 49. The SPA authentication driver was not abandoning authentication and moving
1085 on to the next authenticator when an expansion was forced to fail,
1086 contradicting the general specification for all authenticators. Instead it
1087 was generating a temporary error. It now behaves as specified.
1089 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1090 (the order specifies the preference for clients). The order is now AES256,
1091 AES128, 3DES, ARCFOUR128.
1093 51. Small patch to Sieve code - explicitly set From: when generating an
1096 52. Exim crashed if a remote delivery caused a very long error message to be
1097 recorded - for instance if somebody sent an entire SpamAssassin report back
1098 as a large number of 550 error lines. This bug was coincidentally fixed by
1099 increasing the size of one of Exim's internal buffers (big_buffer) that
1100 happened as part of the Exiscan merge. However, to be on the safe side, I
1101 have made the code more robust (and fixed the comments that describe what
1104 53. Now that there can be additional text after "Completed" in log lines (if
1105 the queue_time_overall log selector is set), a one-byte patch to exigrep
1106 was needed to allow it to recognize "Completed" as not the last thing in
1109 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1110 patch that reportedly fixes this has been added. I am not expert enough to
1111 create a test for it. This is what the patch creator wrote:
1113 "I found a little strange behaviour of ldap code when working with
1114 Windows 2003 AD Domain, where users was placed in more than one
1115 Organization Units. When I tried to give exim partial DN, the exit code
1116 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1117 But simultaneously result of request was absolutely normal ldap result,
1118 so I produce this patch..."
1120 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1121 REFERENCE, so I have modified the code to exclude the patch when that macro
1124 55. Some experimental protocols are using DNS PTR records for new purposes. The
1125 keys for these records are domain names, not reversed IP addresses. The
1126 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1127 leaves it alone. Component reversal etc. now happens only for IP addresses.
1130 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1132 57. Double the size of the debug message buffer (to 2048) so that more of very
1133 long debug lines gets shown.
1135 58. The exicyclog utility now does better if the number of log files to keep
1136 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1138 59. Two changes related to the smtp_active_hostname option:
1140 (1) $smtp_active_hostname is now available as a variable.
1141 (2) The default for smtp_banner uses $smtp_active_hostname instead
1142 of $primary_hostname.
1144 60. The host_aton() function is supposed to be passed a string that is known
1145 to be a valid IP address. However, in the case of IPv6 addresses, it was
1146 not checking this. This is a hostage to fortune. Exim now panics and dies
1147 if the condition is not met. A case was found where this could be provoked
1148 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1149 components; fortuitously, this particular loophole had already been fixed
1150 by change 4.50/55 above.
1152 If there are any other similar loopholes, the new check in host_aton()
1153 itself should stop them being exploited. The report I received stated that
1154 data on the command line could provoke the exploit when Exim was running as
1155 exim, but did not say which command line option was involved. All I could
1156 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1157 running as the user.
1160 61. There was a buffer overflow vulnerability in the SPA authentication code
1161 (which came originally from the Samba project). I have added a test to the
1162 spa_base64_to_bits() function which I hope fixes it.
1165 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1166 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1168 63. The daemon start-up calls getloadavg() while still root for those OS that
1169 need the first call to be done as root, but it missed one case: when
1170 deliver_queue_load_max is set with deliver_drop_privilege. This is
1171 necessary for the benefit of the queue runner, because there is no re-exec
1172 when deliver_drop_privilege is set.
1174 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1175 This has been fixed.
1177 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1178 was in use, was not putting the data itself into the right store pool;
1179 consequently, it could be overwritten for a subsequent message in the same
1180 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1183 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1185 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1186 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1187 the function were treating the return as a boolean value, which happened to
1188 work because 0=false and not-0=true, but is not correct code.
1190 68. The host_aton() function was not handling scoped IPv6 addresses (those
1191 with, for example, "%eth0" on the end) correctly.
1193 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1194 negated items (that is, ~something) in unsigned ints. Some compilers
1195 apparently mutter when there is no cast.
1197 70. If an address verification called from an ACL failed, and did not produce a
1198 user-specific message (i.e. there was only a "system" message), nothing was
1199 put in $acl_verify_message. In this situation, it now puts the system
1202 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1203 session; change 4.31/43 added the unwanted input to the log line - except
1204 that it did not do this in the start of session case. It now does.
1206 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1207 This is wrong and can cause the other end to generate a synchronization
1208 error if it is another Exim or anything else that does the synchronization
1209 check. A QUIT command is no longer sent after a timeout.
1211 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1212 during host lookups.
1214 74. The defer_ok option of callout verification was not working if it was used
1215 when verifying addresses in header lines, that is, for this case:
1217 verify = header_sender/callout=defer_ok
1219 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1220 those file descriptors could be used for SMTP connections. If anything
1221 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1222 could be sent to the SMTP client, causing chaos. The daemon now opens
1223 stdin, stdout, and stderr to /dev/null when it puts itself into the
1226 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1227 log by default. The user can override this with suitable Perl magic.
1229 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1230 the log message when discard triggers, was not working for the DATA ACL or
1231 for the non-SMTP ACL.
1233 78. Error message wording change in sieve.c.
1235 79. If smtp_accept_max_per_host was set, the number of connections could be
1236 restricted to fewer than expected, because the daemon was trying to set up
1237 a new connection before checking whether the processes handling previous
1238 connections had finished. The check for completed processes is now done
1239 earlier. On busy systems, this bug wouldn't be noticed because something
1240 else would have woken the daemon, and it would have reaped the completed
1243 80. If a message was submitted locally by a user whose login name contained one
1244 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1245 It caused a spool format error. I have fixed the spool reading code. A
1246 related problem was that the "from" clause in the Received: line became
1247 illegal because of the space(s). It is now covered by ${quote_local_part.
1249 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1252 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1254 83. Patch from Sieve maintainer to fix unterminated string problem in
1255 "vacation" handling.
1257 84. Some minor changes to the Linux configuration files to help with other
1258 OS variants using glibc.
1260 85. One more patch for Sieve to update vacation handling to latest spec.
1263 ----------------------------------------------------
1264 See the note above about the 4.44 and 4.50 releases.
1265 ----------------------------------------------------
1271 1. Change 4.43/35 introduced a bug that caused file counts to be
1272 incorrectly computed when quota_filecount was set in an appendfile
1275 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1276 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1279 3. Give more explanation in the error message when the command for a transport
1280 filter fails to execute.
1282 4. There are several places where Exim runs a non-Exim command in a
1283 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1284 was being done only for the command run by the queryprogram router. It is
1285 now done for all such subprocesses. The other cases are: ${run, transport
1286 filters, and the commands run by the lmtp and pipe transports.
1288 5. Some older OS have a limit of 256 on the maximum number of file
1289 descriptors. Exim was using setrlimit() to set 1000 as a large value
1290 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1291 systems. I've change it so that if it can't get 1000, it tries for 256.
1293 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1294 was an oversight, and furthermore, ever since the addition of extra
1295 controls (e.g. 4.43/32), the checks on when to allow different forms of
1296 "control" were broken. There should now be diagnostics for all cases when a
1297 control that does not make sense is encountered.
1299 7. $recipients is now available in the predata ACL (oversight).
1301 8. Tidy the search cache before the fork to do a delivery from a message
1302 received from the command line. Otherwise the child will trigger a lookup
1303 failure and thereby defer the delivery if it tries to use (for example) a
1304 cached ldap connection that the parent has called unbind on.
1306 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1307 of $address_data from the recipient verification was clobbered by the
1308 sender verification.
1310 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1311 was its contents. (It was OK if the option was not defined at all.)
1313 11. A "Completed" log line is now written for messages that are removed from
1314 the spool by the -Mrm option.
1316 12. $host_address is now set to the target address during the checking of
1317 ignore_target_hosts.
1319 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1320 being passed; this would have caused $sender_host_name to have been used if
1321 matching the list had actually called for a host name (not very likely,
1322 since this list is usually IP addresses). A host name is now passed as
1325 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1326 code that specifies a non-restarting handler (typically sigaction() in
1327 modern systems) in an attempt to fix a rare and obscure crash bug.
1329 15. Narrowed the window for a race in the daemon that could cause it to ignore
1330 SIGCHLD signals. This is not a major problem, because they are used only to
1331 wake it up if nothing else does.
1333 16. A malformed maildirsize file could cause Exim to calculate negative values
1334 for the mailbox size or file count. Odd effects could occur as a result.
1335 The maildirsize information is now recalculated if the size or filecount
1338 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1339 support for a long time. Removed HAVE_SYS_VFS_H.
1341 18. Updated exipick to current release from John Jetmore.
1343 19. Allow an empty sender to be matched against a lookup in an address list.
1344 Previously the only cases considered were a regular expression, or an
1347 20. Exim went into a mad DNS lookup loop when doing a callout where the
1348 host was specified on the transport, if the DNS lookup yielded more than
1351 21. The RFC2047 encoding function was originally intended for short strings
1352 such as real names; it was not keeping to the 75-character limit for
1353 encoded words that the RFC imposes. It now respects the limit, and
1354 generates multiple encoded words if necessary. To be on the safe side, I
1355 have increased the buffer size for the ${rfc2047: expansion operator from
1358 22. Failure to deliver a bounce message always caused it to be frozen, even if
1359 there was an errors_to setting on the router. The errors_to setting is now
1362 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1363 canonical form (fully expanded) before being placed in
1364 $sender_host_address.
1366 24. Updated eximstats to version 1.33
1368 25. Include certificate and key file names in error message when GnuTLS fails
1369 to set them up, because the GnuTLS error message doesn't include the name
1370 of the failing file when there is a problem reading it.
1372 26. Expand error message when OpenSSL has problems setting up cert/key files.
1375 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1376 (this can affect the format of dates).
1378 28. exim_tidydb, when checking for the continued existence of a message for
1379 which it has found a message-specific retry record, was not finding
1380 messages that were in split spool directories. Consequently, it was
1381 deleting retry records that should have stayed in existence.
1383 29. eximstats updated to version 1.35
1384 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1385 1.35 - bugfix such that pie charts by volume are generated correctly
1387 30. The SPA authentication driver was not abandoning authentication and moving
1388 on to the next authenticator when an expansion was forced to fail,
1389 contradicting the general specification for all authenticators. Instead it
1390 was generating a temporary error. It now behaves as specified.
1392 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1393 (the order specifies the preference for clients). The order is now AES256,
1394 AES128, 3DES, ARCFOUR128.
1396 31. Small patch to Sieve code - explicitly set From: when generating an
1399 32. Exim crashed if a remote delivery caused a very long error message to be
1400 recorded - for instance if somebody sent an entire SpamAssassin report back
1401 as a large number of 550 error lines. This bug was coincidentally fixed by
1402 increasing the size of one of Exim's internal buffers (big_buffer) that
1403 happened as part of the Exiscan merge. However, to be on the safe side, I
1404 have made the code more robust (and fixed the comments that describe what
1407 33. Some experimental protocols are using DNS PTR records for new purposes. The
1408 keys for these records are domain names, not reversed IP addresses. The
1409 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1410 leaves it alone. Component reversal etc. now happens only for IP addresses.
1413 34. The host_aton() function is supposed to be passed a string that is known
1414 to be a valid IP address. However, in the case of IPv6 addresses, it was
1415 not checking this. This is a hostage to fortune. Exim now panics and dies
1416 if the condition is not met. A case was found where this could be provoked
1417 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1418 components; fortuitously, this particular loophole had already been fixed
1419 by change 4.50/55 or 4.44/33 above.
1421 If there are any other similar loopholes, the new check in host_aton()
1422 itself should stop them being exploited. The report I received stated that
1423 data on the command line could provoke the exploit when Exim was running as
1424 exim, but did not say which command line option was involved. All I could
1425 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1426 running as the user.
1429 35. There was a buffer overflow vulnerability in the SPA authentication code
1430 (which came originally from the Samba project). I have added a test to the
1431 spa_base64_to_bits() function which I hope fixes it.
1434 36. The daemon start-up calls getloadavg() while still root for those OS that
1435 need the first call to be done as root, but it missed one case: when
1436 deliver_queue_load_max is set with deliver_drop_privilege. This is
1437 necessary for the benefit of the queue runner, because there is no re-exec
1438 when deliver_drop_privilege is set.
1440 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1441 was in use, was not putting the data itself into the right store pool;
1442 consequently, it could be overwritten for a subsequent message in the same
1443 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1446 38. Sometimes the final signoff response after QUIT could fail to get
1447 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1448 < 0 before doing a fflush(). This bug looks as though it goes back to the
1449 introduction of TLS in release 3.20, but "sometimes" must have been rare
1450 because the tests only now provoked it.
1456 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1457 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1458 Now it returns an integer. A number of calls were still expecting a T/F
1459 return. Fortuitously, in all cases, the tests worked in OK situations,
1460 which is the norm. However, things would have gone wrong on any write
1461 failures on the smtp file descriptor. This function is used when sending
1462 messages over SMTP and also when doing verify callouts.
1464 2. When Exim is called to do synchronous delivery of a locally submitted
1465 message (the -odf or -odi options), it no longer closes stderr before doing
1468 3. Implemented the mua_wrapper option.
1470 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1472 5. Implemented the functions header_remove(), header_testname(),
1473 header_add_at_position(), and receive_remove_recipient(), and exported them
1476 6. If an ACL "warn" statement specified the addition of headers, Exim already
1477 inserted X-ACL-Warn: at the start if there was no header name. However, it
1478 was not making this test for the second and subsequent header lines if
1479 there were newlines in the string. This meant that an invalid header could
1480 be inserted if Exim was badly configured.
1482 7. Allow an ACL "warn" statement to add header lines at the start or after all
1483 the Received: headers, as well as at the end.
1485 8. Added the rcpt_4xx retry error code.
1487 9. Added postmaster_mailfrom=xxx to callout verification option.
1489 10. Added mailfrom=xxxx to the callout verification option, for verify=
1492 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1493 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1495 12. Inserted some casts to stop certain compilers warning when using pointer
1496 differences as field lengths or precisions in printf-type calls (mostly
1497 affecting debugging statements).
1499 13. Added optional readline() support for -be (dynamically loaded).
1501 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1502 same clock tick as a message's arrival, so that its received time was the
1503 same as the "first fail" time on the retry record, and that message
1504 remained on the queue past the ultimate address timeout, every queue runner
1505 would try a delivery (because it was past the ultimate address timeout) but
1506 after another failure, the ultimate address timeout, which should have then
1507 bounced the address, did not kick in. This was a "< instead of <=" error;
1508 in most cases the first failure would have been in the next clock tick
1509 after the received time, and all would be well.
1511 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1512 being recognized when the domain list was tested by the match_domain
1513 condition in an expansion string.
1515 16. Added the ${str2b64: operator.
1517 17. Exim was always calling setrlimit() to set a large limit for the number of
1518 processes, without checking whether the existing limit was already
1519 adequate. (It did check for the limit on file descriptors.) Furthermore,
1520 errors from getrlimit() and setrlimit() were being ignored. Now they are
1521 logged to the main and panic logs, but Exim does carry on, to try to do its
1522 job under whatever limits there are.
1524 18. Imported PCRE 5.0.
1526 19. Trivial typo in log message " temporarily refused connection" (the leading
1529 20. If the log selector return_path_on_delivery was set and an address was
1530 redirected to /dev/null, the delivery process crashed because it assumed
1531 that a return path would always be set for a "successful" delivery. In this
1532 case, the whole delivery is bypassed as an optimization, and therefore no
1535 21. Internal re-arrangement: the function for sending a challenge and reading
1536 a response while authentication was assuming a zero-terminated challenge
1537 string. It's now changed to take a pointer and a length, to allow for
1538 binary data in such strings.
1540 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1542 23. Exim was not respecting finduser_retries when seeking the login of the
1543 uid under which it was called; it was always trying 10 times. (The default
1544 setting of finduser_retries is zero.) Also, it was sleeping after the final
1545 failure, which is pointless.
1547 24. Implemented tls_on_connect_ports.
1549 25. Implemented acl_smtp_predata.
1551 26. If the domain in control=submission is set empty, Exim assumes that the
1552 authenticated id is a complete email address when it generates From: or
1553 Sender: header lines.
1555 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1556 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1557 chown and chgrp in /bin and hostname in /usr/bin.
1559 28. Exim was keeping the "process log" file open after each use, just as it
1560 does for the main log. This opens the possibility of it remaining open for
1561 long periods when the USR1 signal hits a daemon. Occasional processlog
1562 errors were reported, that could have been caused by this. Anyway, it seems
1563 much more sensible not to leave this file open at all, so that is what now
1566 29. The long-running daemon process does not normally write to the log once it
1567 has entered its main loop, and it closes the log before doing so. This is
1568 so that log files can straightforwardly be renamed and moved. However,
1569 there are a couple of unusual error situations where the daemon does write
1570 log entries, and I had neglected to close the log afterwards.
1572 30. The text of an SMTP error response that was received during a remote
1573 delivery was being truncated at 512 bytes. This is too short for some of
1574 the long messages that one sometimes sees. I've increased the limit to
1577 31. It is now possible to make retry rules that apply only when a message has a
1578 specific sender, in particular, an empty sender.
1580 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1581 it possible to be selective about when SMTP synchronization is enforced.
1583 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1585 32. Implemented hosts_connection_nolog.
1587 33. Added an ACL for QUIT.
1589 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1592 35. Added mailbox_size and mailbox_filecount to appendfile.
1594 36. Added control = no_multiline_responses to ACLs.
1596 37. There was a bug in the logic of the code that waits for the clock to tick
1597 in the case where the clock went backwards by a substantial amount such
1598 that the microsecond fraction of "now" was more than the microsecond
1599 fraction of "then" (but the whole seconds number was less).
1601 38. Added support for the libradius Radius client library this is found on
1602 FreeBSD (previously only the radiusclient library was supported).
1608 1. When certain lookups returned multiple values in the form name=value, the
1609 quoting of the values was not always being done properly. Specifically:
1610 (a) If the value started with a double quote, but contained no whitespace,
1612 (b) If the value contained whitespace other than a space character (i.e.
1613 tabs or newlines or carriage returns) it was not quoted.
1614 This fix has been applied to the mysql and pgsql lookups by writing a
1615 separate quoting function and calling it from the lookup code. The fix
1616 should probably also be applied to nisplus, ibase and oracle lookups, but
1617 since I cannot test any of those, I have not disturbed their existing code.
1619 2. A hit in the callout cache for a specific address caused a log line with no
1620 reason for rejecting RCPT. Now it says "Previous (cached) callout
1621 verification failure".
1623 3. There was an off-by-one bug in the queryprogram router. An over-long
1624 return line was truncated at 256 instead of 255 characters, thereby
1625 overflowing its buffer with the terminating zero. As well as fixing this, I
1626 have increased the buffer size to 1024 (and made a note to document this).
1628 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1629 when Exim is waiting for an SMTP response from a remote server, Exim
1630 restarts its select() call on the socket, thereby resetting its timeout.
1631 This is not a problem when such interrupts are rare. Somebody set up a cron
1632 job to run exiwhat every 2 minutes, which is less than the normal select()
1633 timeout (5 or 10 minutes). This meant that the select() timeout never
1634 kicked in because it was always reset. I have fixed this by comparing the
1635 time when an interrupt arrives with the time at the start of the first call
1636 to select(). If more time than the timeout has elapsed, the interrupt is
1637 treated as a timeout.
1639 5. Some internal re-factoring in preparation for the addition of Sieve
1640 extensions (by MH). In particular, the "personal" test is moved to a
1641 separate function, and given an option for scanning Cc: and Bcc: (which is
1642 not set for Exim filters).
1644 6. When Exim created an email address using the login of the caller as the
1645 local part (e.g. when creating a From: or Sender: header line), it was not
1646 quoting the local part when it contained special characters such as @.
1648 7. Installed new OpenBSD configuration files.
1650 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1651 try to make them clearer.
1653 9. Callout options, other than the timeout value, were being ignored when
1654 verifying sender addresses in header lines. For example, when using
1656 verify = header_sender/callout=no_cache
1658 the cache was (incorrectly) being used.
1660 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1661 only the Cygwin environment.
1663 11. When return_path_on_delivery was set as a log selector, if different remote
1664 addresses in the same message used different return paths and parallel
1665 remote delivery occurred, the wrong values would sometimes be logged.
1666 (Whenever a remote delivery process finished, the return path value from
1667 the most recently started remote delivery process was logged.)
1669 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1670 lines when AUTH and/or TLS are in use. This is the "received protocol"
1671 field. Exim used to use "asmtp" for authenticated SMTP, without any
1672 indication (in the protocol name) for TLS use. Now it follows the RFC and
1673 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1674 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1675 names appear in log lines as well as in Received: header lines.
1677 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1678 extensions, and comparison tests, and to fix some bugs.
1680 14. Changes to the "personal" filter test:
1682 (1) The test was buggy in that it was just doing the equivalent of
1683 "contains" tests on header lines. For example, if a user's address was
1684 anne@some.where, the "personal" test would incorrectly be true for
1686 To: susanne@some.where
1688 This test is now done by extracting each address from the header in turn,
1689 and checking the entire address. Other tests that are part of "personal"
1690 are now done using regular expressions (for example, to check local parts
1691 of addresses in From: header lines).
1693 (2) The list of non-personal local parts in From: addresses has been
1694 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1695 taken from the Sieve specification recommendations.
1697 (3) If the message contains any header line starting with "List-" it is
1698 treated as non-personal.
1700 (4) The test for "circular" in the Subject: header line has been removed
1701 because it now seems ill-conceived.
1703 15. Minor typos in src/EDITME comments corrected.
1705 16. Installed latest exipick from John Jetmore.
1707 17. If headers_add on a router specified a text string that was too long for
1708 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1709 of string_sprintf() is now avoided.
1711 18. $message_body_size was not set (it was always zero) when running the DATA
1712 ACL and the local_scan() function.
1714 19. For the "mail" command in an Exim filter, no default was being set for
1715 the once_repeat time, causing a random time value to be used if "once" was
1716 specified. (If the value happened to be <= 0, no repeat happened.) The
1717 default is now 0s, meaning "never repeat". The "vacation" command was OK
1718 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1719 (I found it when inspecting the code).
1721 20. There is now an overall timeout for performing a callout verification. It
1722 defaults to 4 times the callout timeout, which applies to individual SMTP
1723 commands during the callout. The overall timeout applies when there is more
1724 than one host that can be tried. The timeout is checked before trying the
1725 next host. This prevents very long delays if there are a large number of
1726 hosts and all are timing out (e.g. when the network connections are timing
1727 out). The value of the overall timeout can be changed by specifying an
1728 additional sub-option for "callout", called "maxwait". For example:
1730 verify = sender/callout=5s,maxwait=20s
1732 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1733 to the end before writing, but this should make it even safer).
1735 22. Exim was forgetting that it had advertised PIPELINING for the second and
1736 subsequent messages on an SMTP connection. It was also not resetting its
1737 memory on STARTTLS and an internal HELO.
1739 23. When Exim logs an SMTP synchronization error within a session, it now
1740 records whether PIPELINING has been advertised or not.
1742 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1743 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1744 rather than long int.
1746 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1748 26. Added the never_mail option to autoreply.
1754 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1755 crash if the getsockname() call failed; this can happen if a connection is
1756 closed very soon after it is established. The problem was simply in the
1757 order in which certain operations were done, causing Exim to try to write
1758 to the SMTP stream before it had set up the file descriptor. The bug has
1759 been fixed by making things happen in the correct order.
1765 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1766 before the connection was closed, thus losing the rejection response.
1768 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1769 some early Solaris releases, but causes trouble in current releases where
1770 socklen_t is defined.
1772 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1775 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1778 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1779 that releasing the top of it at the end releases what was used for sub-
1780 expansions (unless the block got too big). However, discard this block if
1781 the first thing is a variable or header, so that we can use its block when
1782 it is dynamic (useful for very large $message_headers, for example).
1784 6. Lookups now cache *every* query, not just the most recent. A new, separate
1785 store pool is used for this. It can be recovered when all lookup caches are
1786 flushed. Lookups now release memory at the end of their result strings.
1787 This has involved some general refactoring of the lookup sources.
1789 7. Some code has been added to the store_xxx() functions to reduce the amount
1790 of flapping under certain conditions.
1792 8. log_incoming_interface used to affect only the <= reception log lines. Now
1793 it causes the local interface and port to be added to several more SMTP log
1794 lines, for example "SMTP connection from", and rejection lines.
1796 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1798 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1800 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1801 could be overwritten at the end of the current message (or the start of a
1802 new message if it was set in a HELO ACL). The value is now preserved for
1803 the duration of the SMTP connection.
1805 12. If a transport had a headers_rewrite setting, and a matching header line
1806 contained an unqualified address, that address was qualified, even if it
1807 did not match any rewriting rules. The underlying bug was that the values
1808 of the flags that permit the existence of unqualified sender and recipient
1809 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1810 non-local messages, and by -bnq for local messages) were not being
1811 preserved with the message after it was received.
1813 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1814 "next input=" as part of the text comprising the host identity instead of
1815 the correct text. The code was using the same buffer for two different
1816 strings. However, depending on which order the printing function evaluated
1817 its arguments, the bug did not always show up. Under Linux, for example, my
1818 test suite worked just fine.
1820 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1821 doesn't work with some older versions of Perl. It has been changed to "my",
1822 which in any case is probably the better facility to use.
1824 15. A really picky compiler found some instances of statements for creating
1825 error messages that either had too many or two few arguments for the format
1828 16. The size of the buffer for calls to the DNS resolver has been increased
1829 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1830 for addresses that have a lot of PTR records. This alleviates a problem; it
1831 does not fully solve it.
1833 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1834 buffer now truncates the list and logs the incident, which is the same
1835 action as happens when Exim is looking up a host name and its aliases.
1836 Previously in this situation something unpredictable would happen;
1837 sometimes it was "internal error: store_reset failed".
1839 18. If a server dropped the connection unexpectedly when an Exim client was
1840 using GnuTLS and trying to read a response, the client delivery process
1841 crashed while trying to generate an error log message.
1843 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1844 string, for example:
1846 warn message = H1: something\nH2: something
1848 the text was added as a single header line from Exim's point of view
1849 though it ended up OK in the delivered message. However, searching for the
1850 second and subsequent header lines using $h_h2: did not work. This has been
1851 fixed. Similarly, if a system filter added multiple headers in this way,
1852 the routers could not see them.
1854 20. Expanded the error message when iplsearch is called with an invalid key to
1855 suggest using net-iplsearch in a host list.
1857 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1858 ACLs are no longer actually imposed (and a message to that effect is
1861 22. If a "gecos" field in a passwd entry contained escaped characters, in
1862 particular, if it contained a \" sequence, Exim got it wrong when building
1863 a From: or a Sender: header from that name. A second bug also caused
1864 incorrect handling when an unquoted " was present following a character
1865 that needed quoting.
1867 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1868 was not being matched caselessly.
1870 24. Arranged for all hyphens in the exim.8 source to be escaped with
1873 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1874 properly. Recipient callout cache records were still being keyed to include
1875 the sender, even when use_sender was set false. This led to far more
1876 callouts that were necessary. The sender is no longer included in the key
1877 when use_sender is false.
1879 26. Added "control = submission" modifier to ACLs.
1881 27. Added the ${base62d: operator to decode base 62 numbers.
1883 28. dnsdb lookups can now access SRV records.
1885 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1886 the configuration file.
1888 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1889 (-v) mode. This makes the output for a verbose queue run more intelligible.
1891 31. Added a use_postmaster feature to recipient callouts.
1893 32. Added the $body_zerocount variable, containing the number of binary zero
1894 bytes in the message body.
1896 33. The time of last modification of the "new" subdirectory is now used as the
1897 "mailbox time last read" when there is a quota error for a maildir
1900 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1902 35. Added +ignore_unknown as a special item in host lists.
1904 36. Code for decoding IPv6 addresses in host lists is now included, even if
1905 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1906 address was recognized as an IP address, but was then not correctly decoded
1907 into binary, causing unexpected and incorrect effects when compared with
1914 1. Very minor rewording of debugging text in manualroute to say "list of
1915 hosts" instead of "hostlist".
1917 2. If verify=header_syntax was set, and a header line with an unqualified
1918 address (no domain) and a large number of spaces between the end of the
1919 name and the colon was received, the reception process suffered a buffer
1920 overflow, and (when I tested it) crashed. This was caused by some obsolete
1921 code that should have been removed. The fix is to remove it!
1923 3. When running in the test harness, delay a bit after writing a bounce
1924 message to get a bit more predictability in the log output.
1926 4. Added a call to search_tidyup() just before forking a reception process. In
1927 theory, someone could use a lookup in the expansion of smtp_accept_max_
1928 per_host which, without the tidyup, could leave open a database connection.
1930 5. Added the variables $recipient_data and $sender_data which get set from a
1931 lookup success in an ACL "recipients" or "senders" condition, or a router
1932 "senders" option, similar to $domain_data and $local_part_data.
1934 6. Moved the writing of debug_print from before to after the "senders" test
1937 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1938 problems for message scanning, either using a data ACL, or using
1939 local_scan() because the Received: header was not generated till after they
1940 were called (in order to set the time as the time of reception completion).
1941 I have revised the way this works. The header is now generated after the
1942 body is received, but before the ACL or local_scan() are called. After they
1943 are run, the timestamp in the header is updated.
1949 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1950 before starting a queue runner without re-exec. This happened only when
1951 deliver_drop_privilege was set or when the Exim user was set to root. The
1952 effect of the bug was that timeouts during subsequent deliveries caused
1953 crashes instead of being properly handled. The handler is now left at its
1954 default (and expected) setting.
1956 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1957 message, again when deliver_drop_privilege is set or Exim is run as root.
1958 The bug described in (1) was not present in this case, but the tidying up
1959 of the other signals was missing. I have made the two cases consistent.
1961 3. The ignore_target_hosts setting on a manualroute router was being ignored
1962 for hosts that were looked up using the /MX notation.
1964 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1967 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1968 operated on the sender address. After changing the $sender_address to <>
1969 for the sender address verify, Exim was re-instated it as the original
1970 (before rewriting) address, but remembering that it had rewritten it, so it
1971 wasn't rewriting it again. This bug also had the effect of breaking the
1972 sender address verification caching when the sender address was rewritten.
1974 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1975 This has been changed so that if the ip literal address matches
1976 ignore_target_hosts, the router declines.
1978 7. Added expansion conditions match_domain, match_address, and match_local_
1979 part (NOT match_host).
1981 8. The placeholder for the Received: header didn't have a length field set.
1983 9. Added code to Exim itself and to exim_lock to test for a specific race
1984 condition that could lead to file corruption when using MBX delivery. The
1985 issue is with the lockfile that is created in /tmp. If this file is removed
1986 after a process has opened it but before that process has acquired a lock,
1987 there is the potential for a second process to recreate the file and also
1988 acquire a lock. This could lead to two Exim processes writing to the file
1989 at the same time. The added code performs the same test as UW imapd; it
1990 checks after acquiring the lock that its file descriptor still refers to
1991 the same named file.
1993 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1994 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1997 11. Added the smtp_active_hostname option. If used, this will typically be made
1998 to depend on the incoming interface address. Because $interface_address is
1999 not set up until the daemon has forked a reception process, error responses
2000 that can happen earlier (such as "too many connections") no longer contain
2003 12. If an expansion in a condition on a "warn" statement fails because a lookup
2004 defers, the "warn" statement is abandoned, and the next ACL statement is
2005 processed. Previously this caused the whole ACL to be aborted.
2007 13. Added the iplsearch lookup type.
2009 14. Added ident_timeout as a log selector.
2011 15. Added tls_certificate_verified as a log selector.
2013 16. Added a global option tls_require_ciphers (compare the smtp transport
2014 option of the same name). This controls incoming TLS connections.
2016 17. I finally figured out how to make tls_require_ciphers do a similar thing
2017 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2018 before starting the TLS session.
2020 18. Tabs are now shown as \t in -bP output.
2022 19. If the log selector return_path_on_delivery was set, Exim crashed when
2023 bouncing a message because it had too many Received: header lines.
2025 20. If two routers both had headers_remove settings, and the first one included
2026 a superfluous trailing colon, the final name in the first list and the
2027 first name in the second list were incorrectly joined into one item (with a
2028 colon in the middle).
2034 1. Added -C and -D options to the exinext utility, mainly to make it easier
2035 to include in the automated testing, but these could be helpful when
2036 multiple configurations are in use.
2038 2. The exinext utility was not formatting the output nicely when there was
2039 an alternate port involved in the retry record key, nor when there was a
2040 message id as well (for retries that were specific to a specific message
2041 and a specific host). It was also confused by IPv6 addresses, because of
2042 the additional colons they contain. I have fixed the IPv4 problem, and
2043 patched it up to do a reasonable job for IPv6.
2045 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2046 delivery, the log line now contains "pipelined" if PIPELINING was used.
2048 4. An SMTP transport process used to panic and die if the bind() call to set
2049 an explicit outgoing interface failed. This has been changed; it is now
2050 treated in the same way as a connect() failure.
2052 5. A reference to $sender_host_name in the part of a conditional expansion
2053 that was being skipped was still causing a DNS lookup. This no longer
2056 6. The def: expansion condition was not recognizing references to header lines
2057 that used bh_ and bheader_.
2059 7. Added the _cache feature to named lists.
2061 8. The code for checking quota_filecount in the appendfile transport was
2062 allowing one more file than it should have been.
2064 9. For compatibility with Sendmail, the command line option
2072 and sets the incoming protocol and host name (for trusted callers). The
2073 host name and its colon can be omitted when only the protocol is to be set.
2074 Note the Exim already has two private options, -pd and -ps, that refer to
2075 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2076 "s", but I don't think that's a major issue.
2078 10. A number of refactoring changes to the code, none of which should affect
2081 (a) The number of logging options was getting close to filling up the
2082 32-bit word that was used as a bit map. I have split them into two classes:
2083 those that are passed in the argument to log_write(), and those that are
2084 only ever tested independently outside of that function. These are now in
2085 separate 32-bit words, so there is plenty of room for expansion again.
2086 There is no change in the user interface or the logging behaviour.
2088 (b) When building, for example, log lines, the code previously used a
2089 macro that called string_cat() twice, in order to add two strings. This is
2090 not really sufficiently general. Furthermore, there was one instance where
2091 it was actually wrong because one of the argument was used twice, and in
2092 one call a function was used. (As it happened, calling the function twice
2093 did not affect the overall behaviour.) The macro has been replaced by a
2094 function that can join an arbitrary number of extra strings onto a growing
2097 (c) The code for expansion conditions now uses a table and a binary chop
2098 instead of a serial search (which was left over from when there were very
2099 few conditions). Also, it now recognizes conditions like "pam" even when
2100 the relevant support is not compiled in: a suitably worded error message is
2101 given if an attempt is made to use such a condition.
2103 11. Added ${time_interval:xxxxx}.
2105 12. A bug was causing one of the ddress fields not to be passed back correctly
2106 from remote delivery subprocesses. The field in question was not being
2107 subsequently used, so this caused to problems in practice.
2109 13. Added new log selectors queue_time and deliver_time.
2111 14. Might have fixed a bug in maildirsizefile handling that threw up
2112 "unexpected character" debug warnings, and recalculated the data
2113 unnecessarily. In any case, I expanded the warning message to give more
2116 15. Added the message "Restricted characters in address" to the statements in
2117 the default ACL that block characters like @ and % in local parts.
2119 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2120 Three changes have been made:
2122 (a) There was a serious bug; a negative response to MAIL caused the whole
2123 recipient domain to be cached as invalid, thereby blocking all messages
2124 to all local parts at the same domain, from all senders. This bug has
2125 been fixed. The domain is no longer cached after a negative response to
2126 MAIL if the sender used is not empty.
2128 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2131 (c) A new callout option, "use_sender" has been added for people who want
2132 the modified behaviour.
2138 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2141 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2142 indeed breaks things for older releases.
2144 3. Added additional logging to the case where there is a problem reading data
2145 from a filter that is running in a subprocess using a pipe, in order to
2146 try to track down a specific problem.
2148 4. Testing facility fudge: when running in the test harness and attempting
2149 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2150 getting "No route to host". Convert this to a timeout.
2152 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2155 6. Some OS don't have socklen_t but use size_t instead. This affects the
2156 fifth argument of getsockopt() amongst other things. This is now
2157 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2158 can be set for individual OS. I have set it for SunOS5, OSF1, and
2159 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2160 some earlier ones do not.
2162 7. Change 4.30/15 was not doing the test caselessly.
2164 8. The standard form for an IPv6 address literal was being rejected by address
2165 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2166 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2167 this, as well as the form without the "IPv6" on the front (but only when
2168 address literals are enabled, of course).
2170 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2172 10. Exim crashed if a message with an empty sender address specified by -f
2173 encountered a router with an errors_to setting. This could be provoked only
2174 by a command such as
2178 where an empty string was supplied; "<>" did not hit this bug.
2180 11. Installed PCRE release 4.5.
2182 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2183 remained set. It is now erased.
2185 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2186 times from message ids (which are base 36 rather than the normal 62).
2188 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2189 were being counted as actual protocol errors, and logged if the log
2190 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2191 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2192 and DATA following a set of rejected RCPTs do not count as protocol errors.
2193 In other words, Exim assumes they were pipelined, though this may not
2194 actually be the case. Of course, in all cases the client gets an
2195 appropriate error code.
2197 15. If a lookup fails in an ACL condition, a message about the failure may
2198 be available; it is used if testing the ACL cannot continue, because most
2199 such messages specify what the cause of the deferral is. However, some
2200 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2201 that caused an old message to be retained and used if a later statement
2202 caused a defer, replacing the real cause of the deferral.
2204 16. If an IP address had so many PTR records that the DNS lookup buffer
2205 was not large enough to hold them, Exim could crash while trying to process
2206 the truncated data. It now detects and logs this case.
2208 17. Further to 4.21/58, another change has been made: if (and only if) the
2209 first line of a message (the first header line) ends with CRLF, a bare LF
2210 in a subsequent header line has a space inserted after it, so as not to
2211 terminate the header.
2213 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2214 unnecessarily, used atoi() instead of strtol(), and didn't check the
2215 termination when getting file sizes from file names by regex.
2217 19. Completely re-implemented the support for maildirsize files, in the light
2218 of a number of problems with the previous contributed implementation
2219 (4.30/29). In particular:
2221 . If the quota is zero, the maildirsize file is maintained, but no quota is
2224 . If the maildir directory does not exist, it is created before any attempt
2225 to write a maildirsize file.
2227 . The quota value in the file is just a cache; if the quota is changed in
2228 the transport, the new value overrides.
2230 . A regular expression is available for excluding directories from the
2233 20. The autoreply transport checks the characters in options that define the
2234 message's headers; it allows continued headers, but it was checking with
2235 isspace() after an embedded newline instead of explicitly looking for a
2238 21. If all the "regular" hosts to which an address was routed had passed their
2239 expiry times, and had not reached their retry times, the address was
2240 bounced, even if fallback hosts were defined. Now Exim should go on to try
2243 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2244 equivalent code in the SMTP transport. Some hosts send humungous responses
2245 to HELO/EHLO, more than 1024 it seems.
2247 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2248 gives compiler warnings in some environments. I've now done it "properly",
2251 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2252 (because of problems with the built-in one) was declared to return uschar *
2253 instead of char *, causing compiler failure.
2255 25. Fixed a file descriptor leak when processing alias/forward files.
2257 26. Fixed a minor format string issue in dbfn.c.
2259 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2261 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2262 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2265 29. When a sender address is verified, it is cached, to save repeating the test
2266 when there is more than one recipient in a message. However, when the
2267 verification involves a callout, it is possible for different callout
2268 options to be set for different recipients. It is too complicated to keep
2269 track of this in the cache, so now Exim always runs a verification when a
2270 callout is required, relying on the callout cache for the optimization.
2271 The overhead is duplication of the address routing, but this should not be
2274 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2275 to be verified with callout=postmaster, and the main callout worked but the
2276 postmaster check failed, the verification correctly failed. However, if a
2277 subsequent RCPT command asked for sender verification *without* the
2278 postmaster check, incorrect caching caused this verification also to fail,
2281 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2282 it was not caching the DNS options (qualify_single, search_parents) that
2283 were used when the lookup failed. A subsequent lookup with different
2284 options therefore always gave the same answer, though there were cases
2285 where it should not have. (Example: a "domains = !$mx_any" option on a
2286 dnslookup router: the "domains" option is always processed without any
2287 widening, but the router might have qualify_single set.) Now Exim uses the
2288 cached value only when the same options are set.
2290 32. Added John Jetmore's "exipick" utility to the distribution.
2292 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2293 than a timeout (e.g. a certificate is required, and is not provided), an
2294 Exim server now closes the connection immediately. Previously it waited for
2295 the client to close - but if the client is SSL, it seems that they each
2296 wait for each other, leading to a delay before one of them times out.
2298 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2299 maintained 0.8.x compatibility because I don't think many are using it, and
2300 it is clearly obsolete.
2302 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2305 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2306 client certificate was expired. A simple patch fixes this, though I don't
2307 understand the full logic of why the verify callback is called multiple
2310 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2311 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2312 which causes problems with some clients (such as the Certicom SSL Plus
2313 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2314 disables the coutermeasure allowing Eudora to connect."
2316 38. Exim was not checking that a write() to a log file succeeded. This could
2317 lead to Bad Things if a log got too big, in particular if it hit a file
2318 size limit. Exim now panics and dies if it cannot write to a log file, just
2319 as it does if it cannot open a log file.
2321 39. Modified OS/Makefile-Linux so that it now contains
2323 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2325 The two -D definitions ensure that Exim is compiled with large file
2326 support, which makes it possible to handle log files that are bigger than
2329 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2330 instance) a domain was checked against a named list that involved a lookup,
2331 causing $domain_data to be set, then another domain was checked against the
2332 same list, then the first domain was re-checked, the value of $domain_data
2333 after the final check could be wrong. In particular, if the second check
2334 failed, it could be set empty. This bug probably also applied to
2337 41. The strip_trailing_dot option was not being applied to the address given
2338 with the -f command-line option.
2340 42. The code for reading a message's header from the spool was incrementing
2341 $received_count, but never initializing it. This meant that the value was
2342 incorrect (doubled) while delivering a message in the same process in which
2343 it was received. In the most common configuration of Exim, this never
2344 happens - a fresh exec is done - but it can happen when
2345 deliver_drop_privilege is set.
2347 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2348 it now includes up to 150 characters of the unexpected data in the log
2351 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2352 and building data strings. The size of both of these buffers was 10 000
2353 bytes - far larger than anybody would *ever* want, thought I. Needless to
2354 say, somebody hit the limit. I have increased the maximum line length to
2355 20 000 and the maximum data length of concatenated lines to 100 000. I have
2356 also fixed two bugs, because there was no checking on these buffers. Tsk,
2357 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2358 buffer is too small.
2360 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2361 lsearch lookups. Now it does.
2363 46. When parsing a route_list item in a manualroute router, a fixed-length
2364 buffer was used for the list of hosts. I made this 1024 bytes long,
2365 thinking that nobody would ever have a list of hosts that long. Wrong.
2366 Somebody had a whole pile of complicated expansion conditions, and the
2367 string was silently truncated, leading to an expansion error. It turns out
2368 that it is easier to change to an unlimited length (owing to other changes
2369 that have happened since this code was originally written) than to build
2370 structure for giving a limitation error. The length of the item that
2371 expands into the list of hosts is now unlimited.
2373 47. The lsearch lookup could not handle data where the length of text line was
2374 more than 4095 characters. Such lines were truncated, leading to shortened
2375 data being returned. It should now handle lines of any length.
2377 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2378 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2381 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2382 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2383 no longer happens in comment lines. A list of replacements is now placed
2384 at the head of all of the source files, except those whose only change is
2385 to replace PERL_COMMAND in the very first #! line.
2387 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2388 messages on the queue, with a bottom-up merge sort, using code contributed
2389 by Michael Haardt. This should make operations like -bp somewhat faster on
2390 large queues. It won't affect queue runners, except when queue_run_in_order
2393 51. Installed eximstats 1.31 in the distribution.
2395 52. Added support for SRV lookups to the dnslookup router.
2397 53. If an ACL referred to $message_body or $message_body_end, the value was not
2398 reset for any messages that followed in the same SMTP session.
2400 54. The store-handling optimization for building very long strings was not
2401 differentiating between the different store pools. I don't think this
2402 actually made any difference in practice, but I've tidied it.
2404 55. While running the routers to verify a sender address, $sender_address
2405 was still set to the sender address. This is wrong, because when routing to
2406 send a bounce to the sender, it would be empty. Therefore, I have changed
2407 it so that, while verifying a sender address, $sender_address is set to <>.
2408 (There is no change to what happens when verifying a recipient address.)
2410 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2411 target A or AAAA records (if not already returned) without resetting the
2412 qualify_single or search_parents options of the DNS resolver. These are
2413 inappropriate in this case because the targets of MX and SRV records must
2414 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2415 target that, when qualified, matched something in the local domain. These
2416 two options are now turned off when doing these lookups.
2418 57. It seems that at least some releases of Reiserfs (which does not have the
2419 concept of a fixed number of inodes) returns zero and not -1 for the
2420 number of available inodes. This interacted badly with check_spool_inodes,
2421 which assumed that -1 was the "no such thing" setting. What I have done is
2422 to check that the total number of inodes is greater than zero before doing
2423 the test of how many are available.
2425 58. When a "warn" ACL statement has a log_message modifier, the message is
2426 remembered, and not repeated. This is to avoid a lot of repetition when a
2427 message has many recipients that cause the same warning to be written.
2428 Howewer, Exim was preserving the list of already written lines for an
2429 entire SMTP session, which doesn't seem right. The memory is now reset if a
2430 new message is started.
2432 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2433 debugging output unless log_rewrite was also set.
2435 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2436 of (int)(handle) when we know that handle contains (void *)(-1).
2438 61. The Exim daemon panic-logs an error return when it closes the incoming
2439 connection. However "connection reset by peer" seems to be common, and
2440 isn't really an error worthy of noting specially, so that particular error
2443 62. When Exim is trying to find all the local interfaces, it used to panic and
2444 die if the ioctl to get the interface flags failed. However, it seems that
2445 on at least one OS (Solaris 9) it is possible to have an interface that is
2446 included in the list of interfaces, but for which you get a failure error
2447 for this call. This happens when the interface is not "plumbed" into a
2448 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2449 failure of the "get flags" call assumes that the interface is down.
2451 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2452 makes life easier for people who are doing arithmetic on fields extracted
2453 from dates, where you often get leading zeros that should not be
2454 interpreted as octal.
2456 64. Added qualify_domain to the redirect router, to override the global
2459 65. If a pathologically long header line contained very many addresses (the
2460 report of this problem mentioned 10 000) and each of them was rewritten,
2461 Exim could use up a very large amount of memory. (It kept on making new
2462 copies of the header line as it rewrote, and never released the old ones.)
2463 At the expense of a bit more processing, the header rewriting function has
2464 been changed so that it no longer eats memory in this way.
2466 66. The generation of the Received: header has been moved from the time that a
2467 message starts to be received, to the time that it finishes. The timestamp
2468 in the Received: header should now be very close to that of the <= log
2469 line. There are two side-effects of this change:
2471 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2472 logged header lines no longer include the local Received: line, because
2473 it has not yet been created. The same applies to a copy of the message
2474 that is returned to a non-SMTP sender when a message is rejected.
2476 (b) When a filter file is tested using -bf, no additional Received: header
2477 is added to the test message. After some thought, I decided that this
2480 This change does not affect the value of $received_for. It is still set
2481 after address rewriting, but before local_scan() is called.
2483 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2485 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2486 gave an unhelpful panic error message, and a defer error. I have managed to
2487 change this behaviour so that it now rejects any supplied certificate,
2488 which seems right, as the list of acceptable certificates is empty.
2490 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2491 gave an unhelpful defer error. I have not managed to make this reject any
2492 supplied certificates, but the error message it gives is "no certificate
2493 supplied", which is not helpful.
2495 70. exigrep's output now also includes lines that are not associated with any
2496 message, but which match the given pattern. Implemented by a patch from
2497 Martin Sluka, which also tidied up the Perl a bit.
2499 71. Recipient callout verification, like sender verification, was using <> in
2500 the MAIL FROM command. This isn't really the right thing, since the actual
2501 sender may affect whether the remote host accepts the recipient or not. I
2502 have changed it to use the actual sender in the callout; this means that
2503 the cache record is now keyed on a recipient/sender pair, not just the
2504 recipient address. There doesn't seem to be a real danger of callout loops,
2505 since a callout by the remote host to check the sender would use <>.
2506 [SEE ABOVE: changed after hitting problems.]
2508 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2509 temporary errors. However, in the case of such a code being given after
2510 the end of a data transmission (i.e. after ".") Exim was failing to write
2511 a retry record for the message. (Yes, there was some broken host that was
2512 actually sending 8xx at this point.)
2514 73. An unknown lookup type in a host list could cause Exim to panic-die when
2515 the list was checked. (An example that provoked this was putting <; in the
2516 middle of a list instead of at the start.) If this happened during a DATA
2517 ACL check, a -D file could be left lying around. This kind of configuration
2518 error no longer causes Exim to die; instead it causes a defer errror. The
2519 incident is still logged to the main and panic logs.
2521 74. Buglet left over from Exim 3 conversion. The message "too many messages
2522 in one connection" was written to the rejectlog but not the mainlog, except
2523 when address rewriting (yes!) was being logged.
2525 75. Added write_rejectlog option.
2527 76. When a system filter was run not as root (that is, when system_filter_user
2528 was set), the values of the $n variables were not being returned to the
2529 main process; thus, they were not subsequently available in the $sn
2532 77. Added +return_path_on_delivery log selector.
2534 78. A connection timeout was being treated differently from recipients deferred
2535 when testing hosts_max_try with a message that was older than the host's
2536 retry timeout. (The host should not be counted, thus allowing all hosts to
2537 be tried at least once before bouncing.) This may have been the cause of an
2538 occasionally reported bug whereby a message would remain on the queue
2539 longer than the retry timeout, but would be bounced if a delivery was
2540 forced. I say "may" because I never totally pinned down the problem;
2541 setting up timeout/retry tests is difficult. See also the next item.
2543 79. The ultimate address timeout was not being applied to errors that involved
2544 a combination of host plus message (for example, a timeout on a MAIL
2545 command). When an address resolved to a number of possible hosts, and they
2546 were not all tried for each delivery (e.g. because of hosts_max_try), a
2547 message could remain on the queue longer than the retry timeout.
2549 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2552 81. Fixed an obscure SMTP outgoing bug which required at least the following
2553 conditions: (a) there was another message waiting for the same server;
2554 (b) the server returned 5xx to all RCPT commands in the first message so
2555 that the message was not completed; (c) the server dropped the connection
2556 or gave a negative response to the RSET that Exim sends to abort the
2557 transaction. The observed case was a dropped connection after DATA that had
2558 been sent in pipelining mode. That is, the server had advertised PIPELINING
2559 but was not implementing it correctly. The effect of the bug was incorrect
2560 behaviour, such as trying another host, and this could lead to a crash.
2566 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2567 and daemon.c were passed as pointers to ints; they should have been
2568 pointers to socklen_t variables (which are typically unsigned ints).
2570 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2573 3. Fixed a really odd bug that affected only the testing scheme; patching a
2574 certain fixed string in the binary changed the value of another string that
2575 happened to be identical to the end of the original first string.
2577 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2578 name", it returns that address as the IP address. On some operating
2579 systems (e.g. Solaris), it also passes back the IP address string as the
2580 "host name". However, on others (e.g. Linux), it passes back an empty
2581 string. Exim wasn't checking for this, and was changing the host name to an
2582 empty string, assuming it had been canonicized.
2584 5. Although rare, it is permitted to have more than one PTR record for a given
2585 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2586 all the names associated with an address, because they do in Solaris.
2587 However, it seems that they do not in Linux for data that comes from the
2588 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2589 I found this out when I moved to a new Linux workstation and tried to run
2590 the Exim test suite.
2592 To get round this problem I have changed the code so that it now does its
2593 own call to the DNS to look up PTR records when searching for a host name.
2594 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2595 addresses that are only in /etc/hosts are still found.
2597 This behaviour is, however, controlled by an option called host_lookup_
2598 order, which defaults to "bydns:byaddr". If people want to use the other
2599 order, or indeed, just use one or the other means of lookup, they can
2600 specify it in this variable.
2602 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2603 some operating systems, this comes back from gethostbyaddr() as an empty
2604 string, and this is what Exim used to test for. However, it seems that in
2605 other systems, "." is yielded. Exim now tests for this case too.
2607 7. The values of check_spool_space and check_log_space are now held internally
2608 as a number of kilobytes instead of an absolute number of bytes. If a
2609 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2610 kilobyte. This means that much larger values can be stored.
2612 8. Exim monitor: an attempt to get the action menu when not actually pointing
2613 at a message produces an empty menu entitled "No message selected". This
2614 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2615 no entries in it ("Shell widget menu has zero width and/or height"). So I
2616 have added a single, blank menu entry in this case.
2618 9. Added ${quote_local_part.
2620 10. MIME decoding is now applied to the contents of Subject: header lines when
2623 11. Now that a reference to $sender_host_address automatically causes a reverse
2624 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2625 host lookup before query-style lookups in lists that might use this
2626 variable. This has therefore been abolished, and the "net-" prefix is no
2627 longer necessary for query-style lookups.
2629 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2630 have been a typo for LFLAGS, so it has been changed.
2632 13. The install script calls Exim with "-C /dev/null" in order to find the
2633 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2634 to be output. Howeve, since Exim outputs its version number before the
2635 error, it didn't break the script. It just looked ugly. I fixed this by
2636 always allowing "-C /dev/null" if the caller is root.
2638 14. Ignore overlarge ACL variable number when reading spool file - insurance
2639 against a later release with more variables having written the file.
2641 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2642 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2643 this, as well as the form without the "IPv6" on the front.
2645 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2646 OS/Makefile-Darwin file.
2648 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2649 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2651 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2652 variable. However, some OS use ints for inodes. I've added cast to long int
2653 to get rid of the compiler warning.
2655 19. I had forgotten to lock out "/../" in configuration file names when
2656 ALT_CONFIG_PREFIX was set.
2658 20. Routers used for verification do not need to specify transports. However,
2659 if such a router generated a host list, and callout was configured, Exim
2660 crashed, because it could not find a port number from the (non-existent)
2661 transport. It now assumes port 25 in this circumstance.
2663 21. Added the -t option to exigrep.
2665 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2666 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2667 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2668 them is set, LOOKUP_LSEARCH is forced.
2670 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2672 24. Added sender and host information to the "rejected by local_scan()" log
2673 line; previously there was no indication of these.
2675 25. Added .include_if_exists.
2677 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2678 Linux. It turns out that not all file systems support this. Apparently some
2679 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2680 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2681 not supported on the file descriptor, is now ignored when Exim is trying to
2682 sync a directory. This applies only to Linux.
2684 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2686 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2687 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2690 29. Added support for maildirsize files from supplied patch (modified a bit).
2692 30. The use of :fail: followed by an empty string could lead Exim to respond to
2693 sender verification failures with (e.g.):
2695 550 Verification failed for <xxx>
2696 550 Sender verify failed
2698 where the first response line was missing the '-' that indicates it is not
2699 the final line of the response.
2701 31. The loop for finding the name of the user that called Exim had a hardwired
2702 limit of 10; it now uses the value of finduser_retries, which is used for
2703 all other user lookups.
2705 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2708 33. Exim was neglecting to zero errno before one call of strtol() when
2709 expanding a string and expecting an integer value. On some systems this
2710 resulted in spurious "integer overflow" errors. Also, it was casting the
2711 result into an int without checking.
2713 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2714 did not work. The code looks as if it has *never* worked, though it appears
2715 to have been documented since at least releast 1.62. I have made it work.
2717 35. The "timeout_DNS" error in retry rules, also documented since at least
2718 1.62, also never worked. As it isn't clear exactly what this means, and
2719 clearly it isn't a major issue, I have abolished the feature by treating it
2720 as "timeout", and writing a warning to the main and panic logs.
2722 36. The display of retry rules for -brt wasn't always showing the error code
2725 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2726 timeout_connect_A, timeout_connect_MX.
2728 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2729 to the empty sender.
2731 39. The daemon was not analysing the content of -oX till after it had closed
2732 stderr and disconnected from the controlling terminal. This meant that any
2733 syntax errors were only noted on the panic log, and the return code from
2734 the command was 0. By re-arranging the code a little, I've made the
2735 decoding happen first, so such errors now appear on stderr, and the return
2736 code is 1. However, the actual setting up of the sockets still happens in
2737 the disconnected process, so errors there are still only recorded on the
2740 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2741 connections (as happens on some IP stacks) was logged at start up time as
2742 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2743 it from "IPv6 and IPv4", which means that two separate sockets are being
2746 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2747 says whether AF_INET or AF_INET6 was passed as an argument.
2749 42. Exiwhat output was messed up when time zones were included in log
2752 43. Exiwhat now gives more information about the daemon's listening ports,
2753 and whether -tls-on-connect was used.
2755 44. The "port" option of the smtp transport is now expanded.
2757 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2758 silently ignored. Now an error message is written to the main and panic
2761 46. There's a new ACL modifier called "logwrite" which writes to a log file
2762 as soon as it is encountered.
2764 47. Added $local_user_uid and $local_user_gid at routing time.
2766 48. Exim crashed when trying to verify a sender address that was being
2769 49. Exim was recognizing only a space character after ".include". It now also
2770 recognizes a tab character.
2772 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2773 extracting the relevant information from the specification. The man page no
2774 longer contains scrambled data for the -d option, and I've added a section
2775 at the front about calling Exim under different names.
2777 51. Added "extra_headers" argument to the "mail" command in filter files.
2779 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2782 53. Installed eximstats 1.29.
2784 54. Added transport_filter_timeout as a generic transport option.
2786 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2787 Cc: header lines. This was required by RFC 822, but it not required by RFC
2790 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2791 incoming messages that did not have them. Now it does so only if the
2792 message originates locally, that is, if there is no associated remote host
2793 address. When Resent- header lines are present, this applies to the Resent-
2794 lines rather than the non-Resent- lines.
2796 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2797 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2799 58. Messages for configuration errors now include the name of the main
2800 configuration file - useful now that there may be more than one file in a
2801 list (.included file names were always shown).
2803 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2804 for those rare installations that do not start the daemon as root or run it
2805 setuid root. I've cut out the call to initgroups() if the daemon is not
2808 60. The Exim user and group can now be bound into the binary as text strings
2809 that are looked up at the start of Exim's processing.
2811 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2813 62. Added $mailstore_basename variable.
2815 63. Installed patch to sieve.c from Michael Haardt.
2817 64. When Exim failed to open the panic log after failing to open the main log,
2818 the original message it was trying to log was written to stderr and debug
2819 output, but if they were not available (the usual case in production), it
2820 was lost. Now it is written to syslog before the two lines that record the
2821 failures to open the logs.
2823 65. Users' Exim filters run in subprocesses under the user's uid. It is
2824 possible for a "deliver" command or an alias in a "personal" command to
2825 provoke an address rewrite. If logging of address rewriting is configured,
2826 this fails because the process is not running as root or exim. There may be
2827 a better way of dealing with this, but for the moment (because 4.30 needs
2828 to be released), I have disabled address rewrite logging when running a
2829 filter in a non-root, non-exim process.
2835 1. The buildconfig auxiliary program wasn't quoting the value set for
2836 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2837 not defined. This bug was masked in 4.22 by the effect that was fixed in
2840 2. Some messages that were rejected after a message id was allocated were
2841 shown as "incomplete" by exigrep. It no longer does this for messages that
2842 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2844 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2845 have allow_domain_literals set, the ID did not get logged in the <= line.
2846 Domain literals are now always recognized in Message-ID: header lines.
2848 4. The first argument for a ${extract expansion item is the key name or field
2849 number. Leading and trailing spaces in this item were not being ignored,
2850 causing some misleading effects.
2852 5. When deliver_drop_privilege was set, single queue runner processes started
2853 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2854 same command in the process it spins off) were not dropping privilege.
2856 6. When the daemon running as "exim" started a queue runner, it always
2857 re-executed Exim in the spun-off process. This is a waste of effort when
2858 deliver_drop_privilege is set. The new process now just calls the
2859 queue-runner function directly.
2865 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2868 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2869 ignored. Though the use of -oP was forcing the writing of a pid file, it
2870 was always written to the default place.
2872 3. If the message "no IP address found for host xxxx" is generated during
2873 incoming verification, it is now followed by identification of the incoming
2874 connection (so you can more easily find what provoked it).
2876 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2878 5. Added some features to "harden" Exim a bit more against certain attacks:
2880 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2881 be put in Local/Makefile. This is like the never_users runtime option,
2882 but it cannot be overridden. The default setting is "root".
2884 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2885 prefix string with which any file named in a -C command line option
2888 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2889 is retained for -C and -D only if the caller of Exim is root. Without
2890 it, the exim user may also use -C and -D and retain privilege.
2892 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2893 command line option is disabled.
2895 6. Macro names set by the -D option must start with an upper case letter, just
2896 like macro names defined in the configuration file.
2898 7. Added "dereference=" facility to LDAP.
2900 8. Two instances of the typo "uknown" in the source files are fixed.
2902 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2903 the Configure-Makefile script screwed up while processing it.
2905 10. Incorporated PCRE 4.4.
2907 11. The SMTP synchronization check was not operating right at the start of an
2908 SMTP session. For example, it could not catch a HELO sent before the client
2909 waited for the greeting. There is now a check for outstanding input at the
2910 point when the greeting is written. Because of the duplex, asynchronous
2911 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2912 way, but not yet received, when the check is performed.
2914 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2915 on TCP/IP sockets, because this apparently causes some broken clients to
2918 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2919 unchanged) from the Cygwin maintainer.
2921 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2922 was supported instead of testing for mbx. Effectively a typo.
2924 15. The spa authenticator server code was not checking that the input it
2925 received was valid base64.
2927 16. The debug output line for the "set" modifier in ACLs was not showing the
2928 name of the variable that was being set.
2930 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2932 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2933 lookup on its own. Something else was needed to trigger the lookup. For
2934 example, a match in host_lookup or the need for a host name in a host list.
2935 Now, if $sender_host_name is referenced and the host name has not yet been
2936 looked up, a lookup is performed. If the lookup fails, the variable remains
2937 empty, and $host_lookup_failed is set to "1".
2939 19. Added "eqi" as a case-independent comparison operator.
2941 20. The saslauthd authentication condition could segfault if neither service
2942 nor realm was specified.
2944 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2945 error message that was logged was misleading, and incoming SMTP
2946 connections were dropped. The message is now more accurate, and temporary
2947 errors are given to SMTP connections.
2949 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2950 (except RSET) with a 421 error, until QUIT is received. However, it was
2951 failing to send a response to QUIT.
2953 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2954 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2955 ACL is now run after the helo_try_verify_hosts code.
2957 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2958 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2959 case-independent, so other case variants are also recognized). Apparently
2960 some systems use these upper case variants.
2962 25. If more than two messages were waiting for the same host, and a transport
2963 filter was specified for the transport, Exim sent two messages over the
2964 same TCP/IP connection, and then failed with "socket operation on non-
2965 socket" when it tried to send the third.
2967 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2969 27. The extern definition of crypt16() in expand.c was not being excluded when
2970 the OS had its own crypt16() function.
2972 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2973 as a preferred synonym for return_size_limit, both as an option and as an
2976 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2978 30. Changed the default configuration ACL to relax the local part checking rule
2979 for addresses that are not in any local domains. For these addresses,
2980 slashes and pipe symbols are allowed within local parts, but the sequence
2981 /../ is explicitly forbidden.
2983 31. SPA server authentication was not clearing the challenge buffer before
2986 32. log_message in a "warn" ACL statement was writing to the reject log as
2987 well as to the main log, which contradicts the documentation and doesn't
2988 seem right (because no rejection is happening). So I have stopped it.
2990 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2991 However, I am unable to do any testing of this.
2993 34. Fixed an infelicity in the appendfile transport. When checking directories
2994 for a mailbox, to see if any needed to be created, it was accidentally
2995 using path names with one or more superfluous leading slashes; tracing
2996 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2998 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2999 contents are added to the log line that is written for every discarded
3000 recipient. (Previously a log_message setting was ignored.)
3002 36. The ${quote: operator now quotes the string if it is empty.
3004 37. The install script runs exim in order to find its version number. If for
3005 some reason other than non-existence or emptiness, which it checks, it
3006 could not run './exim', it was installing it with an empty version number,
3007 i.e. as "exim-". This error state is now caught, and the installation is
3010 38. An argument was missing from the function that creates an error message
3011 when Exim fails to connect to the socket for saslauthd authentication.
3012 This could cause Exim to crash, or give a corrupted message.
3014 39. Added isip, isip4, and isip6 to ${if conditions.
3016 40. The ACL variables $acl_xx are now saved with the message, and can be
3017 accessed later in routers, transports, and filters.
3019 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3020 strings in the file are not string-expanded.
3022 42. If a MAIL command specified a SIZE value that was too large to fit into an
3023 int variable, the check against message_size_limit failed. Such values are
3024 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3025 day this will have to be increased, but I don't think I want to be around
3026 when emails are that large.
3033 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3034 iconv() is not standard in FreeBSD.
3036 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3037 IPv6 enabled. The observed symptom was a segmentation fault on return from
3038 the function os_common_find_running_interfaces() in src/os.c.
3040 3. In the check_special_case() function in daemon.c I had used "errno" as an
3041 argument name, which causes warnings on some systems. This was basically a
3042 typo, since it was named "eno" in the comments!
3044 4. The code that waits for the clock to tick (at a resolution of some fraction
3045 of a second) so as to ensure message-id uniqueness was always waiting for
3046 at least one whole tick, when it could have waited for less. [This is
3047 almost certainly not relevant at current processor speeds, where it is
3048 unlikely to ever wait at all. But we try to future-proof.]
3050 5. The function that sleeps for a time interval that includes fractions of a
3051 second contained a race. It did not block SIGALRM between setting the
3052 timer, and suspending (a couple of lines later). If the interval was short
3053 and the sigsuspend() was delayed until after it had expired, the suspension
3054 never ended. On busy systems this could lead to processes getting stuck for
3057 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3058 process, before it forks any delivery processes. The open lookup caching
3059 mechanism meant that the open file or database connection was passed into
3060 the delivery process. The problem was that delivery processes always tidy
3061 up cached lookup data. This could cause a problem for the next delivery
3062 process started by the queue runner, because the external queue runner
3063 process does not know about the closure. So the next delivery process
3064 still has data in the lookup cache. In the case of a file lookup, there was
3065 no problem because closing a file descriptor in a subprocess doesn't affect
3066 the parent. However, if the lookup was caching a connection to a database,
3067 the connection was closed, and the second delivery process was likely to
3068 see errors such as "PGSQL: query failed: server closed the connection
3069 unexpectedly". The problem has been fixed by closing all cached lookups
3070 in a queue runner before running a delivery process.
3072 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3073 seem to have the "const" qualifier which it has on other OS. I've
3076 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3077 *of the same type* (client or server) with the same public name that an
3078 error should be diagnosed.
3080 9. When Exim looked up a host name for an IP address, but failed to find the
3081 original IP address when looking up the host name (a safety check), it
3082 output the message "<ip address> does not match any IP for NULL", which was
3083 confusing, to say the least. The bug was that the host name should have
3084 appeared instead of "NULL".
3086 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3087 user that is built into the binary, and the -C or -D options is used, root
3088 privilege is dropped before the configuration file is read. In addition,
3089 logging is switched to stderr instead of the normal log files. If the
3090 configuration then re-defines the Exim user, the unprivileged environment
3091 is probably not what is expected, so Exim logs a panic warning message (but
3094 However, if deliver_drop_privilege is set, the unprivileged state may well
3095 be exactly what is intended, so the warning has been cut out in that case,
3096 and Exim is allowed to try to write to its normal log files.
3102 1. smtp_return_error_details was not giving details for temporary sender
3103 or receiver verification errors.
3105 2. Diagnose a configuration error if two authenticators have the same public
3108 3. Exim used not to create the message log file for a message until the first
3109 delivery attempt. This could be confusing when incoming messages were held
3110 for policy or load reasons. The message log file is now created at the time
3111 the message is received, and an initial "Received" line is written to it.
3113 4. The automatically generated man page for command line options had a minor
3114 bug that caused no ill effects; however, a more serious problem was that
3115 the procedure for building the man page automatically didn't always
3116 operate. Consequently, release 4.20 contains an out-of-date version. This
3117 shouldn't happen again.
3119 5. When building Exim with embedded Perl support, the script that builds the
3120 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3121 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3123 6. The freeze_tell option was not being used for messages that were frozen on
3124 arrival, either by an ACL or by local_scan().
3126 7. Added the smtp_incomplete_transaction log selector.
3128 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3129 was accepting AUTH without a new EHLO.
3131 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3132 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3135 10. Logging of TCP/IP connections (when configured) now happens in the main
3136 daemon process instead of the child process, so that the TCP/IP connection
3137 count is more accurate (but it can never be perfect).
3139 11. The use of "drop" in a nested ACL was not being handled correctly in the
3140 outer ACL. Now, if condition failure induced by the nested "drop" causes
3141 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3142 or "require"), the connection is dropped.
3144 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3145 that yield "discard" can now be used with an "accept" or a "discard" verb,
3146 but an error is generated for any others (because I can't see a useful way
3147 to define what should happen).
3149 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3150 are now processed in the same way as lines in the Exim configuration file.
3151 In particular, continuation lines are supported.
3153 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3155 15. Added -ti meaning -t -i.
3157 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3158 domains, and warn by logging if others are found.
3160 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3161 returned from GIFCONF when Exim is trying to find the list of interfaces on
3162 a host. The code in os.c has been modified to copy each ifreq to an aligned
3163 structure in all cases.
3165 Also, in some cases, the returned ifreq's were being copied to a 'struct
3166 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3167 means the last couple of bytes of an IPv6 address could be chopped if the
3168 ifreq contained only a normal sockaddr (14 bytes storage).
3170 18. Named domain lists were not supported in the hosts_treat_as_local option.
3171 An entry such as +xxxx was not recognized, and was treated as a literal
3174 19. Ensure that header lines added by a DATA ACL are included in the reject log
3175 if the ACL subsequently rejects the message.
3177 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3178 MD5 (which is deprecated).
3180 21. When testing a filter file using -bf, Exim was writing a message when it
3181 took the sender from a "From " line in the message, but it was not doing so
3182 when it took $return_path from a Return-Path: header line. It now does.
3184 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3185 with a valid header line field name (a series of printing characters
3186 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3188 23. Changed "disc" in the source to "disk" to conform to the documentation and
3189 the book and for uniformity.
3191 24. Ignore Sendmail's -Ooption=value command line item.
3193 25. When execve() failed while trying to run a command in a pipe transport,
3194 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3195 could be confused with a return value of 69 from the command itself. This
3196 has been changed to 127, the value the shell returns if it is asked to run
3197 a non-existent command. The wording for the related log line suggests a
3198 non-existent command as the problem.
3200 26. If received_header_text expands to an empty string, do not add a Received:
3201 header line to the message. (Well, it adds a token one on the spool, but
3202 marks it "old" so that it doesn't get used or transmitted.)
3204 27. Installed eximstats 1.28 (addition of -nt option).
3206 28. There was no check for failure on the call to getsockname() in the daemon
3207 code. This can fail if there is a shortage of resources on the system, with
3208 ENOMEM, for example. A temporary error is now given on failure.
3210 29. Contrary to the C standard, it seems that in some environments, the
3211 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3212 program. Exim now does this explicitly; it affects the formatting of
3213 timestamps using strftime().
3215 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3216 complaints. I've now initialized all the variables, to avoid this.
3218 32. Header lines added by a system filter were not being "seen" during
3219 transport-time rewrites.
3221 33. The info_callback() function passed to OpenSSL is set up with type void
3222 (*)(SSL *, int, int), as described somewhere. However, when calling the
3223 function (actually a macro) that sets it up, the type void(*)() is
3224 expected. I've put in a cast to prevent warnings from picky compilers.
3226 34. If a DNS black list lookup found a CNAME record, but there were no A
3227 records associated with the domain it pointed at, Exim crashed.
3229 35. If a DNS black list lookup returned more than one A record, Exim ignored
3230 all but the first. It now scans all returned addresses if a particular IP
3231 value is being sought. In this situation, the contents of the
3232 $dnslist_value variable are a list of all the addresses, separated by a
3235 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3236 to accept a host name and all its aliases if the forward lookup for any of
3237 them yielded the IP address of the incoming connection. Now it accepts only
3238 those names whose forward lookup yields the correct IP address. Any other
3239 names are discarded. This closes a loophole whereby a rogue DNS
3240 administrator could create reverse DNS records to break through a
3241 wildcarded host restriction in an ACL.
3243 37. If a user filter or a system filter that ran in a subprocess used any of
3244 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3245 the wrong values were passed to the pipe command ($thisaddress had the
3246 value of $0, $0 had the value of $1, etc). This bug was introduced by
3247 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3249 38. Improved the line breaking for long SMTP error messages from ACLs.
3250 Previously, if there was no break point between 40 and 75 characters, Exim
3251 left the rest of the message alone. Two changes have been made: (a) I've
3252 reduced the minimum length to 35 characters; (b) if it can't find a break
3253 point between 35 and 75 characters, it looks ahead and uses the first one
3254 that it finds. This may give the occasional overlong line, but at least the
3255 remaining text gets split now.
3257 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3258 file descriptors might be low, and that setting 1000 would always raise it.
3259 It turns out that in some environments, the limit is already over 1000 and
3260 that lowering it causes trouble. So now Exim takes care not to decrease it.
3262 40. When delivering a message, the value of $return_path is set to $sender_
3263 address at the start of routing (routers may change the value). By an
3264 oversight, this default was not being set up when an address was tested by
3265 -bt or -bv, which affected the outcome if any router or filter referred to
3268 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3269 log only when "message" or "log_message" are set. However, if one of the
3270 conditions was an address verification, or a call to a nested ACL, the
3271 messages generated by the underlying test were being passed through. This
3272 no longer happens. The underlying message is available in $acl_verify_
3273 message for both "message" and "log_message" expansions, so it can be
3274 passed through if needed.
3276 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3277 new expansion $bh_ to give the encoded byte string without charset
3278 translation. Translation happens only if iconv() is available; HAVE_ICONV
3279 indicates this at build time. HEADERS_CHARSET gives the charset to
3280 translate to; headers_charset can change it in the configuration, and
3281 "headers charset" can change it in an individual filter file.
3283 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3284 that creates RFC 2047 encoded "words" labels them as that charset instead
3285 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3286 expansion operator; (ii) when Exim creates a From: line for a local
3287 message; (iii) when a header line is rewritten to include a "phrase" part.
3289 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3290 buggy, causing it to skip the first lines of messages whose message ID
3291 ended in 'D'. This would not have bitten before Exim release 4.14, because
3292 message IDs were unlikely to end in 'D' before then. The effect was to have
3293 incorrect size information for certain domains.
3295 45. #include "config.h" was missing at the start of the crypt16.c module. This
3296 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3299 46. If there was a timeout during a "random" callout check, Exim treated it as
3300 a failure of the random address, and carried on sending RSET and the real
3301 address. If the delay was just some slowness somewhere, the response to the
3302 original RCPT would be taken as a response to RSET and so on, causing
3303 mayhem of various kinds.
3305 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3306 when I implemented it. It didn't allow for the fact that some option values
3307 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3308 the right test for positive values.
3310 48. Domain names in DNS records are case-independent. Exim always looks them up
3311 in lower case. Some resolvers return domain names in exactly the case they
3312 appear in the zone file, that is, they may contain uppercase letters. Not
3313 all resolvers do this - some return always lower case. Exim was treating a
3314 change of case by a resolver as a change of domain, similar to a widening
3315 of a domain abbreviation. This triggered its re-routing code and so it was
3316 trying to route what was effectively the same domain again. This normally
3317 caused routing to fail (because the router wouldn't handle the domain
3318 twice). Now Exim checks for this case specially, and just changes the
3319 casing of the domain that it ultimately uses when it transmits the message
3322 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3325 50. If a filter generated a file delivery with a non-absolute name (possible if
3326 no home directory exists for the router), the forbid_file option was not
3329 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3330 the existing equality matching.
3332 52. Exim was using ints instead of ino_t variables in some places where it was
3333 dealing with inode numbers.
3335 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3336 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3337 called TMPDIR, and if it finds it is different, it changes its value.
3339 54. The smtp_printf() function is now made available to local_scan() so
3340 additional output lines can be written before returning. There is also an
3341 smtp_fflush() function to enable the detection of a dropped connection.
3342 The variables smtp_input and smtp_batched_input are exported to
3345 55. Changed the default runtime configuration: the message "Unknown user"
3346 has been removed from the ACL, and instead placed on the localuser router,
3347 using the cannot_route_message feature. This means that any verification
3348 failures that generate their own messages won't get overridden. Similarly,
3349 the "Unrouteable address" message that was in the ACL for unverifiable
3350 relay addresses has also been removed.
3352 56. Added hosts_avoid_esmtp to the smtp transport.
3354 57. The exicyclog script was not checking for the esoteric option
3355 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3356 will work only if exicyclog is run under the appropriate euid.
3358 58. Following a discussion on the list, the rules by which Exim recognises line
3359 endings on incoming messages have been changed. The -dropcr and drop_cr
3360 options are now no-ops, retained only for backwards compatibility. The
3361 following line terminators are recognized: LF CRLF CR. However, special
3362 processing applies to CR:
3364 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3365 nor a local message in the state where . is a terminator.
3367 (ii) If a bare CR is encountered in a header line, an extra space is added
3368 after the line terminator so as not to end the header. The reasoning
3369 behind this is that bare CRs in header lines are most likely either
3370 to be mistakes, or people trying to play silly games.
3372 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3373 was being incorrectly given as 18 bytes larger than it should have been.
3374 This is a VOB (very old bug).
3376 60. This may never have affected anything current, but just in case it has:
3377 When the local host is found other than at the start of a list of hosts,
3378 the local host, those with the same MX, and any that follow, are discarded.
3379 When the list in question was part of a longer list of hosts, the following
3380 hosts (not currently being processed) were also being discarded. This no
3381 longer happens. I'm not sure if this situation could ever has previously
3384 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3387 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3388 header. This is something that is recommended in a new Internet Draft, and
3389 is something that is documented as being done by Sendmail. There are two
3390 possible values. For messages generated by the autoreply transport, Exim
3393 Auto-Submitted: auto-replied
3395 whereas for all other generated messages (e.g. bounces) it adds
3397 Auto-Submitted: auto-generated
3399 63. The "personal" condition in filters now includes a test for the
3400 Auto-Submitted: header. If it contains the string "auto-" the message it
3401 not considered personal.
3403 64. Added rcpt_include_affixes as a generic transport option.
3405 65. Added queue_only_override (default true).
3407 66. Added the syslog_duplication option.
3409 67. If what should have been the first header line of a message consisted of
3410 a space followed by a colon, Exim was mis-interpreting it as a header line.
3411 It isn't of course - it is syntactically invalid and should therefore be
3412 treated as the start of the message body. The misbehaviour could have
3413 caused a number of strange effects, including loss of data in subsequent
3414 header lines, and spool format errors.
3416 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3417 client host had authenticated. This control can now be exercised by an ACL
3418 for more flexibility.
3420 69. By default, callouts do not happen when testing with -bh. There is now a
3421 variant, -bhc, which does actually run the callout code, including
3422 consulting and updating the callout cache.
3424 70. Added support for saslauthd authentication, courtesy of Alexander
3427 71. If statvfs() failed on the spool or log directories while checking their
3428 size for availability, Exim confusingly gave the error "space shortage".
3429 Furthermore, in debugging mode it crashed with a floating point exception.
3430 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3431 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3432 serious problem, Exim now writes to the main and panic logs when this
3433 happens, with details of the failure. It then refuses to accept the
3434 incoming message, giving the message "spool directory problem" or "log
3435 directory problem" with a 421 code for SMTP messages.
3437 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3438 0, 1, and 2 exist, because some OS complain for execs without them (see
3439 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3440 descriptors. However, the code omitted to check that the open succeeded,
3441 causing mysterious errors if for some reason the permissions on /dev/null
3442 got screwed. Now Exim writes a message to the main and panic logs, and
3443 bombs out if it can't open /dev/null.
3445 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3446 interact so that it is all more flexible. It is supposed to remain
3447 backwards compatible. Also added extra_local_interfaces.
3449 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3450 to bomb out with an assertion failure - to the client this appears as a
3451 connection drop. This problem occurs in the part of the code that was taken
3452 from the Samba project. Fortunately, the assertion is in a very simple
3453 function, so I have fixed this by reproducing the function inline in the
3454 one place where it is called, and arranging for authentication to fail
3455 instead of killing the process with assert().
3457 75. The SPA client code was not working when the server requested OEM rather
3458 than Unicode encoding.
3460 76. Added code to make require_files with a specific uid setting more usable in
3461 the case where statting the file as root fails - usually a non-root-mounted
3462 NFS file system. When this happens and the failure is EACCES, Exim now
3463 forks a subprocess and does the per-uid checking as the relevant uid.
3465 77. Added process_log_path.
3467 78. If log_file_path was not explicitly set, a setting of check_log_space or
3468 check_log_inodes was ignored.
3470 79. If a space check for the spool or log partitions fails, the incident is now
3471 logged. Of course, in the latter case the data may get lost...
3473 80. Added the %p formatting code to string_format() so that it can be used to
3474 print addresses in debug_print(). Adjusted all the address printing in the
3475 debugging in store.c to use %p rather than %d.
3477 81. There was a concern that a line of code in smtp_in.c could overflow a
3478 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3479 initially expressed, the concern was not well-founded, because trailing
3480 spaces are removed early. However, if the trailing spaces were followed by
3481 a NULL, they did not get removed, so the overflow was possible. Two fixes
3484 (a) I re-wrote the offending code in a cleaner fashion.
3485 (b) If an incoming SMTP command contains a NULL character, it is rejected
3488 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3489 runs initgroups(), so that if the Exim user is in any additional groups,
3490 they will be used during message reception.
3496 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob