1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.254 2005/11/11 10:02:04 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.
71 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
72 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
75 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
76 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
78 PH/03 Typo: missing ".o" in src/pcre/Makefile.
80 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
81 header line, restrict the check to what is listed in RFCs 2369 and 2929.
82 Also, for "Auto-Submitted", treat anything other than "no" as
83 non-personal, in accordance with RFC 3834. (Previously it treated
84 anything starting "auto-" as non-personal.)
86 TF/01 The control=submission/name=... option had a problem with syntax
87 errors if the name included a slash character. The /name= option
88 now slurps the rest of the string, so it can include any characters
89 but it must come last in the list of options (after /sender_retain
92 PH/05 Some modifications to the interface to the fake nameserver for the new
99 TK/01 Added the "success_on_redirect" address verification option. See
100 NewStuff for rationale and an example.
102 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
104 PH/02 Patch to exigrep to allow it to work on syslog lines.
106 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
107 fread() to skip over the body file's header line, because in Cygwin the
108 header line is locked and is inaccessible.
110 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
111 co-exist for some time) to make it clear that it is the Exim ID that is
112 referenced, not the Message-ID: header line.
114 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
115 string_format() function, because snprintf() does not exist on all
118 PH/06 The use of forbid_filter_existstest now also locks out the use of the
119 ${stat: expansion item.
121 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
122 protocol synchronization error", to keep the pedants happy.
124 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
125 well as for IRIX systems, when gcc is being used. See the host.c source
128 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
130 PH/10 Named domain lists were not working if used in a queue_smtp_domains
133 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
134 transport and to the smtp transport in LMTP mode.
136 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
138 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
139 run a filter in a subprocess. This could lead to confusion in subsequent
140 lookups in the parent process. There should also be a search_tidyup() at
141 the end of the subprocess.
143 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
144 only if the host matched helo_try_verify_hosts, which caused the
145 verification to occur when the EHLO/HELO command was issued. The ACL just
146 tested the remembered result. Now, if a previous verification attempt has
147 not happened, "verify = helo" does it there and then.
149 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
151 TK/03 Fix log output including CR from clamd.
153 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
154 exist provoked a memory error which could cause a segfault.
156 PH/15 Installed PCRE 6.2
158 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
160 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
161 of the problem. Specifically, suggested +O2 rather than +O1 for the
164 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
166 PH/20 If a delivery was routed to a non-standard port by means of an SRV
167 record, the port was not correctly logged when the outgoing_port log
168 selector was set (it logged the transort's default port).
170 PH/21 Added support for host-specific ports to manualroute, queryprogram,
171 fallback_hosts, and "hosts" in the smtp transport.
173 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
174 host errors such as "Connection refused".
176 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
177 authentication with radiusclient 0.4.9:
179 - Error returned from rc_read_config was caught wrongly
180 - Username/password not passed on to radius server due to wrong length.
182 The presumption is that some radiusclient API changes for 4.51/PH/17
183 were not taken care of correctly. The code is still untested by me (my
184 Linux distribution still has 0.3.2 of radiusclient), but it was
185 contributed by a Radius user.
187 PH/24 When doing a callout, the value of $domain wasn't set correctly when
188 expanding the "port" option of the smtp transport.
190 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
191 while reading a MIME header. Thanks to Tom Hughes for a patch.
193 PH/24 Include config.h inside local_scan.h so that configuration settings are
196 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
197 that in an ACL for RCPT (for example), you can examine exactly what was
200 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
201 commands, but it was not correctly comparing the address with the actual
202 client host address. Thus, it would show the EHLO address in Received:
203 header lines when this was not necessary.
205 PH/27 Added the % operator to ${eval:}.
207 PH/28 Exim tries to create and chdir to its spool directory when it starts;
208 it should be ignoring failures (because with -C, for example, it has lost
209 privilege). It wasn't ignoring creation failures other than "already
212 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
215 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
216 because an input line was too long, either on its own, or by virtue of
217 too many continuations, the temporary file was not being removed, and the
218 return code was incorrect.
220 PH/31 Missing "BOOL" in function definition in filtertest.c.
222 PH/32 Applied Sieve patches from the maintainer.
224 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
226 PH/33 Added "verify = not_blind".
228 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
229 Local/Makefile (with some defaults set). These are used in built scripts
230 such as exicyclog, but they have never been used in the exim_install
231 script (though there are many overriding facilities there). I have
232 arranged that the exim_install script now takes note of these two
235 PH/35 Installed configuration files for Dragonfly.
237 PH/36 When a locally submitted message by a trusted user did not contain a
238 From: header, and the sender address was obtained from -f or from an SMTP
239 MAIL command, and the trusted user did not use -F to supply a sender
240 name, $originator_name was incorrectly used when constructing a From:
241 header. Furthermore, $originator_name was used for submission mode
242 messages from external hosts without From: headers in a similar way,
243 which is clearly wrong.
245 PH/37 Added control=suppress_local_fixups.
247 PH/38 When log_selector = +received_sender was set, and the addition of the
248 sender made the log line's construction buffer exactly full, or one byte
249 less than full, an overflow happened when the terminating "\n" was
252 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
253 when the result of a list match is failure because a DNS lookup failed.
255 PH/40 RM_COMMAND is now used in the building process.
257 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
258 the "build-* directories that it finds.
260 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
261 address in a domain literal was a prefix of an interface address.
263 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
264 when verifying a sender address, unless rewrite_headers is false.
266 PH/44 Wrote a long comment about why errors_to addresses are verified as
267 recipients, not senders.
269 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
270 the ratelimit ACL was added.
272 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
274 PH/46 Added extra information about PostgreSQL errors to the error string.
276 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
277 avoiding the need to install special zones in a real server. This is
278 backwards compatible; if it can't find the fake resolver, it drops back.
279 Thus, both old and new test suites can be run.
281 TF/02 Added util/ratelimit.pl
283 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
284 clock is set back in time.
286 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
287 Candler <B.Candler@pobox.com>.
289 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
290 OK for addresses that are the result of redirections.
292 PH/48 A number of further additions for the benefit of the new test suite,
293 including a fake gethostbyname() that interfaces to the fake DNS resolver
296 TF/06 The fix for widen_domains has also been applied to qualify_single and
297 search_parents which are the other dnslookup options that can cause
300 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
303 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
305 TF/07 Exim produced the error message "an SRV record indicated no SMTP
306 service" if it encountered an MX record with an empty target hostname.
307 The message is now "an MX or SRV record indicated no SMTP service".
309 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
310 if the DNS of the sending site is misconfigured. This is quite a
311 common situation. This change restores the behaviour of treating a
312 helo verification defer as a failure.
314 PH/51 If self=fail was set on a router, the bounce message did not include the
315 actual error message.
321 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
323 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
324 command itself ended in error, the underlying message about the transport
325 filter timeout was being overwritten with the pipe command error. Now the
326 underlying error message should be appended to the second error message.
328 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
330 PH/02 Reduce the amount of output that "make" produces by default. Full output
331 can still be requested.
333 PH/03 The warning log line about a condition test deferring for a "warn" verb
334 was being output only once per connection, rather than after each
335 occurrence (because it was using the same function as for successful
336 "warn" verbs). This seems wrong, so I have changed it.
338 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
339 it should not have, which might have caused a crash in the right
340 circumstances, but probably never did.
342 PH/04 Installed a modified version of Tony Finch's patch to make submission
343 mode fix the return path as well as the Sender: header line, and to
344 add a /name= option so that you can make the user's friendly name appear
347 TF/03 Added the control = fakedefer ACL modifier.
349 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
350 Mark Lowes for thorough testing.
352 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
354 TK/03 Merged latest SRS patch from Miles Wilton.
356 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
357 with the definition in sysexits.h (which is #included earlier).
358 Fortunately, Exim does not actually use EX_OK. The code used to try to
359 preserve the sysexits.h value, by assumimg that macro definitions were
360 scanned for macro replacements. I have been disabused of this notion,
361 so now the code just undefines EX_OK before #including unistd.h.
363 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
364 in the smtp transport. When a block could not be written in a single
365 write() function, the timeout was being re-applied to each part-write.
366 This seems wrong - if the receiver was accepting one byte at a time it
367 would take for ever. The timeout is now adjusted when this happens. It
368 doesn't have to be particularly precise.
370 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
371 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
373 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
374 without a domain if the check to <postmaster@domain> fails.
376 SC/01 Eximstats: added -xls and the ability to specify output files
377 (patch written by Frank Heydlauf).
379 SC/02 Eximstats: use FileHandles for outputing results.
381 SC/03 Eximstats: allow any combination of xls, txt, and html output.
383 SC/04 Eximstats: fixed display of large numbers with -nvr option
385 SC/05 Eximstats: fixed merging of reports with empty tables.
387 SC/06 Eximstats: added the -include_original_destination flag
389 SC/07 Eximstats: removed tabs and trailing whitespace.
391 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
393 TK/06 MBOX spool code: Add real "From " MBOX separator line
394 so the .eml file is really in mbox format (even though
395 most programs do not really care). Patch from Alex Miller.
397 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
398 The latter is generated from $received_to and is only set if the
399 message has one envelope recipient. SA can use these headers,
400 obviously out-of-the-box. Patch from Alex Miller.
402 PH/08 The ${def test on a variable was returning false if the variable's
403 value was "0", contrary to what the specification has always said!
404 The result should be true unless the variable is empty.
406 PH/09 The syntax error of a character other than { following "${if
407 def:variable_name" (after optional whitespace) was not being diagnosed.
408 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
409 accidental colon was present, for example, could give incorrect results.
411 PH/10 Tidied the code in a number of places where the st_size field of a stat()
412 result is used (not including appendfile, where other changes are about
415 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
416 This involved changing a lot of size variables from int to off_t. It
417 should work with maildirs and everything.
419 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
420 spamd dying while we are connected to it.
422 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
425 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
426 function for BDB 4.3.
428 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
430 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
431 the sources for PCRE have been re-arranged and more files are now
434 PH/15 The code I had for printing potentially long long variables in PH/11
435 above was not the best (it lost precision). The length of off_t variables
436 is now inspected at build time, and an appropriate printing format (%ld
437 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
438 to be "long long int" or "long int". This is needed for the internal
439 formatting function string_vformat().
441 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
442 the configuration file to be ":syslog", then the script "guesses" where
443 the logs files are, rather than using the compiled in default. In our
444 case the guess is not the same as the compiled default, so the script
445 suddenly stopped working when I started to use syslog. The patch checks
446 to see if log_file_path is "". If so, it attempts to read it from exim
447 with no configuration file to get the compiled in version, before it
448 falls back to the previous guessing code."
450 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
451 implementing BATV in an Exim configuration. See NewStuff for the gory
454 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
455 Makefile that are specific to HP-UX.
457 PH/18 If the "use_postmaster" option was set for a recipient callout together
458 with the "random" option, the postmaster address was used as the MAIL
459 FROM address for the random test, but not for the subsequent recipient
460 test. It is now used for both.
462 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
463 patch removes a few documentation additions to RFC 3028, because the
464 latest draft now contains them. It adds the new en;ascii-case comparator
465 and a new error check for 8bit text in MIME parts. Comparator and
466 require names are now matched exactly. I enabled the subaddress
467 extension, but it is not well tested yet (read: it works for me)."
469 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
470 rework some of the code of TK/09 above to avoid the hardwired use of
471 "%lld" and "long long". Replaced the call to snprintf() with a call to
474 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
475 records point to non-existent hosts", "retry timeout exceeded", and
476 "retry time not reached for any host after a long failure period".
478 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
479 experimental DomainKeys support:
481 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
482 (2) On an error such as an illegally used "control", the wrong name for
483 the control was given.
485 These problems did NOT occur unless DomainKeys support was compiled.
487 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
489 PH/24 Added ${if match_ip condition.
491 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
492 databases so that it will be absolutely obvious if a crash occurs in the
493 DB library. This is a regular occurrence (often caused by mis-matched
496 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
497 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
498 picked up on a user's system that detects such things. There doesn't seem
499 to be a gcc warning option for this - only an attribute that has to be
500 put on the function's prototype. It seems that in Fedora Core 4 they have
501 set this on a number of new functions. No doubt there will be more in due
504 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
505 specify a transport. However, if an address that was verified by such a
506 router was the subject of a callout, Exim crashed because it tried to
507 read the rcpt_include_affixes from the non-existent transport. Now it
508 just assumes that the setting of that option is false. This bug was
509 introduced by 4.51/PH/31.
511 PH/28 Changed -d+all to exclude +memory, because that information is very
512 rarely of interest, but it makes the output a lot bigger. People tend to
513 do -d+all out of habit.
515 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
516 code in os-type was giving problems when libc.so lives in lib64, like on
519 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
520 aren't the modern standard, and it seems that some systems' include files
521 don't always have them. Exim was already checking for some of the newer
522 ones like T_AAAA, and defining it itself. I've added checks for all the
523 record types that Exim uses.
525 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
526 not automatically generating a new one, as it is supposed to. This
527 prevented TLS from working. If the file did exist, but contained invalid
528 data, a new version was generated, as expected. It was only the case of a
529 non-existent file that was broken.
531 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
532 with a change in libdomainkeys > 0.64.
534 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
535 from DNS. If the selector record carries the flag, it now has
536 precedence over the domain-wide flag.
538 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
540 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
541 the use of a port name that isn't defined in /etc/services) occurred, the
542 message was deferred as in a normal delivery, and thus remained on the
543 spool, instead of being failed because of the mua_wrapper setting. This
544 is now fixed, and I tidied up some of the mua_wrapper messages at the
547 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
548 the messages in a hash of arrays rather than using individual hashes.
549 This is a bit cleaner and results in dramatic memory savings, albeit
550 at a slight CPU cost.
552 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
553 as requested by Marc Sherman.
555 SC/10 Eximstats: added histograms for user specified patterns as requested
558 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
560 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
561 fopen() in the content-scanning modules that did not already have it.
567 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
568 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
570 TK/02 Fix ACL "control" statement not being available in MIME ACL.
572 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
574 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
575 to test Sieve filters that use "vacation".
577 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
578 that changes the way the GnuTLS parameters are stored in the cache file.
579 The new format can be generated externally. For backward compatibility,
580 if the data in the cache doesn't make sense, Exim assumes it has read an
581 old-format file, and it generates new data and writes a new file. This
582 means that you can't go back to an older release without removing the
585 PH/03 A redirect router that has both "unseen" and "one_time" set does not
586 work if there are any delivery delays because "one_time" forces the
587 parent to be marked "delivered", so its unseen clone is never tried
588 again. For this reason, Exim now forbids the simultaneous setting of
591 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
592 redirected to themselves ("homonym" addresses). Read the long ChangeLog
593 entry if you want to know the details. The fix, however, neglected to
594 consider the case when local delivery batching is involved. The test for
595 "previously delivered" was not happening when checking to see if an
596 address could be batched with a previous (undelivered) one; under
597 certain circumstances this could lead to multiple deliveries to the same
600 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
601 in its include files, and this causes problems building Exim.
603 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
604 header_syntax) but Exim was just ignoring anything given after a slash.
605 In particular, this caused confusion with an attempt to use "verify =
606 reverse_host_lookup/defer_ok". An error is now given when options are
607 supplied for verify items that do not have them. (Maybe reverse_host_
608 lookup should have a defer_ok option, but that's a different point.)
610 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
611 defined by RFC 821) to 2048, because there were problems with some AUTH
612 commands, and RFC 1869 says the size should be increased for extended
613 SMTP commands that take arguments.
615 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
618 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
619 "unknown" error; now it says that the functionality isn't in the binary.
621 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
622 an address' error message when a string expansion fails (syntax or
623 whatever). Otherwise the password may appear in the log. Following change
624 PH/42 below, there is no longer a chance of it appearing in a bounce
627 PH/11 Installed exipick version 20050225.0 from John Jetmore.
629 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
630 of its addresses was ever tried. (Bugzilla bug #2.)
632 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
633 the result incorrectly in the debug output. (It correctly added a newline
634 to what was transported.)
636 TF/01 Added $received_time.
638 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
639 commented out examples of how to interface to a virus scanner and to
640 SpamAssassin. Also added commented examples of av_scanner and
641 spamd_address settings.
643 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
644 and controls are allowed in which ACLs. There were a couple of minor
645 errors. Some of the entries in the conditions table (which is a table of
646 where they are NOT allowed) were getting very unwieldy; rewrote them as a
647 negation of where the condition IS allowed.
649 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
651 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
652 header file does not have a version number, so I've had to invent a new
653 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
654 API. The code is untested by me (my Linux distribution still has 0.3.2 of
655 radiusclient), but it was contributed by a Radius user.
657 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
658 files or directories, for OpenSSL.
660 PH/19 When an Exim process that is running as root has to create an Exim log
661 file, it does so in a subprocess that runs as exim:exim so as to get the
662 ownership right at creation (otherwise, other Exim processes might see
663 the file with the wrong ownership). There was no test for failure of this
664 fork() call, which would lead to the process getting stuck as it waited
665 for a non-existent subprocess. Forks do occasionally fail when resources
666 run out. I reviewed all the other calls to fork(); they all seem to check
669 PH/20 When checking for unexpected SMTP input at connect time (before writing
670 the banner), Exim was not dealing correctly with a non-positive return
671 from the read() function. If the client had disconnected by this time,
672 the result was a log entry for a synchronization error with an empty
673 string after "input=" when read() returned zero. If read() returned -1
674 (an event I could not check), uninitialized data bytes were printed.
675 There were reports of junk text (parts of files, etc) appearing after
678 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
680 PH/22 Added support for macro redefinition, and (re)definition in between
681 driver and ACL definitions.
683 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
684 forgetting to use the resulting value; it was using the unexpanded value.
686 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
687 hadn't been configured. The fix is from Juergen Kreileder, who
688 understands it better than I do:
690 "Here's what I see happening with three configured cyrus_sasl
691 authenticators configured (plain, login, cram-md5):
693 On startup auth_cyrus_sasl_init() gets called for each of these.
694 This means three calls to sasl_listmech() without a specified mech_list.
695 => SASL tests which mechs of all available mechs actually work
696 => three warnings about OTP not working
697 => the returned list contains: plain, login, cram-md5, digest-md5, ...
699 With the patch, sasl_listmech() also gets called three times. But now
700 SASL's mech_list option is set to the server_mech specified in the the
701 authenticator. Or in other words, the answer from sasl_listmech()
702 gets limited to just the mech you're testing for (which is different
704 => the return list contains just 'plain' or 'login', 'cram-md5' or
705 nothing depending on the value of ob->server_mech.
707 I've just tested the patch: Authentication still works fine,
708 unavailable mechs specified in the exim configuration are still
709 caught, and the auth.log warnings about OTP are gone."
711 PH/25 When debugging is enabled, the contents of the command line are added
712 to the debugging output, even when log_selector=+arguments is not
715 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
716 answer is "GNU", and only if the return is "GNU/something" is the answer
719 PH/27 $acl_verify_message is now set immediately after the failure of a
720 verification in an ACL, and so is available in subsequent modifiers. In
721 particular, the message can be preserved by coding like this:
723 warn !verify = sender
724 set acl_m0 = $acl_verify_message
726 Previously, $acl_verify_message was set only while expanding "message"
727 and "log_message" when a very denied access.
729 PH/28 Modified OS/os.c-Linux with
731 -#ifndef OS_LOAD_AVERAGE
732 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
734 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
735 nomenclature these days.)
737 PH/29 Installed patch from the Sieve maintainer that adds the options
738 sieve_useraddress and sieve_subaddress to the redirect router.
740 PH/30 In these circumstances:
741 . Two addresses routed to the same list of hosts;
742 . First host does not offer TLS;
743 . First host accepts first address;
744 . First host gives temporary error to second address;
745 . Second host offers TLS and a TLS session is established;
746 . Second host accepts second address.
747 Exim incorrectly logged both deliveries with the TLS parameters (cipher
748 and peerdn, if requested) that were in fact used only for the second
751 PH/31 When doing a callout as part of verifying an address, Exim was not paying
752 attention to any local part prefix or suffix that was matched by the
753 router that accepted the address. It now behaves in the same way as it
754 does for delivery: the affixes are removed from the local part unless
755 rcpt_include_affixes is set on the transport.
757 PH/32 Add the sender address, as F=<...>, to the log line when logging a
758 timeout during the DATA phase of an incoming message.
760 PH/33 Sieve envelope tests were broken for match types other than :is. I have
761 applied a patch sanctioned by the Sieve maintainer.
763 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
764 the uid or gid is negative. A case of a negative gid caused this to be
765 noticed. The fix allows for either to be negative.
767 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
768 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
771 PH/36 The change PH/12 above was broken. Fixed it.
773 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
774 the grounds that routing the same address twice would always produce the
775 same answer. This might have been true once, but it is certainly no
776 longer true now. Routing a child address may depend on the previous
777 routing that produced that child. Some complicated redirection strategies
778 went wrong when messages had multiple recipients, and made Exim's
779 behaviour dependent on the order in which the addresses were given.
781 I have moved the duplicate checking until after the routing is complete.
782 Exim scans the addresses that are assigned to local and remote
783 transports, and removes any duplicates. This means that more work will be
784 done, as duplicates will always all be routed, but duplicates are
785 presumably rare, so I don't expect this is of any significance.
787 For deliveries to pipes, files, and autoreplies, the duplicate checking
788 still happens during the routing process, since they are not going to be
791 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
792 It corrects a timeout issue with spamd. This is Ian's comment: "The
793 background is that sometimes spamd either never reads data from a
794 connection it has accepted, or it never writes response data. The exiscan
795 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
796 blindly assumes that writes won't block so it may never time out."
798 PH/39 Allow G after quota size as well as K and M.
800 PH/40 The value set for $authenticated_id in an authenticator may not contain
801 binary zeroes or newlines because the value is written to log lines and
802 to spool files. There was no check on this. Now the value is run through
803 the string_printing() function so that such characters are converted to
804 printable escape sequences.
806 PH/41 $message_linecount is a new variable that contains the total number of
807 lines in the message. Compare $body_linecount, which is the count for the
810 PH/42 Exim no longer gives details of delivery errors for specific addresses in
811 bounce and delay warning messages, except in certain special cases, which
814 (a) An SMTP error message from a remote host;
815 (b) A message specified in a :fail: redirection;
816 (c) A message specified in a "fail" command in a system filter;
817 (d) A message specified in a FAIL return from the queryprogram router;
818 (e) A message specified by the cannot_route_message router option.
820 In these cases only, Exim does include the error details in bounce and
821 warning messages. There are also a few cases where bland messages such
822 as "unrouteable address" or "local delivery error" are given.
824 PH/43 $value is now also set for the "else" part of a ${run expansion.
826 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
827 being worked on, but at least Exim now implements the latest version to
830 PH/45 In a pipe transport, although a timeout while waiting for the pipe
831 process to complete was treated as a delivery failure, a timeout while
832 writing the message to the pipe was logged, but erroneously treated as a
833 successful delivery. Such timeouts include transport filter timeouts. For
834 consistency with the overall process timeout, these timeouts are now
835 treated as errors, giving rise to delivery failures by default. However,
836 there is now a new Boolean option for the pipe transport called
837 timeout_defer, which, if set TRUE, converts the failures into defers for
838 both kinds of timeout. A transport filter timeout is now identified in
841 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
842 systems where "make" and "gmake" are different, calling "gmake" at top
843 level broke things. I've arranged for the value of $(MAKE) to be passed
844 from the Makefile to this script so that it can call the same version of
848 A note about Exim versions 4.44 and 4.50
849 ----------------------------------------
851 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
852 changes of various kinds. As a consequence, a big documentation update was
853 needed. This delayed the release for rather longer than seemed good, especially
854 in the light of a couple of (minor) security issues. Therefore, the changes
855 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
856 release. So 4.44 and 4.50 are in effect two different branches that both start
859 I have left the 4.50 change log unchanged; it contains all the changes since
860 4.43. The change log for 4.44 is below; many of its items are identical to
861 those for 4.50. This seems to be the most sensible way to preserve the
862 historical information.
868 1. Minor wording change to the doc/README.SIEVE file.
870 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
871 computation of the current number of files was incorrect.
873 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
874 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
877 4. Give more explanation in the error message when the command for a transport
878 filter fails to execute.
880 5. There are several places where Exim runs a non-Exim command in a
881 subprocess. The SIGUSR1 signal should be disabled for these processes. This
882 was being done only for the command run by the queryprogram router. It is
883 now done for all such subprocesses. The other cases are: ${run, transport
884 filters, and the commands run by the lmtp and pipe transports.
886 6. Added CONFIGURE_GROUP build-time option.
888 7. Some older OS have a limit of 256 on the maximum number of file
889 descriptors. Exim was using setrlimit() to set 1000 as a large value
890 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
891 systems. I've change it so that if it can't get 1000, it tries for 256.
893 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
894 was an oversight, and furthermore, ever since the addition of extra
895 controls (e.g. 4.43/32), the checks on when to allow different forms of
896 "control" were broken. There should now be diagnostics for all cases when a
897 control that does not make sense is encountered.
899 9. Added the /retain_sender option to "control=submission".
901 10. $recipients is now available in the predata ACL (oversight).
903 11. Tidy the search cache before the fork to do a delivery from a message
904 received from the command line. Otherwise the child will trigger a lookup
905 failure and thereby defer the delivery if it tries to use (for example) a
906 cached ldap connection that the parent has called unbind on.
908 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
909 of $address_data from the recipient verification was clobbered by the
912 13. The value of address_data from a sender verification is now available in
913 $sender_address_data in subsequent conditions in the ACL statement.
915 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
917 15. Added a new option "connect=<time>" to callout options, to set a different
920 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
921 was its contents. (It was OK if the option was not defined at all.)
923 17. A "Completed" log line is now written for messages that are removed from
924 the spool by the -Mrm option.
926 18. New variables $sender_verify_failure and $recipient_verify_failure contain
927 information about exactly what failed.
929 19. Added -dd to debug only the daemon process.
931 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
932 handles timeouts, both on the server side and network timeouts. Renamed the
933 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
935 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
936 It is now set to "smtps".
938 22. $host_address is now set to the target address during the checking of
941 23. When checking ignore_target_hosts for an ipliteral router, no host name was
942 being passed; this would have caused $sender_host_name to have been used if
943 matching the list had actually called for a host name (not very likely,
944 since this list is usually IP addresses). A host name is now passed as
947 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
948 code that specifies a non-restarting handler (typically sigaction() in
949 modern systems) in an attempt to fix a rare and obscure crash bug.
951 25. Narrowed the window for a race in the daemon that could cause it to ignore
952 SIGCHLD signals. This is not a major problem, because they are used only to
953 wake it up if nothing else does.
955 26. A malformed maildirsize file could cause Exim to calculate negative values
956 for the mailbox size or file count. Odd effects could occur as a result.
957 The maildirsize information is now recalculated if the size or filecount
960 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
961 support for a long time. Removed HAVE_SYS_VFS_H.
963 28. Installed the latest version of exipick from John Jetmore.
965 29. In an address list, if the pattern was not a regular expression, an empty
966 subject address (from a bounce message) matched only if the pattern was an
967 empty string. Non-empty patterns were not even tested. This was the wrong
968 because it is perfectly reasonable to use an empty address as part of a
969 database query. An empty address is now tested by patterns that are
970 lookups. However, all the other forms of pattern expect the subject to
971 contain a local part and a domain, and therefore, for them, an empty
972 address still always fails if the pattern is not itself empty.
974 30. Exim went into a mad DNS loop when attempting to do a callout where the
975 host was specified on an smtp transport, and looking it up yielded more
978 31. Re-factored the code for checking spool and log partition space into a
979 function that finds that data and another that does the check. The former
980 is then used to implement four new variables: $spool_space, $log_space,
981 $spool_inodes, and $log_inodes.
983 32. The RFC2047 encoding function was originally intended for short strings
984 such as real names; it was not keeping to the 75-character limit for
985 encoded words that the RFC imposes. It now respects the limit, and
986 generates multiple encoded words if necessary. To be on the safe side, I
987 have increased the buffer size for the ${rfc2047: expansion operator from
990 33. It is now permitted to omit both strings after an "if" condition; if the
991 condition is true, the result is "true". As before, when the second string
992 is omitted, a false condition yields an empty string. This makes it less
993 cumbersome to write custom ACL and router conditions.
995 34. Failure to deliver a bounce message always caused it to be frozen, even if
996 there was an errors_to setting on the router. The errors_to setting is now
999 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1000 canonical form (fully expanded) before being placed in
1001 $sender_host_address.
1003 36. The table in the code that translates DNS record types into text (T_A to
1004 "A" for instance) was missing entries for NS and CNAME. It is just possible
1005 that this could have caused confusion if both these types were looked up
1006 for the same domain, because the text type is used as part of Exim's
1007 per-process caching. But the chance of anyone hitting this buglet seems
1010 37. The dnsdb lookup has been extended in a number of ways.
1012 (1) There is a new type, "zns", which walks up the domain tree until it
1013 finds some nameserver records. It should be used with care.
1015 (2) There is a new type, "mxh", which is like "mx" except that it returns
1016 just the host names, not the priorities.
1018 (3) It is now possible to give a list of domains (or IP addresses) to be
1019 looked up. The behaviour when one of the lookups defers can be
1020 controlled by a keyword.
1022 (4) It is now possible to specify the separator character for use when
1023 multiple records are returned.
1025 38. The dnslists ACL condition has been extended: it is now possible to supply
1026 a list of IP addresses and/or domains to be looked up in a particular DNS
1029 39. Added log_selector=+queue_time_overall.
1031 40. When running the queue in the test harness, wait just a tad after forking a
1032 delivery process, to get repeatability of debugging output.
1034 41. Include certificate and key file names in error message when GnuTLS fails
1035 to set them up, because the GnuTLS error message doesn't include the name
1036 of the failing file when there is a problem reading it.
1038 42. Allow both -bf and -bF in the same test run.
1040 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1042 44. The "Exiscan patch" is now merged into the mainline Exim source.
1044 45. Sometimes the final signoff response after QUIT could fail to get
1045 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1046 < 0 before doing a fflush(). This bug looks as though it goes back to the
1047 introduction of TLS in release 3.20, but "sometimes" must have been rare
1048 because the tests only now provoked it.
1050 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1051 (this can affect the format of dates).
1053 47. exim_tidydb, when checking for the continued existence of a message for
1054 which it has found a message-specific retry record, was not finding
1055 messages that were in split spool directories. Consequently, it was
1056 deleting retry records that should have stayed in existence.
1058 48. Steve fixed some bugs in eximstats.
1060 49. The SPA authentication driver was not abandoning authentication and moving
1061 on to the next authenticator when an expansion was forced to fail,
1062 contradicting the general specification for all authenticators. Instead it
1063 was generating a temporary error. It now behaves as specified.
1065 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1066 (the order specifies the preference for clients). The order is now AES256,
1067 AES128, 3DES, ARCFOUR128.
1069 51. Small patch to Sieve code - explicitly set From: when generating an
1072 52. Exim crashed if a remote delivery caused a very long error message to be
1073 recorded - for instance if somebody sent an entire SpamAssassin report back
1074 as a large number of 550 error lines. This bug was coincidentally fixed by
1075 increasing the size of one of Exim's internal buffers (big_buffer) that
1076 happened as part of the Exiscan merge. However, to be on the safe side, I
1077 have made the code more robust (and fixed the comments that describe what
1080 53. Now that there can be additional text after "Completed" in log lines (if
1081 the queue_time_overall log selector is set), a one-byte patch to exigrep
1082 was needed to allow it to recognize "Completed" as not the last thing in
1085 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1086 patch that reportedly fixes this has been added. I am not expert enough to
1087 create a test for it. This is what the patch creator wrote:
1089 "I found a little strange behaviour of ldap code when working with
1090 Windows 2003 AD Domain, where users was placed in more than one
1091 Organization Units. When I tried to give exim partial DN, the exit code
1092 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1093 But simultaneously result of request was absolutely normal ldap result,
1094 so I produce this patch..."
1096 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1097 REFERENCE, so I have modified the code to exclude the patch when that macro
1100 55. Some experimental protocols are using DNS PTR records for new purposes. The
1101 keys for these records are domain names, not reversed IP addresses. The
1102 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1103 leaves it alone. Component reversal etc. now happens only for IP addresses.
1106 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1108 57. Double the size of the debug message buffer (to 2048) so that more of very
1109 long debug lines gets shown.
1111 58. The exicyclog utility now does better if the number of log files to keep
1112 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1114 59. Two changes related to the smtp_active_hostname option:
1116 (1) $smtp_active_hostname is now available as a variable.
1117 (2) The default for smtp_banner uses $smtp_active_hostname instead
1118 of $primary_hostname.
1120 60. The host_aton() function is supposed to be passed a string that is known
1121 to be a valid IP address. However, in the case of IPv6 addresses, it was
1122 not checking this. This is a hostage to fortune. Exim now panics and dies
1123 if the condition is not met. A case was found where this could be provoked
1124 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1125 components; fortuitously, this particular loophole had already been fixed
1126 by change 4.50/55 above.
1128 If there are any other similar loopholes, the new check in host_aton()
1129 itself should stop them being exploited. The report I received stated that
1130 data on the command line could provoke the exploit when Exim was running as
1131 exim, but did not say which command line option was involved. All I could
1132 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1133 running as the user.
1136 61. There was a buffer overflow vulnerability in the SPA authentication code
1137 (which came originally from the Samba project). I have added a test to the
1138 spa_base64_to_bits() function which I hope fixes it.
1141 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1142 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1144 63. The daemon start-up calls getloadavg() while still root for those OS that
1145 need the first call to be done as root, but it missed one case: when
1146 deliver_queue_load_max is set with deliver_drop_privilege. This is
1147 necessary for the benefit of the queue runner, because there is no re-exec
1148 when deliver_drop_privilege is set.
1150 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1151 This has been fixed.
1153 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1154 was in use, was not putting the data itself into the right store pool;
1155 consequently, it could be overwritten for a subsequent message in the same
1156 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1159 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1161 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1162 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1163 the function were treating the return as a boolean value, which happened to
1164 work because 0=false and not-0=true, but is not correct code.
1166 68. The host_aton() function was not handling scoped IPv6 addresses (those
1167 with, for example, "%eth0" on the end) correctly.
1169 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1170 negated items (that is, ~something) in unsigned ints. Some compilers
1171 apparently mutter when there is no cast.
1173 70. If an address verification called from an ACL failed, and did not produce a
1174 user-specific message (i.e. there was only a "system" message), nothing was
1175 put in $acl_verify_message. In this situation, it now puts the system
1178 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1179 session; change 4.31/43 added the unwanted input to the log line - except
1180 that it did not do this in the start of session case. It now does.
1182 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1183 This is wrong and can cause the other end to generate a synchronization
1184 error if it is another Exim or anything else that does the synchronization
1185 check. A QUIT command is no longer sent after a timeout.
1187 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1188 during host lookups.
1190 74. The defer_ok option of callout verification was not working if it was used
1191 when verifying addresses in header lines, that is, for this case:
1193 verify = header_sender/callout=defer_ok
1195 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1196 those file descriptors could be used for SMTP connections. If anything
1197 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1198 could be sent to the SMTP client, causing chaos. The daemon now opens
1199 stdin, stdout, and stderr to /dev/null when it puts itself into the
1202 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1203 log by default. The user can override this with suitable Perl magic.
1205 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1206 the log message when discard triggers, was not working for the DATA ACL or
1207 for the non-SMTP ACL.
1209 78. Error message wording change in sieve.c.
1211 79. If smtp_accept_max_per_host was set, the number of connections could be
1212 restricted to fewer than expected, because the daemon was trying to set up
1213 a new connection before checking whether the processes handling previous
1214 connections had finished. The check for completed processes is now done
1215 earlier. On busy systems, this bug wouldn't be noticed because something
1216 else would have woken the daemon, and it would have reaped the completed
1219 80. If a message was submitted locally by a user whose login name contained one
1220 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1221 It caused a spool format error. I have fixed the spool reading code. A
1222 related problem was that the "from" clause in the Received: line became
1223 illegal because of the space(s). It is now covered by ${quote_local_part.
1225 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1228 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1230 83. Patch from Sieve maintainer to fix unterminated string problem in
1231 "vacation" handling.
1233 84. Some minor changes to the Linux configuration files to help with other
1234 OS variants using glibc.
1236 85. One more patch for Sieve to update vacation handling to latest spec.
1239 ----------------------------------------------------
1240 See the note above about the 4.44 and 4.50 releases.
1241 ----------------------------------------------------
1247 1. Change 4.43/35 introduced a bug that caused file counts to be
1248 incorrectly computed when quota_filecount was set in an appendfile
1251 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1252 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1255 3. Give more explanation in the error message when the command for a transport
1256 filter fails to execute.
1258 4. There are several places where Exim runs a non-Exim command in a
1259 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1260 was being done only for the command run by the queryprogram router. It is
1261 now done for all such subprocesses. The other cases are: ${run, transport
1262 filters, and the commands run by the lmtp and pipe transports.
1264 5. Some older OS have a limit of 256 on the maximum number of file
1265 descriptors. Exim was using setrlimit() to set 1000 as a large value
1266 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1267 systems. I've change it so that if it can't get 1000, it tries for 256.
1269 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1270 was an oversight, and furthermore, ever since the addition of extra
1271 controls (e.g. 4.43/32), the checks on when to allow different forms of
1272 "control" were broken. There should now be diagnostics for all cases when a
1273 control that does not make sense is encountered.
1275 7. $recipients is now available in the predata ACL (oversight).
1277 8. Tidy the search cache before the fork to do a delivery from a message
1278 received from the command line. Otherwise the child will trigger a lookup
1279 failure and thereby defer the delivery if it tries to use (for example) a
1280 cached ldap connection that the parent has called unbind on.
1282 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1283 of $address_data from the recipient verification was clobbered by the
1284 sender verification.
1286 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1287 was its contents. (It was OK if the option was not defined at all.)
1289 11. A "Completed" log line is now written for messages that are removed from
1290 the spool by the -Mrm option.
1292 12. $host_address is now set to the target address during the checking of
1293 ignore_target_hosts.
1295 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1296 being passed; this would have caused $sender_host_name to have been used if
1297 matching the list had actually called for a host name (not very likely,
1298 since this list is usually IP addresses). A host name is now passed as
1301 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1302 code that specifies a non-restarting handler (typically sigaction() in
1303 modern systems) in an attempt to fix a rare and obscure crash bug.
1305 15. Narrowed the window for a race in the daemon that could cause it to ignore
1306 SIGCHLD signals. This is not a major problem, because they are used only to
1307 wake it up if nothing else does.
1309 16. A malformed maildirsize file could cause Exim to calculate negative values
1310 for the mailbox size or file count. Odd effects could occur as a result.
1311 The maildirsize information is now recalculated if the size or filecount
1314 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1315 support for a long time. Removed HAVE_SYS_VFS_H.
1317 18. Updated exipick to current release from John Jetmore.
1319 19. Allow an empty sender to be matched against a lookup in an address list.
1320 Previously the only cases considered were a regular expression, or an
1323 20. Exim went into a mad DNS lookup loop when doing a callout where the
1324 host was specified on the transport, if the DNS lookup yielded more than
1327 21. The RFC2047 encoding function was originally intended for short strings
1328 such as real names; it was not keeping to the 75-character limit for
1329 encoded words that the RFC imposes. It now respects the limit, and
1330 generates multiple encoded words if necessary. To be on the safe side, I
1331 have increased the buffer size for the ${rfc2047: expansion operator from
1334 22. Failure to deliver a bounce message always caused it to be frozen, even if
1335 there was an errors_to setting on the router. The errors_to setting is now
1338 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1339 canonical form (fully expanded) before being placed in
1340 $sender_host_address.
1342 24. Updated eximstats to version 1.33
1344 25. Include certificate and key file names in error message when GnuTLS fails
1345 to set them up, because the GnuTLS error message doesn't include the name
1346 of the failing file when there is a problem reading it.
1348 26. Expand error message when OpenSSL has problems setting up cert/key files.
1351 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1352 (this can affect the format of dates).
1354 28. exim_tidydb, when checking for the continued existence of a message for
1355 which it has found a message-specific retry record, was not finding
1356 messages that were in split spool directories. Consequently, it was
1357 deleting retry records that should have stayed in existence.
1359 29. eximstats updated to version 1.35
1360 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1361 1.35 - bugfix such that pie charts by volume are generated correctly
1363 30. The SPA authentication driver was not abandoning authentication and moving
1364 on to the next authenticator when an expansion was forced to fail,
1365 contradicting the general specification for all authenticators. Instead it
1366 was generating a temporary error. It now behaves as specified.
1368 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1369 (the order specifies the preference for clients). The order is now AES256,
1370 AES128, 3DES, ARCFOUR128.
1372 31. Small patch to Sieve code - explicitly set From: when generating an
1375 32. Exim crashed if a remote delivery caused a very long error message to be
1376 recorded - for instance if somebody sent an entire SpamAssassin report back
1377 as a large number of 550 error lines. This bug was coincidentally fixed by
1378 increasing the size of one of Exim's internal buffers (big_buffer) that
1379 happened as part of the Exiscan merge. However, to be on the safe side, I
1380 have made the code more robust (and fixed the comments that describe what
1383 33. Some experimental protocols are using DNS PTR records for new purposes. The
1384 keys for these records are domain names, not reversed IP addresses. The
1385 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1386 leaves it alone. Component reversal etc. now happens only for IP addresses.
1389 34. The host_aton() function is supposed to be passed a string that is known
1390 to be a valid IP address. However, in the case of IPv6 addresses, it was
1391 not checking this. This is a hostage to fortune. Exim now panics and dies
1392 if the condition is not met. A case was found where this could be provoked
1393 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1394 components; fortuitously, this particular loophole had already been fixed
1395 by change 4.50/55 or 4.44/33 above.
1397 If there are any other similar loopholes, the new check in host_aton()
1398 itself should stop them being exploited. The report I received stated that
1399 data on the command line could provoke the exploit when Exim was running as
1400 exim, but did not say which command line option was involved. All I could
1401 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1402 running as the user.
1405 35. There was a buffer overflow vulnerability in the SPA authentication code
1406 (which came originally from the Samba project). I have added a test to the
1407 spa_base64_to_bits() function which I hope fixes it.
1410 36. The daemon start-up calls getloadavg() while still root for those OS that
1411 need the first call to be done as root, but it missed one case: when
1412 deliver_queue_load_max is set with deliver_drop_privilege. This is
1413 necessary for the benefit of the queue runner, because there is no re-exec
1414 when deliver_drop_privilege is set.
1416 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1417 was in use, was not putting the data itself into the right store pool;
1418 consequently, it could be overwritten for a subsequent message in the same
1419 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1422 38. Sometimes the final signoff response after QUIT could fail to get
1423 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1424 < 0 before doing a fflush(). This bug looks as though it goes back to the
1425 introduction of TLS in release 3.20, but "sometimes" must have been rare
1426 because the tests only now provoked it.
1432 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1433 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1434 Now it returns an integer. A number of calls were still expecting a T/F
1435 return. Fortuitously, in all cases, the tests worked in OK situations,
1436 which is the norm. However, things would have gone wrong on any write
1437 failures on the smtp file descriptor. This function is used when sending
1438 messages over SMTP and also when doing verify callouts.
1440 2. When Exim is called to do synchronous delivery of a locally submitted
1441 message (the -odf or -odi options), it no longer closes stderr before doing
1444 3. Implemented the mua_wrapper option.
1446 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1448 5. Implemented the functions header_remove(), header_testname(),
1449 header_add_at_position(), and receive_remove_recipient(), and exported them
1452 6. If an ACL "warn" statement specified the addition of headers, Exim already
1453 inserted X-ACL-Warn: at the start if there was no header name. However, it
1454 was not making this test for the second and subsequent header lines if
1455 there were newlines in the string. This meant that an invalid header could
1456 be inserted if Exim was badly configured.
1458 7. Allow an ACL "warn" statement to add header lines at the start or after all
1459 the Received: headers, as well as at the end.
1461 8. Added the rcpt_4xx retry error code.
1463 9. Added postmaster_mailfrom=xxx to callout verification option.
1465 10. Added mailfrom=xxxx to the callout verification option, for verify=
1468 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1469 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1471 12. Inserted some casts to stop certain compilers warning when using pointer
1472 differences as field lengths or precisions in printf-type calls (mostly
1473 affecting debugging statements).
1475 13. Added optional readline() support for -be (dynamically loaded).
1477 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1478 same clock tick as a message's arrival, so that its received time was the
1479 same as the "first fail" time on the retry record, and that message
1480 remained on the queue past the ultimate address timeout, every queue runner
1481 would try a delivery (because it was past the ultimate address timeout) but
1482 after another failure, the ultimate address timeout, which should have then
1483 bounced the address, did not kick in. This was a "< instead of <=" error;
1484 in most cases the first failure would have been in the next clock tick
1485 after the received time, and all would be well.
1487 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1488 being recognized when the domain list was tested by the match_domain
1489 condition in an expansion string.
1491 16. Added the ${str2b64: operator.
1493 17. Exim was always calling setrlimit() to set a large limit for the number of
1494 processes, without checking whether the existing limit was already
1495 adequate. (It did check for the limit on file descriptors.) Furthermore,
1496 errors from getrlimit() and setrlimit() were being ignored. Now they are
1497 logged to the main and panic logs, but Exim does carry on, to try to do its
1498 job under whatever limits there are.
1500 18. Imported PCRE 5.0.
1502 19. Trivial typo in log message " temporarily refused connection" (the leading
1505 20. If the log selector return_path_on_delivery was set and an address was
1506 redirected to /dev/null, the delivery process crashed because it assumed
1507 that a return path would always be set for a "successful" delivery. In this
1508 case, the whole delivery is bypassed as an optimization, and therefore no
1511 21. Internal re-arrangement: the function for sending a challenge and reading
1512 a response while authentication was assuming a zero-terminated challenge
1513 string. It's now changed to take a pointer and a length, to allow for
1514 binary data in such strings.
1516 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1518 23. Exim was not respecting finduser_retries when seeking the login of the
1519 uid under which it was called; it was always trying 10 times. (The default
1520 setting of finduser_retries is zero.) Also, it was sleeping after the final
1521 failure, which is pointless.
1523 24. Implemented tls_on_connect_ports.
1525 25. Implemented acl_smtp_predata.
1527 26. If the domain in control=submission is set empty, Exim assumes that the
1528 authenticated id is a complete email address when it generates From: or
1529 Sender: header lines.
1531 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1532 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1533 chown and chgrp in /bin and hostname in /usr/bin.
1535 28. Exim was keeping the "process log" file open after each use, just as it
1536 does for the main log. This opens the possibility of it remaining open for
1537 long periods when the USR1 signal hits a daemon. Occasional processlog
1538 errors were reported, that could have been caused by this. Anyway, it seems
1539 much more sensible not to leave this file open at all, so that is what now
1542 29. The long-running daemon process does not normally write to the log once it
1543 has entered its main loop, and it closes the log before doing so. This is
1544 so that log files can straightforwardly be renamed and moved. However,
1545 there are a couple of unusual error situations where the daemon does write
1546 log entries, and I had neglected to close the log afterwards.
1548 30. The text of an SMTP error response that was received during a remote
1549 delivery was being truncated at 512 bytes. This is too short for some of
1550 the long messages that one sometimes sees. I've increased the limit to
1553 31. It is now possible to make retry rules that apply only when a message has a
1554 specific sender, in particular, an empty sender.
1556 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1557 it possible to be selective about when SMTP synchronization is enforced.
1559 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1561 32. Implemented hosts_connection_nolog.
1563 33. Added an ACL for QUIT.
1565 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1568 35. Added mailbox_size and mailbox_filecount to appendfile.
1570 36. Added control = no_multiline_responses to ACLs.
1572 37. There was a bug in the logic of the code that waits for the clock to tick
1573 in the case where the clock went backwards by a substantial amount such
1574 that the microsecond fraction of "now" was more than the microsecond
1575 fraction of "then" (but the whole seconds number was less).
1577 38. Added support for the libradius Radius client library this is found on
1578 FreeBSD (previously only the radiusclient library was supported).
1584 1. When certain lookups returned multiple values in the form name=value, the
1585 quoting of the values was not always being done properly. Specifically:
1586 (a) If the value started with a double quote, but contained no whitespace,
1588 (b) If the value contained whitespace other than a space character (i.e.
1589 tabs or newlines or carriage returns) it was not quoted.
1590 This fix has been applied to the mysql and pgsql lookups by writing a
1591 separate quoting function and calling it from the lookup code. The fix
1592 should probably also be applied to nisplus, ibase and oracle lookups, but
1593 since I cannot test any of those, I have not disturbed their existing code.
1595 2. A hit in the callout cache for a specific address caused a log line with no
1596 reason for rejecting RCPT. Now it says "Previous (cached) callout
1597 verification failure".
1599 3. There was an off-by-one bug in the queryprogram router. An over-long
1600 return line was truncated at 256 instead of 255 characters, thereby
1601 overflowing its buffer with the terminating zero. As well as fixing this, I
1602 have increased the buffer size to 1024 (and made a note to document this).
1604 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1605 when Exim is waiting for an SMTP response from a remote server, Exim
1606 restarts its select() call on the socket, thereby resetting its timeout.
1607 This is not a problem when such interrupts are rare. Somebody set up a cron
1608 job to run exiwhat every 2 minutes, which is less than the normal select()
1609 timeout (5 or 10 minutes). This meant that the select() timeout never
1610 kicked in because it was always reset. I have fixed this by comparing the
1611 time when an interrupt arrives with the time at the start of the first call
1612 to select(). If more time than the timeout has elapsed, the interrupt is
1613 treated as a timeout.
1615 5. Some internal re-factoring in preparation for the addition of Sieve
1616 extensions (by MH). In particular, the "personal" test is moved to a
1617 separate function, and given an option for scanning Cc: and Bcc: (which is
1618 not set for Exim filters).
1620 6. When Exim created an email address using the login of the caller as the
1621 local part (e.g. when creating a From: or Sender: header line), it was not
1622 quoting the local part when it contained special characters such as @.
1624 7. Installed new OpenBSD configuration files.
1626 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1627 try to make them clearer.
1629 9. Callout options, other than the timeout value, were being ignored when
1630 verifying sender addresses in header lines. For example, when using
1632 verify = header_sender/callout=no_cache
1634 the cache was (incorrectly) being used.
1636 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1637 only the Cygwin environment.
1639 11. When return_path_on_delivery was set as a log selector, if different remote
1640 addresses in the same message used different return paths and parallel
1641 remote delivery occurred, the wrong values would sometimes be logged.
1642 (Whenever a remote delivery process finished, the return path value from
1643 the most recently started remote delivery process was logged.)
1645 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1646 lines when AUTH and/or TLS are in use. This is the "received protocol"
1647 field. Exim used to use "asmtp" for authenticated SMTP, without any
1648 indication (in the protocol name) for TLS use. Now it follows the RFC and
1649 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1650 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1651 names appear in log lines as well as in Received: header lines.
1653 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1654 extensions, and comparison tests, and to fix some bugs.
1656 14. Changes to the "personal" filter test:
1658 (1) The test was buggy in that it was just doing the equivalent of
1659 "contains" tests on header lines. For example, if a user's address was
1660 anne@some.where, the "personal" test would incorrectly be true for
1662 To: susanne@some.where
1664 This test is now done by extracting each address from the header in turn,
1665 and checking the entire address. Other tests that are part of "personal"
1666 are now done using regular expressions (for example, to check local parts
1667 of addresses in From: header lines).
1669 (2) The list of non-personal local parts in From: addresses has been
1670 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1671 taken from the Sieve specification recommendations.
1673 (3) If the message contains any header line starting with "List-" it is
1674 treated as non-personal.
1676 (4) The test for "circular" in the Subject: header line has been removed
1677 because it now seems ill-conceived.
1679 15. Minor typos in src/EDITME comments corrected.
1681 16. Installed latest exipick from John Jetmore.
1683 17. If headers_add on a router specified a text string that was too long for
1684 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1685 of string_sprintf() is now avoided.
1687 18. $message_body_size was not set (it was always zero) when running the DATA
1688 ACL and the local_scan() function.
1690 19. For the "mail" command in an Exim filter, no default was being set for
1691 the once_repeat time, causing a random time value to be used if "once" was
1692 specified. (If the value happened to be <= 0, no repeat happened.) The
1693 default is now 0s, meaning "never repeat". The "vacation" command was OK
1694 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1695 (I found it when inspecting the code).
1697 20. There is now an overall timeout for performing a callout verification. It
1698 defaults to 4 times the callout timeout, which applies to individual SMTP
1699 commands during the callout. The overall timeout applies when there is more
1700 than one host that can be tried. The timeout is checked before trying the
1701 next host. This prevents very long delays if there are a large number of
1702 hosts and all are timing out (e.g. when the network connections are timing
1703 out). The value of the overall timeout can be changed by specifying an
1704 additional sub-option for "callout", called "maxwait". For example:
1706 verify = sender/callout=5s,maxwait=20s
1708 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1709 to the end before writing, but this should make it even safer).
1711 22. Exim was forgetting that it had advertised PIPELINING for the second and
1712 subsequent messages on an SMTP connection. It was also not resetting its
1713 memory on STARTTLS and an internal HELO.
1715 23. When Exim logs an SMTP synchronization error within a session, it now
1716 records whether PIPELINING has been advertised or not.
1718 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1719 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1720 rather than long int.
1722 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1724 26. Added the never_mail option to autoreply.
1730 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1731 crash if the getsockname() call failed; this can happen if a connection is
1732 closed very soon after it is established. The problem was simply in the
1733 order in which certain operations were done, causing Exim to try to write
1734 to the SMTP stream before it had set up the file descriptor. The bug has
1735 been fixed by making things happen in the correct order.
1741 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1742 before the connection was closed, thus losing the rejection response.
1744 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1745 some early Solaris releases, but causes trouble in current releases where
1746 socklen_t is defined.
1748 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1751 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1754 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1755 that releasing the top of it at the end releases what was used for sub-
1756 expansions (unless the block got too big). However, discard this block if
1757 the first thing is a variable or header, so that we can use its block when
1758 it is dynamic (useful for very large $message_headers, for example).
1760 6. Lookups now cache *every* query, not just the most recent. A new, separate
1761 store pool is used for this. It can be recovered when all lookup caches are
1762 flushed. Lookups now release memory at the end of their result strings.
1763 This has involved some general refactoring of the lookup sources.
1765 7. Some code has been added to the store_xxx() functions to reduce the amount
1766 of flapping under certain conditions.
1768 8. log_incoming_interface used to affect only the <= reception log lines. Now
1769 it causes the local interface and port to be added to several more SMTP log
1770 lines, for example "SMTP connection from", and rejection lines.
1772 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1774 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1776 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1777 could be overwritten at the end of the current message (or the start of a
1778 new message if it was set in a HELO ACL). The value is now preserved for
1779 the duration of the SMTP connection.
1781 12. If a transport had a headers_rewrite setting, and a matching header line
1782 contained an unqualified address, that address was qualified, even if it
1783 did not match any rewriting rules. The underlying bug was that the values
1784 of the flags that permit the existence of unqualified sender and recipient
1785 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1786 non-local messages, and by -bnq for local messages) were not being
1787 preserved with the message after it was received.
1789 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1790 "next input=" as part of the text comprising the host identity instead of
1791 the correct text. The code was using the same buffer for two different
1792 strings. However, depending on which order the printing function evaluated
1793 its arguments, the bug did not always show up. Under Linux, for example, my
1794 test suite worked just fine.
1796 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1797 doesn't work with some older versions of Perl. It has been changed to "my",
1798 which in any case is probably the better facility to use.
1800 15. A really picky compiler found some instances of statements for creating
1801 error messages that either had too many or two few arguments for the format
1804 16. The size of the buffer for calls to the DNS resolver has been increased
1805 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1806 for addresses that have a lot of PTR records. This alleviates a problem; it
1807 does not fully solve it.
1809 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1810 buffer now truncates the list and logs the incident, which is the same
1811 action as happens when Exim is looking up a host name and its aliases.
1812 Previously in this situation something unpredictable would happen;
1813 sometimes it was "internal error: store_reset failed".
1815 18. If a server dropped the connection unexpectedly when an Exim client was
1816 using GnuTLS and trying to read a response, the client delivery process
1817 crashed while trying to generate an error log message.
1819 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1820 string, for example:
1822 warn message = H1: something\nH2: something
1824 the text was added as a single header line from Exim's point of view
1825 though it ended up OK in the delivered message. However, searching for the
1826 second and subsequent header lines using $h_h2: did not work. This has been
1827 fixed. Similarly, if a system filter added multiple headers in this way,
1828 the routers could not see them.
1830 20. Expanded the error message when iplsearch is called with an invalid key to
1831 suggest using net-iplsearch in a host list.
1833 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1834 ACLs are no longer actually imposed (and a message to that effect is
1837 22. If a "gecos" field in a passwd entry contained escaped characters, in
1838 particular, if it contained a \" sequence, Exim got it wrong when building
1839 a From: or a Sender: header from that name. A second bug also caused
1840 incorrect handling when an unquoted " was present following a character
1841 that needed quoting.
1843 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1844 was not being matched caselessly.
1846 24. Arranged for all hyphens in the exim.8 source to be escaped with
1849 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1850 properly. Recipient callout cache records were still being keyed to include
1851 the sender, even when use_sender was set false. This led to far more
1852 callouts that were necessary. The sender is no longer included in the key
1853 when use_sender is false.
1855 26. Added "control = submission" modifier to ACLs.
1857 27. Added the ${base62d: operator to decode base 62 numbers.
1859 28. dnsdb lookups can now access SRV records.
1861 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1862 the configuration file.
1864 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1865 (-v) mode. This makes the output for a verbose queue run more intelligible.
1867 31. Added a use_postmaster feature to recipient callouts.
1869 32. Added the $body_zerocount variable, containing the number of binary zero
1870 bytes in the message body.
1872 33. The time of last modification of the "new" subdirectory is now used as the
1873 "mailbox time last read" when there is a quota error for a maildir
1876 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1878 35. Added +ignore_unknown as a special item in host lists.
1880 36. Code for decoding IPv6 addresses in host lists is now included, even if
1881 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1882 address was recognized as an IP address, but was then not correctly decoded
1883 into binary, causing unexpected and incorrect effects when compared with
1890 1. Very minor rewording of debugging text in manualroute to say "list of
1891 hosts" instead of "hostlist".
1893 2. If verify=header_syntax was set, and a header line with an unqualified
1894 address (no domain) and a large number of spaces between the end of the
1895 name and the colon was received, the reception process suffered a buffer
1896 overflow, and (when I tested it) crashed. This was caused by some obsolete
1897 code that should have been removed. The fix is to remove it!
1899 3. When running in the test harness, delay a bit after writing a bounce
1900 message to get a bit more predictability in the log output.
1902 4. Added a call to search_tidyup() just before forking a reception process. In
1903 theory, someone could use a lookup in the expansion of smtp_accept_max_
1904 per_host which, without the tidyup, could leave open a database connection.
1906 5. Added the variables $recipient_data and $sender_data which get set from a
1907 lookup success in an ACL "recipients" or "senders" condition, or a router
1908 "senders" option, similar to $domain_data and $local_part_data.
1910 6. Moved the writing of debug_print from before to after the "senders" test
1913 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1914 problems for message scanning, either using a data ACL, or using
1915 local_scan() because the Received: header was not generated till after they
1916 were called (in order to set the time as the time of reception completion).
1917 I have revised the way this works. The header is now generated after the
1918 body is received, but before the ACL or local_scan() are called. After they
1919 are run, the timestamp in the header is updated.
1925 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1926 before starting a queue runner without re-exec. This happened only when
1927 deliver_drop_privilege was set or when the Exim user was set to root. The
1928 effect of the bug was that timeouts during subsequent deliveries caused
1929 crashes instead of being properly handled. The handler is now left at its
1930 default (and expected) setting.
1932 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1933 message, again when deliver_drop_privilege is set or Exim is run as root.
1934 The bug described in (1) was not present in this case, but the tidying up
1935 of the other signals was missing. I have made the two cases consistent.
1937 3. The ignore_target_hosts setting on a manualroute router was being ignored
1938 for hosts that were looked up using the /MX notation.
1940 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1943 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1944 operated on the sender address. After changing the $sender_address to <>
1945 for the sender address verify, Exim was re-instated it as the original
1946 (before rewriting) address, but remembering that it had rewritten it, so it
1947 wasn't rewriting it again. This bug also had the effect of breaking the
1948 sender address verification caching when the sender address was rewritten.
1950 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1951 This has been changed so that if the ip literal address matches
1952 ignore_target_hosts, the router declines.
1954 7. Added expansion conditions match_domain, match_address, and match_local_
1955 part (NOT match_host).
1957 8. The placeholder for the Received: header didn't have a length field set.
1959 9. Added code to Exim itself and to exim_lock to test for a specific race
1960 condition that could lead to file corruption when using MBX delivery. The
1961 issue is with the lockfile that is created in /tmp. If this file is removed
1962 after a process has opened it but before that process has acquired a lock,
1963 there is the potential for a second process to recreate the file and also
1964 acquire a lock. This could lead to two Exim processes writing to the file
1965 at the same time. The added code performs the same test as UW imapd; it
1966 checks after acquiring the lock that its file descriptor still refers to
1967 the same named file.
1969 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1970 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1973 11. Added the smtp_active_hostname option. If used, this will typically be made
1974 to depend on the incoming interface address. Because $interface_address is
1975 not set up until the daemon has forked a reception process, error responses
1976 that can happen earlier (such as "too many connections") no longer contain
1979 12. If an expansion in a condition on a "warn" statement fails because a lookup
1980 defers, the "warn" statement is abandoned, and the next ACL statement is
1981 processed. Previously this caused the whole ACL to be aborted.
1983 13. Added the iplsearch lookup type.
1985 14. Added ident_timeout as a log selector.
1987 15. Added tls_certificate_verified as a log selector.
1989 16. Added a global option tls_require_ciphers (compare the smtp transport
1990 option of the same name). This controls incoming TLS connections.
1992 17. I finally figured out how to make tls_require_ciphers do a similar thing
1993 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1994 before starting the TLS session.
1996 18. Tabs are now shown as \t in -bP output.
1998 19. If the log selector return_path_on_delivery was set, Exim crashed when
1999 bouncing a message because it had too many Received: header lines.
2001 20. If two routers both had headers_remove settings, and the first one included
2002 a superfluous trailing colon, the final name in the first list and the
2003 first name in the second list were incorrectly joined into one item (with a
2004 colon in the middle).
2010 1. Added -C and -D options to the exinext utility, mainly to make it easier
2011 to include in the automated testing, but these could be helpful when
2012 multiple configurations are in use.
2014 2. The exinext utility was not formatting the output nicely when there was
2015 an alternate port involved in the retry record key, nor when there was a
2016 message id as well (for retries that were specific to a specific message
2017 and a specific host). It was also confused by IPv6 addresses, because of
2018 the additional colons they contain. I have fixed the IPv4 problem, and
2019 patched it up to do a reasonable job for IPv6.
2021 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2022 delivery, the log line now contains "pipelined" if PIPELINING was used.
2024 4. An SMTP transport process used to panic and die if the bind() call to set
2025 an explicit outgoing interface failed. This has been changed; it is now
2026 treated in the same way as a connect() failure.
2028 5. A reference to $sender_host_name in the part of a conditional expansion
2029 that was being skipped was still causing a DNS lookup. This no longer
2032 6. The def: expansion condition was not recognizing references to header lines
2033 that used bh_ and bheader_.
2035 7. Added the _cache feature to named lists.
2037 8. The code for checking quota_filecount in the appendfile transport was
2038 allowing one more file than it should have been.
2040 9. For compatibility with Sendmail, the command line option
2048 and sets the incoming protocol and host name (for trusted callers). The
2049 host name and its colon can be omitted when only the protocol is to be set.
2050 Note the Exim already has two private options, -pd and -ps, that refer to
2051 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2052 "s", but I don't think that's a major issue.
2054 10. A number of refactoring changes to the code, none of which should affect
2057 (a) The number of logging options was getting close to filling up the
2058 32-bit word that was used as a bit map. I have split them into two classes:
2059 those that are passed in the argument to log_write(), and those that are
2060 only ever tested independently outside of that function. These are now in
2061 separate 32-bit words, so there is plenty of room for expansion again.
2062 There is no change in the user interface or the logging behaviour.
2064 (b) When building, for example, log lines, the code previously used a
2065 macro that called string_cat() twice, in order to add two strings. This is
2066 not really sufficiently general. Furthermore, there was one instance where
2067 it was actually wrong because one of the argument was used twice, and in
2068 one call a function was used. (As it happened, calling the function twice
2069 did not affect the overall behaviour.) The macro has been replaced by a
2070 function that can join an arbitrary number of extra strings onto a growing
2073 (c) The code for expansion conditions now uses a table and a binary chop
2074 instead of a serial search (which was left over from when there were very
2075 few conditions). Also, it now recognizes conditions like "pam" even when
2076 the relevant support is not compiled in: a suitably worded error message is
2077 given if an attempt is made to use such a condition.
2079 11. Added ${time_interval:xxxxx}.
2081 12. A bug was causing one of the ddress fields not to be passed back correctly
2082 from remote delivery subprocesses. The field in question was not being
2083 subsequently used, so this caused to problems in practice.
2085 13. Added new log selectors queue_time and deliver_time.
2087 14. Might have fixed a bug in maildirsizefile handling that threw up
2088 "unexpected character" debug warnings, and recalculated the data
2089 unnecessarily. In any case, I expanded the warning message to give more
2092 15. Added the message "Restricted characters in address" to the statements in
2093 the default ACL that block characters like @ and % in local parts.
2095 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2096 Three changes have been made:
2098 (a) There was a serious bug; a negative response to MAIL caused the whole
2099 recipient domain to be cached as invalid, thereby blocking all messages
2100 to all local parts at the same domain, from all senders. This bug has
2101 been fixed. The domain is no longer cached after a negative response to
2102 MAIL if the sender used is not empty.
2104 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2107 (c) A new callout option, "use_sender" has been added for people who want
2108 the modified behaviour.
2114 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2117 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2118 indeed breaks things for older releases.
2120 3. Added additional logging to the case where there is a problem reading data
2121 from a filter that is running in a subprocess using a pipe, in order to
2122 try to track down a specific problem.
2124 4. Testing facility fudge: when running in the test harness and attempting
2125 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2126 getting "No route to host". Convert this to a timeout.
2128 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2131 6. Some OS don't have socklen_t but use size_t instead. This affects the
2132 fifth argument of getsockopt() amongst other things. This is now
2133 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2134 can be set for individual OS. I have set it for SunOS5, OSF1, and
2135 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2136 some earlier ones do not.
2138 7. Change 4.30/15 was not doing the test caselessly.
2140 8. The standard form for an IPv6 address literal was being rejected by address
2141 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2142 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2143 this, as well as the form without the "IPv6" on the front (but only when
2144 address literals are enabled, of course).
2146 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2148 10. Exim crashed if a message with an empty sender address specified by -f
2149 encountered a router with an errors_to setting. This could be provoked only
2150 by a command such as
2154 where an empty string was supplied; "<>" did not hit this bug.
2156 11. Installed PCRE release 4.5.
2158 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2159 remained set. It is now erased.
2161 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2162 times from message ids (which are base 36 rather than the normal 62).
2164 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2165 were being counted as actual protocol errors, and logged if the log
2166 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2167 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2168 and DATA following a set of rejected RCPTs do not count as protocol errors.
2169 In other words, Exim assumes they were pipelined, though this may not
2170 actually be the case. Of course, in all cases the client gets an
2171 appropriate error code.
2173 15. If a lookup fails in an ACL condition, a message about the failure may
2174 be available; it is used if testing the ACL cannot continue, because most
2175 such messages specify what the cause of the deferral is. However, some
2176 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2177 that caused an old message to be retained and used if a later statement
2178 caused a defer, replacing the real cause of the deferral.
2180 16. If an IP address had so many PTR records that the DNS lookup buffer
2181 was not large enough to hold them, Exim could crash while trying to process
2182 the truncated data. It now detects and logs this case.
2184 17. Further to 4.21/58, another change has been made: if (and only if) the
2185 first line of a message (the first header line) ends with CRLF, a bare LF
2186 in a subsequent header line has a space inserted after it, so as not to
2187 terminate the header.
2189 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2190 unnecessarily, used atoi() instead of strtol(), and didn't check the
2191 termination when getting file sizes from file names by regex.
2193 19. Completely re-implemented the support for maildirsize files, in the light
2194 of a number of problems with the previous contributed implementation
2195 (4.30/29). In particular:
2197 . If the quota is zero, the maildirsize file is maintained, but no quota is
2200 . If the maildir directory does not exist, it is created before any attempt
2201 to write a maildirsize file.
2203 . The quota value in the file is just a cache; if the quota is changed in
2204 the transport, the new value overrides.
2206 . A regular expression is available for excluding directories from the
2209 20. The autoreply transport checks the characters in options that define the
2210 message's headers; it allows continued headers, but it was checking with
2211 isspace() after an embedded newline instead of explicitly looking for a
2214 21. If all the "regular" hosts to which an address was routed had passed their
2215 expiry times, and had not reached their retry times, the address was
2216 bounced, even if fallback hosts were defined. Now Exim should go on to try
2219 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2220 equivalent code in the SMTP transport. Some hosts send humungous responses
2221 to HELO/EHLO, more than 1024 it seems.
2223 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2224 gives compiler warnings in some environments. I've now done it "properly",
2227 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2228 (because of problems with the built-in one) was declared to return uschar *
2229 instead of char *, causing compiler failure.
2231 25. Fixed a file descriptor leak when processing alias/forward files.
2233 26. Fixed a minor format string issue in dbfn.c.
2235 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2237 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2238 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2241 29. When a sender address is verified, it is cached, to save repeating the test
2242 when there is more than one recipient in a message. However, when the
2243 verification involves a callout, it is possible for different callout
2244 options to be set for different recipients. It is too complicated to keep
2245 track of this in the cache, so now Exim always runs a verification when a
2246 callout is required, relying on the callout cache for the optimization.
2247 The overhead is duplication of the address routing, but this should not be
2250 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2251 to be verified with callout=postmaster, and the main callout worked but the
2252 postmaster check failed, the verification correctly failed. However, if a
2253 subsequent RCPT command asked for sender verification *without* the
2254 postmaster check, incorrect caching caused this verification also to fail,
2257 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2258 it was not caching the DNS options (qualify_single, search_parents) that
2259 were used when the lookup failed. A subsequent lookup with different
2260 options therefore always gave the same answer, though there were cases
2261 where it should not have. (Example: a "domains = !$mx_any" option on a
2262 dnslookup router: the "domains" option is always processed without any
2263 widening, but the router might have qualify_single set.) Now Exim uses the
2264 cached value only when the same options are set.
2266 32. Added John Jetmore's "exipick" utility to the distribution.
2268 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2269 than a timeout (e.g. a certificate is required, and is not provided), an
2270 Exim server now closes the connection immediately. Previously it waited for
2271 the client to close - but if the client is SSL, it seems that they each
2272 wait for each other, leading to a delay before one of them times out.
2274 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2275 maintained 0.8.x compatibility because I don't think many are using it, and
2276 it is clearly obsolete.
2278 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2281 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2282 client certificate was expired. A simple patch fixes this, though I don't
2283 understand the full logic of why the verify callback is called multiple
2286 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2287 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2288 which causes problems with some clients (such as the Certicom SSL Plus
2289 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2290 disables the coutermeasure allowing Eudora to connect."
2292 38. Exim was not checking that a write() to a log file succeeded. This could
2293 lead to Bad Things if a log got too big, in particular if it hit a file
2294 size limit. Exim now panics and dies if it cannot write to a log file, just
2295 as it does if it cannot open a log file.
2297 39. Modified OS/Makefile-Linux so that it now contains
2299 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2301 The two -D definitions ensure that Exim is compiled with large file
2302 support, which makes it possible to handle log files that are bigger than
2305 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2306 instance) a domain was checked against a named list that involved a lookup,
2307 causing $domain_data to be set, then another domain was checked against the
2308 same list, then the first domain was re-checked, the value of $domain_data
2309 after the final check could be wrong. In particular, if the second check
2310 failed, it could be set empty. This bug probably also applied to
2313 41. The strip_trailing_dot option was not being applied to the address given
2314 with the -f command-line option.
2316 42. The code for reading a message's header from the spool was incrementing
2317 $received_count, but never initializing it. This meant that the value was
2318 incorrect (doubled) while delivering a message in the same process in which
2319 it was received. In the most common configuration of Exim, this never
2320 happens - a fresh exec is done - but it can happen when
2321 deliver_drop_privilege is set.
2323 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2324 it now includes up to 150 characters of the unexpected data in the log
2327 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2328 and building data strings. The size of both of these buffers was 10 000
2329 bytes - far larger than anybody would *ever* want, thought I. Needless to
2330 say, somebody hit the limit. I have increased the maximum line length to
2331 20 000 and the maximum data length of concatenated lines to 100 000. I have
2332 also fixed two bugs, because there was no checking on these buffers. Tsk,
2333 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2334 buffer is too small.
2336 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2337 lsearch lookups. Now it does.
2339 46. When parsing a route_list item in a manualroute router, a fixed-length
2340 buffer was used for the list of hosts. I made this 1024 bytes long,
2341 thinking that nobody would ever have a list of hosts that long. Wrong.
2342 Somebody had a whole pile of complicated expansion conditions, and the
2343 string was silently truncated, leading to an expansion error. It turns out
2344 that it is easier to change to an unlimited length (owing to other changes
2345 that have happened since this code was originally written) than to build
2346 structure for giving a limitation error. The length of the item that
2347 expands into the list of hosts is now unlimited.
2349 47. The lsearch lookup could not handle data where the length of text line was
2350 more than 4095 characters. Such lines were truncated, leading to shortened
2351 data being returned. It should now handle lines of any length.
2353 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2354 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2357 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2358 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2359 no longer happens in comment lines. A list of replacements is now placed
2360 at the head of all of the source files, except those whose only change is
2361 to replace PERL_COMMAND in the very first #! line.
2363 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2364 messages on the queue, with a bottom-up merge sort, using code contributed
2365 by Michael Haardt. This should make operations like -bp somewhat faster on
2366 large queues. It won't affect queue runners, except when queue_run_in_order
2369 51. Installed eximstats 1.31 in the distribution.
2371 52. Added support for SRV lookups to the dnslookup router.
2373 53. If an ACL referred to $message_body or $message_body_end, the value was not
2374 reset for any messages that followed in the same SMTP session.
2376 54. The store-handling optimization for building very long strings was not
2377 differentiating between the different store pools. I don't think this
2378 actually made any difference in practice, but I've tidied it.
2380 55. While running the routers to verify a sender address, $sender_address
2381 was still set to the sender address. This is wrong, because when routing to
2382 send a bounce to the sender, it would be empty. Therefore, I have changed
2383 it so that, while verifying a sender address, $sender_address is set to <>.
2384 (There is no change to what happens when verifying a recipient address.)
2386 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2387 target A or AAAA records (if not already returned) without resetting the
2388 qualify_single or search_parents options of the DNS resolver. These are
2389 inappropriate in this case because the targets of MX and SRV records must
2390 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2391 target that, when qualified, matched something in the local domain. These
2392 two options are now turned off when doing these lookups.
2394 57. It seems that at least some releases of Reiserfs (which does not have the
2395 concept of a fixed number of inodes) returns zero and not -1 for the
2396 number of available inodes. This interacted badly with check_spool_inodes,
2397 which assumed that -1 was the "no such thing" setting. What I have done is
2398 to check that the total number of inodes is greater than zero before doing
2399 the test of how many are available.
2401 58. When a "warn" ACL statement has a log_message modifier, the message is
2402 remembered, and not repeated. This is to avoid a lot of repetition when a
2403 message has many recipients that cause the same warning to be written.
2404 Howewer, Exim was preserving the list of already written lines for an
2405 entire SMTP session, which doesn't seem right. The memory is now reset if a
2406 new message is started.
2408 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2409 debugging output unless log_rewrite was also set.
2411 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2412 of (int)(handle) when we know that handle contains (void *)(-1).
2414 61. The Exim daemon panic-logs an error return when it closes the incoming
2415 connection. However "connection reset by peer" seems to be common, and
2416 isn't really an error worthy of noting specially, so that particular error
2419 62. When Exim is trying to find all the local interfaces, it used to panic and
2420 die if the ioctl to get the interface flags failed. However, it seems that
2421 on at least one OS (Solaris 9) it is possible to have an interface that is
2422 included in the list of interfaces, but for which you get a failure error
2423 for this call. This happens when the interface is not "plumbed" into a
2424 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2425 failure of the "get flags" call assumes that the interface is down.
2427 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2428 makes life easier for people who are doing arithmetic on fields extracted
2429 from dates, where you often get leading zeros that should not be
2430 interpreted as octal.
2432 64. Added qualify_domain to the redirect router, to override the global
2435 65. If a pathologically long header line contained very many addresses (the
2436 report of this problem mentioned 10 000) and each of them was rewritten,
2437 Exim could use up a very large amount of memory. (It kept on making new
2438 copies of the header line as it rewrote, and never released the old ones.)
2439 At the expense of a bit more processing, the header rewriting function has
2440 been changed so that it no longer eats memory in this way.
2442 66. The generation of the Received: header has been moved from the time that a
2443 message starts to be received, to the time that it finishes. The timestamp
2444 in the Received: header should now be very close to that of the <= log
2445 line. There are two side-effects of this change:
2447 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2448 logged header lines no longer include the local Received: line, because
2449 it has not yet been created. The same applies to a copy of the message
2450 that is returned to a non-SMTP sender when a message is rejected.
2452 (b) When a filter file is tested using -bf, no additional Received: header
2453 is added to the test message. After some thought, I decided that this
2456 This change does not affect the value of $received_for. It is still set
2457 after address rewriting, but before local_scan() is called.
2459 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2461 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2462 gave an unhelpful panic error message, and a defer error. I have managed to
2463 change this behaviour so that it now rejects any supplied certificate,
2464 which seems right, as the list of acceptable certificates is empty.
2466 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2467 gave an unhelpful defer error. I have not managed to make this reject any
2468 supplied certificates, but the error message it gives is "no certificate
2469 supplied", which is not helpful.
2471 70. exigrep's output now also includes lines that are not associated with any
2472 message, but which match the given pattern. Implemented by a patch from
2473 Martin Sluka, which also tidied up the Perl a bit.
2475 71. Recipient callout verification, like sender verification, was using <> in
2476 the MAIL FROM command. This isn't really the right thing, since the actual
2477 sender may affect whether the remote host accepts the recipient or not. I
2478 have changed it to use the actual sender in the callout; this means that
2479 the cache record is now keyed on a recipient/sender pair, not just the
2480 recipient address. There doesn't seem to be a real danger of callout loops,
2481 since a callout by the remote host to check the sender would use <>.
2482 [SEE ABOVE: changed after hitting problems.]
2484 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2485 temporary errors. However, in the case of such a code being given after
2486 the end of a data transmission (i.e. after ".") Exim was failing to write
2487 a retry record for the message. (Yes, there was some broken host that was
2488 actually sending 8xx at this point.)
2490 73. An unknown lookup type in a host list could cause Exim to panic-die when
2491 the list was checked. (An example that provoked this was putting <; in the
2492 middle of a list instead of at the start.) If this happened during a DATA
2493 ACL check, a -D file could be left lying around. This kind of configuration
2494 error no longer causes Exim to die; instead it causes a defer errror. The
2495 incident is still logged to the main and panic logs.
2497 74. Buglet left over from Exim 3 conversion. The message "too many messages
2498 in one connection" was written to the rejectlog but not the mainlog, except
2499 when address rewriting (yes!) was being logged.
2501 75. Added write_rejectlog option.
2503 76. When a system filter was run not as root (that is, when system_filter_user
2504 was set), the values of the $n variables were not being returned to the
2505 main process; thus, they were not subsequently available in the $sn
2508 77. Added +return_path_on_delivery log selector.
2510 78. A connection timeout was being treated differently from recipients deferred
2511 when testing hosts_max_try with a message that was older than the host's
2512 retry timeout. (The host should not be counted, thus allowing all hosts to
2513 be tried at least once before bouncing.) This may have been the cause of an
2514 occasionally reported bug whereby a message would remain on the queue
2515 longer than the retry timeout, but would be bounced if a delivery was
2516 forced. I say "may" because I never totally pinned down the problem;
2517 setting up timeout/retry tests is difficult. See also the next item.
2519 79. The ultimate address timeout was not being applied to errors that involved
2520 a combination of host plus message (for example, a timeout on a MAIL
2521 command). When an address resolved to a number of possible hosts, and they
2522 were not all tried for each delivery (e.g. because of hosts_max_try), a
2523 message could remain on the queue longer than the retry timeout.
2525 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2528 81. Fixed an obscure SMTP outgoing bug which required at least the following
2529 conditions: (a) there was another message waiting for the same server;
2530 (b) the server returned 5xx to all RCPT commands in the first message so
2531 that the message was not completed; (c) the server dropped the connection
2532 or gave a negative response to the RSET that Exim sends to abort the
2533 transaction. The observed case was a dropped connection after DATA that had
2534 been sent in pipelining mode. That is, the server had advertised PIPELINING
2535 but was not implementing it correctly. The effect of the bug was incorrect
2536 behaviour, such as trying another host, and this could lead to a crash.
2542 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2543 and daemon.c were passed as pointers to ints; they should have been
2544 pointers to socklen_t variables (which are typically unsigned ints).
2546 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2549 3. Fixed a really odd bug that affected only the testing scheme; patching a
2550 certain fixed string in the binary changed the value of another string that
2551 happened to be identical to the end of the original first string.
2553 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2554 name", it returns that address as the IP address. On some operating
2555 systems (e.g. Solaris), it also passes back the IP address string as the
2556 "host name". However, on others (e.g. Linux), it passes back an empty
2557 string. Exim wasn't checking for this, and was changing the host name to an
2558 empty string, assuming it had been canonicized.
2560 5. Although rare, it is permitted to have more than one PTR record for a given
2561 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2562 all the names associated with an address, because they do in Solaris.
2563 However, it seems that they do not in Linux for data that comes from the
2564 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2565 I found this out when I moved to a new Linux workstation and tried to run
2566 the Exim test suite.
2568 To get round this problem I have changed the code so that it now does its
2569 own call to the DNS to look up PTR records when searching for a host name.
2570 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2571 addresses that are only in /etc/hosts are still found.
2573 This behaviour is, however, controlled by an option called host_lookup_
2574 order, which defaults to "bydns:byaddr". If people want to use the other
2575 order, or indeed, just use one or the other means of lookup, they can
2576 specify it in this variable.
2578 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2579 some operating systems, this comes back from gethostbyaddr() as an empty
2580 string, and this is what Exim used to test for. However, it seems that in
2581 other systems, "." is yielded. Exim now tests for this case too.
2583 7. The values of check_spool_space and check_log_space are now held internally
2584 as a number of kilobytes instead of an absolute number of bytes. If a
2585 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2586 kilobyte. This means that much larger values can be stored.
2588 8. Exim monitor: an attempt to get the action menu when not actually pointing
2589 at a message produces an empty menu entitled "No message selected". This
2590 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2591 no entries in it ("Shell widget menu has zero width and/or height"). So I
2592 have added a single, blank menu entry in this case.
2594 9. Added ${quote_local_part.
2596 10. MIME decoding is now applied to the contents of Subject: header lines when
2599 11. Now that a reference to $sender_host_address automatically causes a reverse
2600 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2601 host lookup before query-style lookups in lists that might use this
2602 variable. This has therefore been abolished, and the "net-" prefix is no
2603 longer necessary for query-style lookups.
2605 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2606 have been a typo for LFLAGS, so it has been changed.
2608 13. The install script calls Exim with "-C /dev/null" in order to find the
2609 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2610 to be output. Howeve, since Exim outputs its version number before the
2611 error, it didn't break the script. It just looked ugly. I fixed this by
2612 always allowing "-C /dev/null" if the caller is root.
2614 14. Ignore overlarge ACL variable number when reading spool file - insurance
2615 against a later release with more variables having written the file.
2617 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2618 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2619 this, as well as the form without the "IPv6" on the front.
2621 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2622 OS/Makefile-Darwin file.
2624 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2625 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2627 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2628 variable. However, some OS use ints for inodes. I've added cast to long int
2629 to get rid of the compiler warning.
2631 19. I had forgotten to lock out "/../" in configuration file names when
2632 ALT_CONFIG_PREFIX was set.
2634 20. Routers used for verification do not need to specify transports. However,
2635 if such a router generated a host list, and callout was configured, Exim
2636 crashed, because it could not find a port number from the (non-existent)
2637 transport. It now assumes port 25 in this circumstance.
2639 21. Added the -t option to exigrep.
2641 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2642 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2643 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2644 them is set, LOOKUP_LSEARCH is forced.
2646 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2648 24. Added sender and host information to the "rejected by local_scan()" log
2649 line; previously there was no indication of these.
2651 25. Added .include_if_exists.
2653 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2654 Linux. It turns out that not all file systems support this. Apparently some
2655 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2656 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2657 not supported on the file descriptor, is now ignored when Exim is trying to
2658 sync a directory. This applies only to Linux.
2660 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2662 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2663 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2666 29. Added support for maildirsize files from supplied patch (modified a bit).
2668 30. The use of :fail: followed by an empty string could lead Exim to respond to
2669 sender verification failures with (e.g.):
2671 550 Verification failed for <xxx>
2672 550 Sender verify failed
2674 where the first response line was missing the '-' that indicates it is not
2675 the final line of the response.
2677 31. The loop for finding the name of the user that called Exim had a hardwired
2678 limit of 10; it now uses the value of finduser_retries, which is used for
2679 all other user lookups.
2681 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2684 33. Exim was neglecting to zero errno before one call of strtol() when
2685 expanding a string and expecting an integer value. On some systems this
2686 resulted in spurious "integer overflow" errors. Also, it was casting the
2687 result into an int without checking.
2689 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2690 did not work. The code looks as if it has *never* worked, though it appears
2691 to have been documented since at least releast 1.62. I have made it work.
2693 35. The "timeout_DNS" error in retry rules, also documented since at least
2694 1.62, also never worked. As it isn't clear exactly what this means, and
2695 clearly it isn't a major issue, I have abolished the feature by treating it
2696 as "timeout", and writing a warning to the main and panic logs.
2698 36. The display of retry rules for -brt wasn't always showing the error code
2701 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2702 timeout_connect_A, timeout_connect_MX.
2704 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2705 to the empty sender.
2707 39. The daemon was not analysing the content of -oX till after it had closed
2708 stderr and disconnected from the controlling terminal. This meant that any
2709 syntax errors were only noted on the panic log, and the return code from
2710 the command was 0. By re-arranging the code a little, I've made the
2711 decoding happen first, so such errors now appear on stderr, and the return
2712 code is 1. However, the actual setting up of the sockets still happens in
2713 the disconnected process, so errors there are still only recorded on the
2716 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2717 connections (as happens on some IP stacks) was logged at start up time as
2718 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2719 it from "IPv6 and IPv4", which means that two separate sockets are being
2722 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2723 says whether AF_INET or AF_INET6 was passed as an argument.
2725 42. Exiwhat output was messed up when time zones were included in log
2728 43. Exiwhat now gives more information about the daemon's listening ports,
2729 and whether -tls-on-connect was used.
2731 44. The "port" option of the smtp transport is now expanded.
2733 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2734 silently ignored. Now an error message is written to the main and panic
2737 46. There's a new ACL modifier called "logwrite" which writes to a log file
2738 as soon as it is encountered.
2740 47. Added $local_user_uid and $local_user_gid at routing time.
2742 48. Exim crashed when trying to verify a sender address that was being
2745 49. Exim was recognizing only a space character after ".include". It now also
2746 recognizes a tab character.
2748 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2749 extracting the relevant information from the specification. The man page no
2750 longer contains scrambled data for the -d option, and I've added a section
2751 at the front about calling Exim under different names.
2753 51. Added "extra_headers" argument to the "mail" command in filter files.
2755 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2758 53. Installed eximstats 1.29.
2760 54. Added transport_filter_timeout as a generic transport option.
2762 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2763 Cc: header lines. This was required by RFC 822, but it not required by RFC
2766 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2767 incoming messages that did not have them. Now it does so only if the
2768 message originates locally, that is, if there is no associated remote host
2769 address. When Resent- header lines are present, this applies to the Resent-
2770 lines rather than the non-Resent- lines.
2772 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2773 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2775 58. Messages for configuration errors now include the name of the main
2776 configuration file - useful now that there may be more than one file in a
2777 list (.included file names were always shown).
2779 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2780 for those rare installations that do not start the daemon as root or run it
2781 setuid root. I've cut out the call to initgroups() if the daemon is not
2784 60. The Exim user and group can now be bound into the binary as text strings
2785 that are looked up at the start of Exim's processing.
2787 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2789 62. Added $mailstore_basename variable.
2791 63. Installed patch to sieve.c from Michael Haardt.
2793 64. When Exim failed to open the panic log after failing to open the main log,
2794 the original message it was trying to log was written to stderr and debug
2795 output, but if they were not available (the usual case in production), it
2796 was lost. Now it is written to syslog before the two lines that record the
2797 failures to open the logs.
2799 65. Users' Exim filters run in subprocesses under the user's uid. It is
2800 possible for a "deliver" command or an alias in a "personal" command to
2801 provoke an address rewrite. If logging of address rewriting is configured,
2802 this fails because the process is not running as root or exim. There may be
2803 a better way of dealing with this, but for the moment (because 4.30 needs
2804 to be released), I have disabled address rewrite logging when running a
2805 filter in a non-root, non-exim process.
2811 1. The buildconfig auxiliary program wasn't quoting the value set for
2812 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2813 not defined. This bug was masked in 4.22 by the effect that was fixed in
2816 2. Some messages that were rejected after a message id was allocated were
2817 shown as "incomplete" by exigrep. It no longer does this for messages that
2818 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2820 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2821 have allow_domain_literals set, the ID did not get logged in the <= line.
2822 Domain literals are now always recognized in Message-ID: header lines.
2824 4. The first argument for a ${extract expansion item is the key name or field
2825 number. Leading and trailing spaces in this item were not being ignored,
2826 causing some misleading effects.
2828 5. When deliver_drop_privilege was set, single queue runner processes started
2829 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2830 same command in the process it spins off) were not dropping privilege.
2832 6. When the daemon running as "exim" started a queue runner, it always
2833 re-executed Exim in the spun-off process. This is a waste of effort when
2834 deliver_drop_privilege is set. The new process now just calls the
2835 queue-runner function directly.
2841 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2844 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2845 ignored. Though the use of -oP was forcing the writing of a pid file, it
2846 was always written to the default place.
2848 3. If the message "no IP address found for host xxxx" is generated during
2849 incoming verification, it is now followed by identification of the incoming
2850 connection (so you can more easily find what provoked it).
2852 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2854 5. Added some features to "harden" Exim a bit more against certain attacks:
2856 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2857 be put in Local/Makefile. This is like the never_users runtime option,
2858 but it cannot be overridden. The default setting is "root".
2860 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2861 prefix string with which any file named in a -C command line option
2864 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2865 is retained for -C and -D only if the caller of Exim is root. Without
2866 it, the exim user may also use -C and -D and retain privilege.
2868 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2869 command line option is disabled.
2871 6. Macro names set by the -D option must start with an upper case letter, just
2872 like macro names defined in the configuration file.
2874 7. Added "dereference=" facility to LDAP.
2876 8. Two instances of the typo "uknown" in the source files are fixed.
2878 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2879 the Configure-Makefile script screwed up while processing it.
2881 10. Incorporated PCRE 4.4.
2883 11. The SMTP synchronization check was not operating right at the start of an
2884 SMTP session. For example, it could not catch a HELO sent before the client
2885 waited for the greeting. There is now a check for outstanding input at the
2886 point when the greeting is written. Because of the duplex, asynchronous
2887 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2888 way, but not yet received, when the check is performed.
2890 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2891 on TCP/IP sockets, because this apparently causes some broken clients to
2894 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2895 unchanged) from the Cygwin maintainer.
2897 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2898 was supported instead of testing for mbx. Effectively a typo.
2900 15. The spa authenticator server code was not checking that the input it
2901 received was valid base64.
2903 16. The debug output line for the "set" modifier in ACLs was not showing the
2904 name of the variable that was being set.
2906 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2908 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2909 lookup on its own. Something else was needed to trigger the lookup. For
2910 example, a match in host_lookup or the need for a host name in a host list.
2911 Now, if $sender_host_name is referenced and the host name has not yet been
2912 looked up, a lookup is performed. If the lookup fails, the variable remains
2913 empty, and $host_lookup_failed is set to "1".
2915 19. Added "eqi" as a case-independent comparison operator.
2917 20. The saslauthd authentication condition could segfault if neither service
2918 nor realm was specified.
2920 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2921 error message that was logged was misleading, and incoming SMTP
2922 connections were dropped. The message is now more accurate, and temporary
2923 errors are given to SMTP connections.
2925 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2926 (except RSET) with a 421 error, until QUIT is received. However, it was
2927 failing to send a response to QUIT.
2929 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2930 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2931 ACL is now run after the helo_try_verify_hosts code.
2933 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2934 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2935 case-independent, so other case variants are also recognized). Apparently
2936 some systems use these upper case variants.
2938 25. If more than two messages were waiting for the same host, and a transport
2939 filter was specified for the transport, Exim sent two messages over the
2940 same TCP/IP connection, and then failed with "socket operation on non-
2941 socket" when it tried to send the third.
2943 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2945 27. The extern definition of crypt16() in expand.c was not being excluded when
2946 the OS had its own crypt16() function.
2948 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2949 as a preferred synonym for return_size_limit, both as an option and as an
2952 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2954 30. Changed the default configuration ACL to relax the local part checking rule
2955 for addresses that are not in any local domains. For these addresses,
2956 slashes and pipe symbols are allowed within local parts, but the sequence
2957 /../ is explicitly forbidden.
2959 31. SPA server authentication was not clearing the challenge buffer before
2962 32. log_message in a "warn" ACL statement was writing to the reject log as
2963 well as to the main log, which contradicts the documentation and doesn't
2964 seem right (because no rejection is happening). So I have stopped it.
2966 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2967 However, I am unable to do any testing of this.
2969 34. Fixed an infelicity in the appendfile transport. When checking directories
2970 for a mailbox, to see if any needed to be created, it was accidentally
2971 using path names with one or more superfluous leading slashes; tracing
2972 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2974 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2975 contents are added to the log line that is written for every discarded
2976 recipient. (Previously a log_message setting was ignored.)
2978 36. The ${quote: operator now quotes the string if it is empty.
2980 37. The install script runs exim in order to find its version number. If for
2981 some reason other than non-existence or emptiness, which it checks, it
2982 could not run './exim', it was installing it with an empty version number,
2983 i.e. as "exim-". This error state is now caught, and the installation is
2986 38. An argument was missing from the function that creates an error message
2987 when Exim fails to connect to the socket for saslauthd authentication.
2988 This could cause Exim to crash, or give a corrupted message.
2990 39. Added isip, isip4, and isip6 to ${if conditions.
2992 40. The ACL variables $acl_xx are now saved with the message, and can be
2993 accessed later in routers, transports, and filters.
2995 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2996 strings in the file are not string-expanded.
2998 42. If a MAIL command specified a SIZE value that was too large to fit into an
2999 int variable, the check against message_size_limit failed. Such values are
3000 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3001 day this will have to be increased, but I don't think I want to be around
3002 when emails are that large.
3009 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3010 iconv() is not standard in FreeBSD.
3012 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3013 IPv6 enabled. The observed symptom was a segmentation fault on return from
3014 the function os_common_find_running_interfaces() in src/os.c.
3016 3. In the check_special_case() function in daemon.c I had used "errno" as an
3017 argument name, which causes warnings on some systems. This was basically a
3018 typo, since it was named "eno" in the comments!
3020 4. The code that waits for the clock to tick (at a resolution of some fraction
3021 of a second) so as to ensure message-id uniqueness was always waiting for
3022 at least one whole tick, when it could have waited for less. [This is
3023 almost certainly not relevant at current processor speeds, where it is
3024 unlikely to ever wait at all. But we try to future-proof.]
3026 5. The function that sleeps for a time interval that includes fractions of a
3027 second contained a race. It did not block SIGALRM between setting the
3028 timer, and suspending (a couple of lines later). If the interval was short
3029 and the sigsuspend() was delayed until after it had expired, the suspension
3030 never ended. On busy systems this could lead to processes getting stuck for
3033 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3034 process, before it forks any delivery processes. The open lookup caching
3035 mechanism meant that the open file or database connection was passed into
3036 the delivery process. The problem was that delivery processes always tidy
3037 up cached lookup data. This could cause a problem for the next delivery
3038 process started by the queue runner, because the external queue runner
3039 process does not know about the closure. So the next delivery process
3040 still has data in the lookup cache. In the case of a file lookup, there was
3041 no problem because closing a file descriptor in a subprocess doesn't affect
3042 the parent. However, if the lookup was caching a connection to a database,
3043 the connection was closed, and the second delivery process was likely to
3044 see errors such as "PGSQL: query failed: server closed the connection
3045 unexpectedly". The problem has been fixed by closing all cached lookups
3046 in a queue runner before running a delivery process.
3048 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3049 seem to have the "const" qualifier which it has on other OS. I've
3052 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3053 *of the same type* (client or server) with the same public name that an
3054 error should be diagnosed.
3056 9. When Exim looked up a host name for an IP address, but failed to find the
3057 original IP address when looking up the host name (a safety check), it
3058 output the message "<ip address> does not match any IP for NULL", which was
3059 confusing, to say the least. The bug was that the host name should have
3060 appeared instead of "NULL".
3062 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3063 user that is built into the binary, and the -C or -D options is used, root
3064 privilege is dropped before the configuration file is read. In addition,
3065 logging is switched to stderr instead of the normal log files. If the
3066 configuration then re-defines the Exim user, the unprivileged environment
3067 is probably not what is expected, so Exim logs a panic warning message (but
3070 However, if deliver_drop_privilege is set, the unprivileged state may well
3071 be exactly what is intended, so the warning has been cut out in that case,
3072 and Exim is allowed to try to write to its normal log files.
3078 1. smtp_return_error_details was not giving details for temporary sender
3079 or receiver verification errors.
3081 2. Diagnose a configuration error if two authenticators have the same public
3084 3. Exim used not to create the message log file for a message until the first
3085 delivery attempt. This could be confusing when incoming messages were held
3086 for policy or load reasons. The message log file is now created at the time
3087 the message is received, and an initial "Received" line is written to it.
3089 4. The automatically generated man page for command line options had a minor
3090 bug that caused no ill effects; however, a more serious problem was that
3091 the procedure for building the man page automatically didn't always
3092 operate. Consequently, release 4.20 contains an out-of-date version. This
3093 shouldn't happen again.
3095 5. When building Exim with embedded Perl support, the script that builds the
3096 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3097 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3099 6. The freeze_tell option was not being used for messages that were frozen on
3100 arrival, either by an ACL or by local_scan().
3102 7. Added the smtp_incomplete_transaction log selector.
3104 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3105 was accepting AUTH without a new EHLO.
3107 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3108 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3111 10. Logging of TCP/IP connections (when configured) now happens in the main
3112 daemon process instead of the child process, so that the TCP/IP connection
3113 count is more accurate (but it can never be perfect).
3115 11. The use of "drop" in a nested ACL was not being handled correctly in the
3116 outer ACL. Now, if condition failure induced by the nested "drop" causes
3117 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3118 or "require"), the connection is dropped.
3120 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3121 that yield "discard" can now be used with an "accept" or a "discard" verb,
3122 but an error is generated for any others (because I can't see a useful way
3123 to define what should happen).
3125 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3126 are now processed in the same way as lines in the Exim configuration file.
3127 In particular, continuation lines are supported.
3129 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3131 15. Added -ti meaning -t -i.
3133 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3134 domains, and warn by logging if others are found.
3136 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3137 returned from GIFCONF when Exim is trying to find the list of interfaces on
3138 a host. The code in os.c has been modified to copy each ifreq to an aligned
3139 structure in all cases.
3141 Also, in some cases, the returned ifreq's were being copied to a 'struct
3142 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3143 means the last couple of bytes of an IPv6 address could be chopped if the
3144 ifreq contained only a normal sockaddr (14 bytes storage).
3146 18. Named domain lists were not supported in the hosts_treat_as_local option.
3147 An entry such as +xxxx was not recognized, and was treated as a literal
3150 19. Ensure that header lines added by a DATA ACL are included in the reject log
3151 if the ACL subsequently rejects the message.
3153 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3154 MD5 (which is deprecated).
3156 21. When testing a filter file using -bf, Exim was writing a message when it
3157 took the sender from a "From " line in the message, but it was not doing so
3158 when it took $return_path from a Return-Path: header line. It now does.
3160 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3161 with a valid header line field name (a series of printing characters
3162 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3164 23. Changed "disc" in the source to "disk" to conform to the documentation and
3165 the book and for uniformity.
3167 24. Ignore Sendmail's -Ooption=value command line item.
3169 25. When execve() failed while trying to run a command in a pipe transport,
3170 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3171 could be confused with a return value of 69 from the command itself. This
3172 has been changed to 127, the value the shell returns if it is asked to run
3173 a non-existent command. The wording for the related log line suggests a
3174 non-existent command as the problem.
3176 26. If received_header_text expands to an empty string, do not add a Received:
3177 header line to the message. (Well, it adds a token one on the spool, but
3178 marks it "old" so that it doesn't get used or transmitted.)
3180 27. Installed eximstats 1.28 (addition of -nt option).
3182 28. There was no check for failure on the call to getsockname() in the daemon
3183 code. This can fail if there is a shortage of resources on the system, with
3184 ENOMEM, for example. A temporary error is now given on failure.
3186 29. Contrary to the C standard, it seems that in some environments, the
3187 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3188 program. Exim now does this explicitly; it affects the formatting of
3189 timestamps using strftime().
3191 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3192 complaints. I've now initialized all the variables, to avoid this.
3194 32. Header lines added by a system filter were not being "seen" during
3195 transport-time rewrites.
3197 33. The info_callback() function passed to OpenSSL is set up with type void
3198 (*)(SSL *, int, int), as described somewhere. However, when calling the
3199 function (actually a macro) that sets it up, the type void(*)() is
3200 expected. I've put in a cast to prevent warnings from picky compilers.
3202 34. If a DNS black list lookup found a CNAME record, but there were no A
3203 records associated with the domain it pointed at, Exim crashed.
3205 35. If a DNS black list lookup returned more than one A record, Exim ignored
3206 all but the first. It now scans all returned addresses if a particular IP
3207 value is being sought. In this situation, the contents of the
3208 $dnslist_value variable are a list of all the addresses, separated by a
3211 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3212 to accept a host name and all its aliases if the forward lookup for any of
3213 them yielded the IP address of the incoming connection. Now it accepts only
3214 those names whose forward lookup yields the correct IP address. Any other
3215 names are discarded. This closes a loophole whereby a rogue DNS
3216 administrator could create reverse DNS records to break through a
3217 wildcarded host restriction in an ACL.
3219 37. If a user filter or a system filter that ran in a subprocess used any of
3220 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3221 the wrong values were passed to the pipe command ($thisaddress had the
3222 value of $0, $0 had the value of $1, etc). This bug was introduced by
3223 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3225 38. Improved the line breaking for long SMTP error messages from ACLs.
3226 Previously, if there was no break point between 40 and 75 characters, Exim
3227 left the rest of the message alone. Two changes have been made: (a) I've
3228 reduced the minimum length to 35 characters; (b) if it can't find a break
3229 point between 35 and 75 characters, it looks ahead and uses the first one
3230 that it finds. This may give the occasional overlong line, but at least the
3231 remaining text gets split now.
3233 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3234 file descriptors might be low, and that setting 1000 would always raise it.
3235 It turns out that in some environments, the limit is already over 1000 and
3236 that lowering it causes trouble. So now Exim takes care not to decrease it.
3238 40. When delivering a message, the value of $return_path is set to $sender_
3239 address at the start of routing (routers may change the value). By an
3240 oversight, this default was not being set up when an address was tested by
3241 -bt or -bv, which affected the outcome if any router or filter referred to
3244 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3245 log only when "message" or "log_message" are set. However, if one of the
3246 conditions was an address verification, or a call to a nested ACL, the
3247 messages generated by the underlying test were being passed through. This
3248 no longer happens. The underlying message is available in $acl_verify_
3249 message for both "message" and "log_message" expansions, so it can be
3250 passed through if needed.
3252 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3253 new expansion $bh_ to give the encoded byte string without charset
3254 translation. Translation happens only if iconv() is available; HAVE_ICONV
3255 indicates this at build time. HEADERS_CHARSET gives the charset to
3256 translate to; headers_charset can change it in the configuration, and
3257 "headers charset" can change it in an individual filter file.
3259 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3260 that creates RFC 2047 encoded "words" labels them as that charset instead
3261 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3262 expansion operator; (ii) when Exim creates a From: line for a local
3263 message; (iii) when a header line is rewritten to include a "phrase" part.
3265 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3266 buggy, causing it to skip the first lines of messages whose message ID
3267 ended in 'D'. This would not have bitten before Exim release 4.14, because
3268 message IDs were unlikely to end in 'D' before then. The effect was to have
3269 incorrect size information for certain domains.
3271 45. #include "config.h" was missing at the start of the crypt16.c module. This
3272 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3275 46. If there was a timeout during a "random" callout check, Exim treated it as
3276 a failure of the random address, and carried on sending RSET and the real
3277 address. If the delay was just some slowness somewhere, the response to the
3278 original RCPT would be taken as a response to RSET and so on, causing
3279 mayhem of various kinds.
3281 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3282 when I implemented it. It didn't allow for the fact that some option values
3283 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3284 the right test for positive values.
3286 48. Domain names in DNS records are case-independent. Exim always looks them up
3287 in lower case. Some resolvers return domain names in exactly the case they
3288 appear in the zone file, that is, they may contain uppercase letters. Not
3289 all resolvers do this - some return always lower case. Exim was treating a
3290 change of case by a resolver as a change of domain, similar to a widening
3291 of a domain abbreviation. This triggered its re-routing code and so it was
3292 trying to route what was effectively the same domain again. This normally
3293 caused routing to fail (because the router wouldn't handle the domain
3294 twice). Now Exim checks for this case specially, and just changes the
3295 casing of the domain that it ultimately uses when it transmits the message
3298 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3301 50. If a filter generated a file delivery with a non-absolute name (possible if
3302 no home directory exists for the router), the forbid_file option was not
3305 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3306 the existing equality matching.
3308 52. Exim was using ints instead of ino_t variables in some places where it was
3309 dealing with inode numbers.
3311 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3312 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3313 called TMPDIR, and if it finds it is different, it changes its value.
3315 54. The smtp_printf() function is now made available to local_scan() so
3316 additional output lines can be written before returning. There is also an
3317 smtp_fflush() function to enable the detection of a dropped connection.
3318 The variables smtp_input and smtp_batched_input are exported to
3321 55. Changed the default runtime configuration: the message "Unknown user"
3322 has been removed from the ACL, and instead placed on the localuser router,
3323 using the cannot_route_message feature. This means that any verification
3324 failures that generate their own messages won't get overridden. Similarly,
3325 the "Unrouteable address" message that was in the ACL for unverifiable
3326 relay addresses has also been removed.
3328 56. Added hosts_avoid_esmtp to the smtp transport.
3330 57. The exicyclog script was not checking for the esoteric option
3331 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3332 will work only if exicyclog is run under the appropriate euid.
3334 58. Following a discussion on the list, the rules by which Exim recognises line
3335 endings on incoming messages have been changed. The -dropcr and drop_cr
3336 options are now no-ops, retained only for backwards compatibility. The
3337 following line terminators are recognized: LF CRLF CR. However, special
3338 processing applies to CR:
3340 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3341 nor a local message in the state where . is a terminator.
3343 (ii) If a bare CR is encountered in a header line, an extra space is added
3344 after the line terminator so as not to end the header. The reasoning
3345 behind this is that bare CRs in header lines are most likely either
3346 to be mistakes, or people trying to play silly games.
3348 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3349 was being incorrectly given as 18 bytes larger than it should have been.
3350 This is a VOB (very old bug).
3352 60. This may never have affected anything current, but just in case it has:
3353 When the local host is found other than at the start of a list of hosts,
3354 the local host, those with the same MX, and any that follow, are discarded.
3355 When the list in question was part of a longer list of hosts, the following
3356 hosts (not currently being processed) were also being discarded. This no
3357 longer happens. I'm not sure if this situation could ever has previously
3360 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3363 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3364 header. This is something that is recommended in a new Internet Draft, and
3365 is something that is documented as being done by Sendmail. There are two
3366 possible values. For messages generated by the autoreply transport, Exim
3369 Auto-Submitted: auto-replied
3371 whereas for all other generated messages (e.g. bounces) it adds
3373 Auto-Submitted: auto-generated
3375 63. The "personal" condition in filters now includes a test for the
3376 Auto-Submitted: header. If it contains the string "auto-" the message it
3377 not considered personal.
3379 64. Added rcpt_include_affixes as a generic transport option.
3381 65. Added queue_only_override (default true).
3383 66. Added the syslog_duplication option.
3385 67. If what should have been the first header line of a message consisted of
3386 a space followed by a colon, Exim was mis-interpreting it as a header line.
3387 It isn't of course - it is syntactically invalid and should therefore be
3388 treated as the start of the message body. The misbehaviour could have
3389 caused a number of strange effects, including loss of data in subsequent
3390 header lines, and spool format errors.
3392 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3393 client host had authenticated. This control can now be exercised by an ACL
3394 for more flexibility.
3396 69. By default, callouts do not happen when testing with -bh. There is now a
3397 variant, -bhc, which does actually run the callout code, including
3398 consulting and updating the callout cache.
3400 70. Added support for saslauthd authentication, courtesy of Alexander
3403 71. If statvfs() failed on the spool or log directories while checking their
3404 size for availability, Exim confusingly gave the error "space shortage".
3405 Furthermore, in debugging mode it crashed with a floating point exception.
3406 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3407 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3408 serious problem, Exim now writes to the main and panic logs when this
3409 happens, with details of the failure. It then refuses to accept the
3410 incoming message, giving the message "spool directory problem" or "log
3411 directory problem" with a 421 code for SMTP messages.
3413 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3414 0, 1, and 2 exist, because some OS complain for execs without them (see
3415 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3416 descriptors. However, the code omitted to check that the open succeeded,
3417 causing mysterious errors if for some reason the permissions on /dev/null
3418 got screwed. Now Exim writes a message to the main and panic logs, and
3419 bombs out if it can't open /dev/null.
3421 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3422 interact so that it is all more flexible. It is supposed to remain
3423 backwards compatible. Also added extra_local_interfaces.
3425 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3426 to bomb out with an assertion failure - to the client this appears as a
3427 connection drop. This problem occurs in the part of the code that was taken
3428 from the Samba project. Fortunately, the assertion is in a very simple
3429 function, so I have fixed this by reproducing the function inline in the
3430 one place where it is called, and arranging for authentication to fail
3431 instead of killing the process with assert().
3433 75. The SPA client code was not working when the server requested OEM rather
3434 than Unicode encoding.
3436 76. Added code to make require_files with a specific uid setting more usable in
3437 the case where statting the file as root fails - usually a non-root-mounted
3438 NFS file system. When this happens and the failure is EACCES, Exim now
3439 forks a subprocess and does the per-uid checking as the relevant uid.
3441 77. Added process_log_path.
3443 78. If log_file_path was not explicitly set, a setting of check_log_space or
3444 check_log_inodes was ignored.
3446 79. If a space check for the spool or log partitions fails, the incident is now
3447 logged. Of course, in the latter case the data may get lost...
3449 80. Added the %p formatting code to string_format() so that it can be used to
3450 print addresses in debug_print(). Adjusted all the address printing in the
3451 debugging in store.c to use %p rather than %d.
3453 81. There was a concern that a line of code in smtp_in.c could overflow a
3454 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3455 initially expressed, the concern was not well-founded, because trailing
3456 spaces are removed early. However, if the trailing spaces were followed by
3457 a NULL, they did not get removed, so the overflow was possible. Two fixes
3460 (a) I re-wrote the offending code in a cleaner fashion.
3461 (b) If an incoming SMTP command contains a NULL character, it is rejected
3464 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3465 runs initgroups(), so that if the Exim user is in any additional groups,
3466 they will be used during message reception.
3472 The change log for 4.20 and earlier releases has been archived.