1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.253 2005/11/10 15:00:46 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> ...
59 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
60 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
63 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
64 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
66 PH/03 Typo: missing ".o" in src/pcre/Makefile.
68 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
69 header line, restrict the check to what is listed in RFCs 2369 and 2929.
70 Also, for "Auto-Submitted", treat anything other than "no" as
71 non-personal, in accordance with RFC 3834. (Previously it treated
72 anything starting "auto-" as non-personal.)
74 TF/01 The control=submission/name=... option had a problem with syntax
75 errors if the name included a slash character. The /name= option
76 now slurps the rest of the string, so it can include any characters
77 but it must come last in the list of options (after /sender_retain
80 PH/05 Some modifications to the interface to the fake nameserver for the new
87 TK/01 Added the "success_on_redirect" address verification option. See
88 NewStuff for rationale and an example.
90 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
92 PH/02 Patch to exigrep to allow it to work on syslog lines.
94 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
95 fread() to skip over the body file's header line, because in Cygwin the
96 header line is locked and is inaccessible.
98 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
99 co-exist for some time) to make it clear that it is the Exim ID that is
100 referenced, not the Message-ID: header line.
102 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
103 string_format() function, because snprintf() does not exist on all
106 PH/06 The use of forbid_filter_existstest now also locks out the use of the
107 ${stat: expansion item.
109 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
110 protocol synchronization error", to keep the pedants happy.
112 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
113 well as for IRIX systems, when gcc is being used. See the host.c source
116 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
118 PH/10 Named domain lists were not working if used in a queue_smtp_domains
121 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
122 transport and to the smtp transport in LMTP mode.
124 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
126 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
127 run a filter in a subprocess. This could lead to confusion in subsequent
128 lookups in the parent process. There should also be a search_tidyup() at
129 the end of the subprocess.
131 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
132 only if the host matched helo_try_verify_hosts, which caused the
133 verification to occur when the EHLO/HELO command was issued. The ACL just
134 tested the remembered result. Now, if a previous verification attempt has
135 not happened, "verify = helo" does it there and then.
137 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
139 TK/03 Fix log output including CR from clamd.
141 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
142 exist provoked a memory error which could cause a segfault.
144 PH/15 Installed PCRE 6.2
146 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
148 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
149 of the problem. Specifically, suggested +O2 rather than +O1 for the
152 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
154 PH/20 If a delivery was routed to a non-standard port by means of an SRV
155 record, the port was not correctly logged when the outgoing_port log
156 selector was set (it logged the transort's default port).
158 PH/21 Added support for host-specific ports to manualroute, queryprogram,
159 fallback_hosts, and "hosts" in the smtp transport.
161 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
162 host errors such as "Connection refused".
164 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
165 authentication with radiusclient 0.4.9:
167 - Error returned from rc_read_config was caught wrongly
168 - Username/password not passed on to radius server due to wrong length.
170 The presumption is that some radiusclient API changes for 4.51/PH/17
171 were not taken care of correctly. The code is still untested by me (my
172 Linux distribution still has 0.3.2 of radiusclient), but it was
173 contributed by a Radius user.
175 PH/24 When doing a callout, the value of $domain wasn't set correctly when
176 expanding the "port" option of the smtp transport.
178 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
179 while reading a MIME header. Thanks to Tom Hughes for a patch.
181 PH/24 Include config.h inside local_scan.h so that configuration settings are
184 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
185 that in an ACL for RCPT (for example), you can examine exactly what was
188 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
189 commands, but it was not correctly comparing the address with the actual
190 client host address. Thus, it would show the EHLO address in Received:
191 header lines when this was not necessary.
193 PH/27 Added the % operator to ${eval:}.
195 PH/28 Exim tries to create and chdir to its spool directory when it starts;
196 it should be ignoring failures (because with -C, for example, it has lost
197 privilege). It wasn't ignoring creation failures other than "already
200 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
203 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
204 because an input line was too long, either on its own, or by virtue of
205 too many continuations, the temporary file was not being removed, and the
206 return code was incorrect.
208 PH/31 Missing "BOOL" in function definition in filtertest.c.
210 PH/32 Applied Sieve patches from the maintainer.
212 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
214 PH/33 Added "verify = not_blind".
216 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
217 Local/Makefile (with some defaults set). These are used in built scripts
218 such as exicyclog, but they have never been used in the exim_install
219 script (though there are many overriding facilities there). I have
220 arranged that the exim_install script now takes note of these two
223 PH/35 Installed configuration files for Dragonfly.
225 PH/36 When a locally submitted message by a trusted user did not contain a
226 From: header, and the sender address was obtained from -f or from an SMTP
227 MAIL command, and the trusted user did not use -F to supply a sender
228 name, $originator_name was incorrectly used when constructing a From:
229 header. Furthermore, $originator_name was used for submission mode
230 messages from external hosts without From: headers in a similar way,
231 which is clearly wrong.
233 PH/37 Added control=suppress_local_fixups.
235 PH/38 When log_selector = +received_sender was set, and the addition of the
236 sender made the log line's construction buffer exactly full, or one byte
237 less than full, an overflow happened when the terminating "\n" was
240 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
241 when the result of a list match is failure because a DNS lookup failed.
243 PH/40 RM_COMMAND is now used in the building process.
245 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
246 the "build-* directories that it finds.
248 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
249 address in a domain literal was a prefix of an interface address.
251 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
252 when verifying a sender address, unless rewrite_headers is false.
254 PH/44 Wrote a long comment about why errors_to addresses are verified as
255 recipients, not senders.
257 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
258 the ratelimit ACL was added.
260 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
262 PH/46 Added extra information about PostgreSQL errors to the error string.
264 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
265 avoiding the need to install special zones in a real server. This is
266 backwards compatible; if it can't find the fake resolver, it drops back.
267 Thus, both old and new test suites can be run.
269 TF/02 Added util/ratelimit.pl
271 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
272 clock is set back in time.
274 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
275 Candler <B.Candler@pobox.com>.
277 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
278 OK for addresses that are the result of redirections.
280 PH/48 A number of further additions for the benefit of the new test suite,
281 including a fake gethostbyname() that interfaces to the fake DNS resolver
284 TF/06 The fix for widen_domains has also been applied to qualify_single and
285 search_parents which are the other dnslookup options that can cause
288 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
291 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
293 TF/07 Exim produced the error message "an SRV record indicated no SMTP
294 service" if it encountered an MX record with an empty target hostname.
295 The message is now "an MX or SRV record indicated no SMTP service".
297 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
298 if the DNS of the sending site is misconfigured. This is quite a
299 common situation. This change restores the behaviour of treating a
300 helo verification defer as a failure.
302 PH/51 If self=fail was set on a router, the bounce message did not include the
303 actual error message.
309 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
311 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
312 command itself ended in error, the underlying message about the transport
313 filter timeout was being overwritten with the pipe command error. Now the
314 underlying error message should be appended to the second error message.
316 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
318 PH/02 Reduce the amount of output that "make" produces by default. Full output
319 can still be requested.
321 PH/03 The warning log line about a condition test deferring for a "warn" verb
322 was being output only once per connection, rather than after each
323 occurrence (because it was using the same function as for successful
324 "warn" verbs). This seems wrong, so I have changed it.
326 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
327 it should not have, which might have caused a crash in the right
328 circumstances, but probably never did.
330 PH/04 Installed a modified version of Tony Finch's patch to make submission
331 mode fix the return path as well as the Sender: header line, and to
332 add a /name= option so that you can make the user's friendly name appear
335 TF/03 Added the control = fakedefer ACL modifier.
337 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
338 Mark Lowes for thorough testing.
340 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
342 TK/03 Merged latest SRS patch from Miles Wilton.
344 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
345 with the definition in sysexits.h (which is #included earlier).
346 Fortunately, Exim does not actually use EX_OK. The code used to try to
347 preserve the sysexits.h value, by assumimg that macro definitions were
348 scanned for macro replacements. I have been disabused of this notion,
349 so now the code just undefines EX_OK before #including unistd.h.
351 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
352 in the smtp transport. When a block could not be written in a single
353 write() function, the timeout was being re-applied to each part-write.
354 This seems wrong - if the receiver was accepting one byte at a time it
355 would take for ever. The timeout is now adjusted when this happens. It
356 doesn't have to be particularly precise.
358 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
359 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
361 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
362 without a domain if the check to <postmaster@domain> fails.
364 SC/01 Eximstats: added -xls and the ability to specify output files
365 (patch written by Frank Heydlauf).
367 SC/02 Eximstats: use FileHandles for outputing results.
369 SC/03 Eximstats: allow any combination of xls, txt, and html output.
371 SC/04 Eximstats: fixed display of large numbers with -nvr option
373 SC/05 Eximstats: fixed merging of reports with empty tables.
375 SC/06 Eximstats: added the -include_original_destination flag
377 SC/07 Eximstats: removed tabs and trailing whitespace.
379 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
381 TK/06 MBOX spool code: Add real "From " MBOX separator line
382 so the .eml file is really in mbox format (even though
383 most programs do not really care). Patch from Alex Miller.
385 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
386 The latter is generated from $received_to and is only set if the
387 message has one envelope recipient. SA can use these headers,
388 obviously out-of-the-box. Patch from Alex Miller.
390 PH/08 The ${def test on a variable was returning false if the variable's
391 value was "0", contrary to what the specification has always said!
392 The result should be true unless the variable is empty.
394 PH/09 The syntax error of a character other than { following "${if
395 def:variable_name" (after optional whitespace) was not being diagnosed.
396 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
397 accidental colon was present, for example, could give incorrect results.
399 PH/10 Tidied the code in a number of places where the st_size field of a stat()
400 result is used (not including appendfile, where other changes are about
403 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
404 This involved changing a lot of size variables from int to off_t. It
405 should work with maildirs and everything.
407 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
408 spamd dying while we are connected to it.
410 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
413 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
414 function for BDB 4.3.
416 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
418 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
419 the sources for PCRE have been re-arranged and more files are now
422 PH/15 The code I had for printing potentially long long variables in PH/11
423 above was not the best (it lost precision). The length of off_t variables
424 is now inspected at build time, and an appropriate printing format (%ld
425 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
426 to be "long long int" or "long int". This is needed for the internal
427 formatting function string_vformat().
429 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
430 the configuration file to be ":syslog", then the script "guesses" where
431 the logs files are, rather than using the compiled in default. In our
432 case the guess is not the same as the compiled default, so the script
433 suddenly stopped working when I started to use syslog. The patch checks
434 to see if log_file_path is "". If so, it attempts to read it from exim
435 with no configuration file to get the compiled in version, before it
436 falls back to the previous guessing code."
438 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
439 implementing BATV in an Exim configuration. See NewStuff for the gory
442 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
443 Makefile that are specific to HP-UX.
445 PH/18 If the "use_postmaster" option was set for a recipient callout together
446 with the "random" option, the postmaster address was used as the MAIL
447 FROM address for the random test, but not for the subsequent recipient
448 test. It is now used for both.
450 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
451 patch removes a few documentation additions to RFC 3028, because the
452 latest draft now contains them. It adds the new en;ascii-case comparator
453 and a new error check for 8bit text in MIME parts. Comparator and
454 require names are now matched exactly. I enabled the subaddress
455 extension, but it is not well tested yet (read: it works for me)."
457 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
458 rework some of the code of TK/09 above to avoid the hardwired use of
459 "%lld" and "long long". Replaced the call to snprintf() with a call to
462 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
463 records point to non-existent hosts", "retry timeout exceeded", and
464 "retry time not reached for any host after a long failure period".
466 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
467 experimental DomainKeys support:
469 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
470 (2) On an error such as an illegally used "control", the wrong name for
471 the control was given.
473 These problems did NOT occur unless DomainKeys support was compiled.
475 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
477 PH/24 Added ${if match_ip condition.
479 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
480 databases so that it will be absolutely obvious if a crash occurs in the
481 DB library. This is a regular occurrence (often caused by mis-matched
484 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
485 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
486 picked up on a user's system that detects such things. There doesn't seem
487 to be a gcc warning option for this - only an attribute that has to be
488 put on the function's prototype. It seems that in Fedora Core 4 they have
489 set this on a number of new functions. No doubt there will be more in due
492 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
493 specify a transport. However, if an address that was verified by such a
494 router was the subject of a callout, Exim crashed because it tried to
495 read the rcpt_include_affixes from the non-existent transport. Now it
496 just assumes that the setting of that option is false. This bug was
497 introduced by 4.51/PH/31.
499 PH/28 Changed -d+all to exclude +memory, because that information is very
500 rarely of interest, but it makes the output a lot bigger. People tend to
501 do -d+all out of habit.
503 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
504 code in os-type was giving problems when libc.so lives in lib64, like on
507 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
508 aren't the modern standard, and it seems that some systems' include files
509 don't always have them. Exim was already checking for some of the newer
510 ones like T_AAAA, and defining it itself. I've added checks for all the
511 record types that Exim uses.
513 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
514 not automatically generating a new one, as it is supposed to. This
515 prevented TLS from working. If the file did exist, but contained invalid
516 data, a new version was generated, as expected. It was only the case of a
517 non-existent file that was broken.
519 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
520 with a change in libdomainkeys > 0.64.
522 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
523 from DNS. If the selector record carries the flag, it now has
524 precedence over the domain-wide flag.
526 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
528 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
529 the use of a port name that isn't defined in /etc/services) occurred, the
530 message was deferred as in a normal delivery, and thus remained on the
531 spool, instead of being failed because of the mua_wrapper setting. This
532 is now fixed, and I tidied up some of the mua_wrapper messages at the
535 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
536 the messages in a hash of arrays rather than using individual hashes.
537 This is a bit cleaner and results in dramatic memory savings, albeit
538 at a slight CPU cost.
540 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
541 as requested by Marc Sherman.
543 SC/10 Eximstats: added histograms for user specified patterns as requested
546 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
548 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
549 fopen() in the content-scanning modules that did not already have it.
555 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
556 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
558 TK/02 Fix ACL "control" statement not being available in MIME ACL.
560 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
562 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
563 to test Sieve filters that use "vacation".
565 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
566 that changes the way the GnuTLS parameters are stored in the cache file.
567 The new format can be generated externally. For backward compatibility,
568 if the data in the cache doesn't make sense, Exim assumes it has read an
569 old-format file, and it generates new data and writes a new file. This
570 means that you can't go back to an older release without removing the
573 PH/03 A redirect router that has both "unseen" and "one_time" set does not
574 work if there are any delivery delays because "one_time" forces the
575 parent to be marked "delivered", so its unseen clone is never tried
576 again. For this reason, Exim now forbids the simultaneous setting of
579 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
580 redirected to themselves ("homonym" addresses). Read the long ChangeLog
581 entry if you want to know the details. The fix, however, neglected to
582 consider the case when local delivery batching is involved. The test for
583 "previously delivered" was not happening when checking to see if an
584 address could be batched with a previous (undelivered) one; under
585 certain circumstances this could lead to multiple deliveries to the same
588 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
589 in its include files, and this causes problems building Exim.
591 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
592 header_syntax) but Exim was just ignoring anything given after a slash.
593 In particular, this caused confusion with an attempt to use "verify =
594 reverse_host_lookup/defer_ok". An error is now given when options are
595 supplied for verify items that do not have them. (Maybe reverse_host_
596 lookup should have a defer_ok option, but that's a different point.)
598 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
599 defined by RFC 821) to 2048, because there were problems with some AUTH
600 commands, and RFC 1869 says the size should be increased for extended
601 SMTP commands that take arguments.
603 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
606 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
607 "unknown" error; now it says that the functionality isn't in the binary.
609 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
610 an address' error message when a string expansion fails (syntax or
611 whatever). Otherwise the password may appear in the log. Following change
612 PH/42 below, there is no longer a chance of it appearing in a bounce
615 PH/11 Installed exipick version 20050225.0 from John Jetmore.
617 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
618 of its addresses was ever tried. (Bugzilla bug #2.)
620 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
621 the result incorrectly in the debug output. (It correctly added a newline
622 to what was transported.)
624 TF/01 Added $received_time.
626 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
627 commented out examples of how to interface to a virus scanner and to
628 SpamAssassin. Also added commented examples of av_scanner and
629 spamd_address settings.
631 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
632 and controls are allowed in which ACLs. There were a couple of minor
633 errors. Some of the entries in the conditions table (which is a table of
634 where they are NOT allowed) were getting very unwieldy; rewrote them as a
635 negation of where the condition IS allowed.
637 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
639 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
640 header file does not have a version number, so I've had to invent a new
641 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
642 API. The code is untested by me (my Linux distribution still has 0.3.2 of
643 radiusclient), but it was contributed by a Radius user.
645 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
646 files or directories, for OpenSSL.
648 PH/19 When an Exim process that is running as root has to create an Exim log
649 file, it does so in a subprocess that runs as exim:exim so as to get the
650 ownership right at creation (otherwise, other Exim processes might see
651 the file with the wrong ownership). There was no test for failure of this
652 fork() call, which would lead to the process getting stuck as it waited
653 for a non-existent subprocess. Forks do occasionally fail when resources
654 run out. I reviewed all the other calls to fork(); they all seem to check
657 PH/20 When checking for unexpected SMTP input at connect time (before writing
658 the banner), Exim was not dealing correctly with a non-positive return
659 from the read() function. If the client had disconnected by this time,
660 the result was a log entry for a synchronization error with an empty
661 string after "input=" when read() returned zero. If read() returned -1
662 (an event I could not check), uninitialized data bytes were printed.
663 There were reports of junk text (parts of files, etc) appearing after
666 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
668 PH/22 Added support for macro redefinition, and (re)definition in between
669 driver and ACL definitions.
671 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
672 forgetting to use the resulting value; it was using the unexpanded value.
674 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
675 hadn't been configured. The fix is from Juergen Kreileder, who
676 understands it better than I do:
678 "Here's what I see happening with three configured cyrus_sasl
679 authenticators configured (plain, login, cram-md5):
681 On startup auth_cyrus_sasl_init() gets called for each of these.
682 This means three calls to sasl_listmech() without a specified mech_list.
683 => SASL tests which mechs of all available mechs actually work
684 => three warnings about OTP not working
685 => the returned list contains: plain, login, cram-md5, digest-md5, ...
687 With the patch, sasl_listmech() also gets called three times. But now
688 SASL's mech_list option is set to the server_mech specified in the the
689 authenticator. Or in other words, the answer from sasl_listmech()
690 gets limited to just the mech you're testing for (which is different
692 => the return list contains just 'plain' or 'login', 'cram-md5' or
693 nothing depending on the value of ob->server_mech.
695 I've just tested the patch: Authentication still works fine,
696 unavailable mechs specified in the exim configuration are still
697 caught, and the auth.log warnings about OTP are gone."
699 PH/25 When debugging is enabled, the contents of the command line are added
700 to the debugging output, even when log_selector=+arguments is not
703 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
704 answer is "GNU", and only if the return is "GNU/something" is the answer
707 PH/27 $acl_verify_message is now set immediately after the failure of a
708 verification in an ACL, and so is available in subsequent modifiers. In
709 particular, the message can be preserved by coding like this:
711 warn !verify = sender
712 set acl_m0 = $acl_verify_message
714 Previously, $acl_verify_message was set only while expanding "message"
715 and "log_message" when a very denied access.
717 PH/28 Modified OS/os.c-Linux with
719 -#ifndef OS_LOAD_AVERAGE
720 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
722 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
723 nomenclature these days.)
725 PH/29 Installed patch from the Sieve maintainer that adds the options
726 sieve_useraddress and sieve_subaddress to the redirect router.
728 PH/30 In these circumstances:
729 . Two addresses routed to the same list of hosts;
730 . First host does not offer TLS;
731 . First host accepts first address;
732 . First host gives temporary error to second address;
733 . Second host offers TLS and a TLS session is established;
734 . Second host accepts second address.
735 Exim incorrectly logged both deliveries with the TLS parameters (cipher
736 and peerdn, if requested) that were in fact used only for the second
739 PH/31 When doing a callout as part of verifying an address, Exim was not paying
740 attention to any local part prefix or suffix that was matched by the
741 router that accepted the address. It now behaves in the same way as it
742 does for delivery: the affixes are removed from the local part unless
743 rcpt_include_affixes is set on the transport.
745 PH/32 Add the sender address, as F=<...>, to the log line when logging a
746 timeout during the DATA phase of an incoming message.
748 PH/33 Sieve envelope tests were broken for match types other than :is. I have
749 applied a patch sanctioned by the Sieve maintainer.
751 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
752 the uid or gid is negative. A case of a negative gid caused this to be
753 noticed. The fix allows for either to be negative.
755 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
756 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
759 PH/36 The change PH/12 above was broken. Fixed it.
761 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
762 the grounds that routing the same address twice would always produce the
763 same answer. This might have been true once, but it is certainly no
764 longer true now. Routing a child address may depend on the previous
765 routing that produced that child. Some complicated redirection strategies
766 went wrong when messages had multiple recipients, and made Exim's
767 behaviour dependent on the order in which the addresses were given.
769 I have moved the duplicate checking until after the routing is complete.
770 Exim scans the addresses that are assigned to local and remote
771 transports, and removes any duplicates. This means that more work will be
772 done, as duplicates will always all be routed, but duplicates are
773 presumably rare, so I don't expect this is of any significance.
775 For deliveries to pipes, files, and autoreplies, the duplicate checking
776 still happens during the routing process, since they are not going to be
779 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
780 It corrects a timeout issue with spamd. This is Ian's comment: "The
781 background is that sometimes spamd either never reads data from a
782 connection it has accepted, or it never writes response data. The exiscan
783 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
784 blindly assumes that writes won't block so it may never time out."
786 PH/39 Allow G after quota size as well as K and M.
788 PH/40 The value set for $authenticated_id in an authenticator may not contain
789 binary zeroes or newlines because the value is written to log lines and
790 to spool files. There was no check on this. Now the value is run through
791 the string_printing() function so that such characters are converted to
792 printable escape sequences.
794 PH/41 $message_linecount is a new variable that contains the total number of
795 lines in the message. Compare $body_linecount, which is the count for the
798 PH/42 Exim no longer gives details of delivery errors for specific addresses in
799 bounce and delay warning messages, except in certain special cases, which
802 (a) An SMTP error message from a remote host;
803 (b) A message specified in a :fail: redirection;
804 (c) A message specified in a "fail" command in a system filter;
805 (d) A message specified in a FAIL return from the queryprogram router;
806 (e) A message specified by the cannot_route_message router option.
808 In these cases only, Exim does include the error details in bounce and
809 warning messages. There are also a few cases where bland messages such
810 as "unrouteable address" or "local delivery error" are given.
812 PH/43 $value is now also set for the "else" part of a ${run expansion.
814 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
815 being worked on, but at least Exim now implements the latest version to
818 PH/45 In a pipe transport, although a timeout while waiting for the pipe
819 process to complete was treated as a delivery failure, a timeout while
820 writing the message to the pipe was logged, but erroneously treated as a
821 successful delivery. Such timeouts include transport filter timeouts. For
822 consistency with the overall process timeout, these timeouts are now
823 treated as errors, giving rise to delivery failures by default. However,
824 there is now a new Boolean option for the pipe transport called
825 timeout_defer, which, if set TRUE, converts the failures into defers for
826 both kinds of timeout. A transport filter timeout is now identified in
829 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
830 systems where "make" and "gmake" are different, calling "gmake" at top
831 level broke things. I've arranged for the value of $(MAKE) to be passed
832 from the Makefile to this script so that it can call the same version of
836 A note about Exim versions 4.44 and 4.50
837 ----------------------------------------
839 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
840 changes of various kinds. As a consequence, a big documentation update was
841 needed. This delayed the release for rather longer than seemed good, especially
842 in the light of a couple of (minor) security issues. Therefore, the changes
843 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
844 release. So 4.44 and 4.50 are in effect two different branches that both start
847 I have left the 4.50 change log unchanged; it contains all the changes since
848 4.43. The change log for 4.44 is below; many of its items are identical to
849 those for 4.50. This seems to be the most sensible way to preserve the
850 historical information.
856 1. Minor wording change to the doc/README.SIEVE file.
858 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
859 computation of the current number of files was incorrect.
861 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
862 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
865 4. Give more explanation in the error message when the command for a transport
866 filter fails to execute.
868 5. There are several places where Exim runs a non-Exim command in a
869 subprocess. The SIGUSR1 signal should be disabled for these processes. This
870 was being done only for the command run by the queryprogram router. It is
871 now done for all such subprocesses. The other cases are: ${run, transport
872 filters, and the commands run by the lmtp and pipe transports.
874 6. Added CONFIGURE_GROUP build-time option.
876 7. Some older OS have a limit of 256 on the maximum number of file
877 descriptors. Exim was using setrlimit() to set 1000 as a large value
878 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
879 systems. I've change it so that if it can't get 1000, it tries for 256.
881 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
882 was an oversight, and furthermore, ever since the addition of extra
883 controls (e.g. 4.43/32), the checks on when to allow different forms of
884 "control" were broken. There should now be diagnostics for all cases when a
885 control that does not make sense is encountered.
887 9. Added the /retain_sender option to "control=submission".
889 10. $recipients is now available in the predata ACL (oversight).
891 11. Tidy the search cache before the fork to do a delivery from a message
892 received from the command line. Otherwise the child will trigger a lookup
893 failure and thereby defer the delivery if it tries to use (for example) a
894 cached ldap connection that the parent has called unbind on.
896 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
897 of $address_data from the recipient verification was clobbered by the
900 13. The value of address_data from a sender verification is now available in
901 $sender_address_data in subsequent conditions in the ACL statement.
903 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
905 15. Added a new option "connect=<time>" to callout options, to set a different
908 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
909 was its contents. (It was OK if the option was not defined at all.)
911 17. A "Completed" log line is now written for messages that are removed from
912 the spool by the -Mrm option.
914 18. New variables $sender_verify_failure and $recipient_verify_failure contain
915 information about exactly what failed.
917 19. Added -dd to debug only the daemon process.
919 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
920 handles timeouts, both on the server side and network timeouts. Renamed the
921 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
923 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
924 It is now set to "smtps".
926 22. $host_address is now set to the target address during the checking of
929 23. When checking ignore_target_hosts for an ipliteral router, no host name was
930 being passed; this would have caused $sender_host_name to have been used if
931 matching the list had actually called for a host name (not very likely,
932 since this list is usually IP addresses). A host name is now passed as
935 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
936 code that specifies a non-restarting handler (typically sigaction() in
937 modern systems) in an attempt to fix a rare and obscure crash bug.
939 25. Narrowed the window for a race in the daemon that could cause it to ignore
940 SIGCHLD signals. This is not a major problem, because they are used only to
941 wake it up if nothing else does.
943 26. A malformed maildirsize file could cause Exim to calculate negative values
944 for the mailbox size or file count. Odd effects could occur as a result.
945 The maildirsize information is now recalculated if the size or filecount
948 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
949 support for a long time. Removed HAVE_SYS_VFS_H.
951 28. Installed the latest version of exipick from John Jetmore.
953 29. In an address list, if the pattern was not a regular expression, an empty
954 subject address (from a bounce message) matched only if the pattern was an
955 empty string. Non-empty patterns were not even tested. This was the wrong
956 because it is perfectly reasonable to use an empty address as part of a
957 database query. An empty address is now tested by patterns that are
958 lookups. However, all the other forms of pattern expect the subject to
959 contain a local part and a domain, and therefore, for them, an empty
960 address still always fails if the pattern is not itself empty.
962 30. Exim went into a mad DNS loop when attempting to do a callout where the
963 host was specified on an smtp transport, and looking it up yielded more
966 31. Re-factored the code for checking spool and log partition space into a
967 function that finds that data and another that does the check. The former
968 is then used to implement four new variables: $spool_space, $log_space,
969 $spool_inodes, and $log_inodes.
971 32. The RFC2047 encoding function was originally intended for short strings
972 such as real names; it was not keeping to the 75-character limit for
973 encoded words that the RFC imposes. It now respects the limit, and
974 generates multiple encoded words if necessary. To be on the safe side, I
975 have increased the buffer size for the ${rfc2047: expansion operator from
978 33. It is now permitted to omit both strings after an "if" condition; if the
979 condition is true, the result is "true". As before, when the second string
980 is omitted, a false condition yields an empty string. This makes it less
981 cumbersome to write custom ACL and router conditions.
983 34. Failure to deliver a bounce message always caused it to be frozen, even if
984 there was an errors_to setting on the router. The errors_to setting is now
987 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
988 canonical form (fully expanded) before being placed in
989 $sender_host_address.
991 36. The table in the code that translates DNS record types into text (T_A to
992 "A" for instance) was missing entries for NS and CNAME. It is just possible
993 that this could have caused confusion if both these types were looked up
994 for the same domain, because the text type is used as part of Exim's
995 per-process caching. But the chance of anyone hitting this buglet seems
998 37. The dnsdb lookup has been extended in a number of ways.
1000 (1) There is a new type, "zns", which walks up the domain tree until it
1001 finds some nameserver records. It should be used with care.
1003 (2) There is a new type, "mxh", which is like "mx" except that it returns
1004 just the host names, not the priorities.
1006 (3) It is now possible to give a list of domains (or IP addresses) to be
1007 looked up. The behaviour when one of the lookups defers can be
1008 controlled by a keyword.
1010 (4) It is now possible to specify the separator character for use when
1011 multiple records are returned.
1013 38. The dnslists ACL condition has been extended: it is now possible to supply
1014 a list of IP addresses and/or domains to be looked up in a particular DNS
1017 39. Added log_selector=+queue_time_overall.
1019 40. When running the queue in the test harness, wait just a tad after forking a
1020 delivery process, to get repeatability of debugging output.
1022 41. Include certificate and key file names in error message when GnuTLS fails
1023 to set them up, because the GnuTLS error message doesn't include the name
1024 of the failing file when there is a problem reading it.
1026 42. Allow both -bf and -bF in the same test run.
1028 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1030 44. The "Exiscan patch" is now merged into the mainline Exim source.
1032 45. Sometimes the final signoff response after QUIT could fail to get
1033 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1034 < 0 before doing a fflush(). This bug looks as though it goes back to the
1035 introduction of TLS in release 3.20, but "sometimes" must have been rare
1036 because the tests only now provoked it.
1038 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1039 (this can affect the format of dates).
1041 47. exim_tidydb, when checking for the continued existence of a message for
1042 which it has found a message-specific retry record, was not finding
1043 messages that were in split spool directories. Consequently, it was
1044 deleting retry records that should have stayed in existence.
1046 48. Steve fixed some bugs in eximstats.
1048 49. The SPA authentication driver was not abandoning authentication and moving
1049 on to the next authenticator when an expansion was forced to fail,
1050 contradicting the general specification for all authenticators. Instead it
1051 was generating a temporary error. It now behaves as specified.
1053 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1054 (the order specifies the preference for clients). The order is now AES256,
1055 AES128, 3DES, ARCFOUR128.
1057 51. Small patch to Sieve code - explicitly set From: when generating an
1060 52. Exim crashed if a remote delivery caused a very long error message to be
1061 recorded - for instance if somebody sent an entire SpamAssassin report back
1062 as a large number of 550 error lines. This bug was coincidentally fixed by
1063 increasing the size of one of Exim's internal buffers (big_buffer) that
1064 happened as part of the Exiscan merge. However, to be on the safe side, I
1065 have made the code more robust (and fixed the comments that describe what
1068 53. Now that there can be additional text after "Completed" in log lines (if
1069 the queue_time_overall log selector is set), a one-byte patch to exigrep
1070 was needed to allow it to recognize "Completed" as not the last thing in
1073 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1074 patch that reportedly fixes this has been added. I am not expert enough to
1075 create a test for it. This is what the patch creator wrote:
1077 "I found a little strange behaviour of ldap code when working with
1078 Windows 2003 AD Domain, where users was placed in more than one
1079 Organization Units. When I tried to give exim partial DN, the exit code
1080 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1081 But simultaneously result of request was absolutely normal ldap result,
1082 so I produce this patch..."
1084 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1085 REFERENCE, so I have modified the code to exclude the patch when that macro
1088 55. Some experimental protocols are using DNS PTR records for new purposes. The
1089 keys for these records are domain names, not reversed IP addresses. The
1090 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1091 leaves it alone. Component reversal etc. now happens only for IP addresses.
1094 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1096 57. Double the size of the debug message buffer (to 2048) so that more of very
1097 long debug lines gets shown.
1099 58. The exicyclog utility now does better if the number of log files to keep
1100 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1102 59. Two changes related to the smtp_active_hostname option:
1104 (1) $smtp_active_hostname is now available as a variable.
1105 (2) The default for smtp_banner uses $smtp_active_hostname instead
1106 of $primary_hostname.
1108 60. The host_aton() function is supposed to be passed a string that is known
1109 to be a valid IP address. However, in the case of IPv6 addresses, it was
1110 not checking this. This is a hostage to fortune. Exim now panics and dies
1111 if the condition is not met. A case was found where this could be provoked
1112 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1113 components; fortuitously, this particular loophole had already been fixed
1114 by change 4.50/55 above.
1116 If there are any other similar loopholes, the new check in host_aton()
1117 itself should stop them being exploited. The report I received stated that
1118 data on the command line could provoke the exploit when Exim was running as
1119 exim, but did not say which command line option was involved. All I could
1120 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1121 running as the user.
1124 61. There was a buffer overflow vulnerability in the SPA authentication code
1125 (which came originally from the Samba project). I have added a test to the
1126 spa_base64_to_bits() function which I hope fixes it.
1129 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1130 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1132 63. The daemon start-up calls getloadavg() while still root for those OS that
1133 need the first call to be done as root, but it missed one case: when
1134 deliver_queue_load_max is set with deliver_drop_privilege. This is
1135 necessary for the benefit of the queue runner, because there is no re-exec
1136 when deliver_drop_privilege is set.
1138 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1139 This has been fixed.
1141 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1142 was in use, was not putting the data itself into the right store pool;
1143 consequently, it could be overwritten for a subsequent message in the same
1144 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1147 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1149 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1150 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1151 the function were treating the return as a boolean value, which happened to
1152 work because 0=false and not-0=true, but is not correct code.
1154 68. The host_aton() function was not handling scoped IPv6 addresses (those
1155 with, for example, "%eth0" on the end) correctly.
1157 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1158 negated items (that is, ~something) in unsigned ints. Some compilers
1159 apparently mutter when there is no cast.
1161 70. If an address verification called from an ACL failed, and did not produce a
1162 user-specific message (i.e. there was only a "system" message), nothing was
1163 put in $acl_verify_message. In this situation, it now puts the system
1166 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1167 session; change 4.31/43 added the unwanted input to the log line - except
1168 that it did not do this in the start of session case. It now does.
1170 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1171 This is wrong and can cause the other end to generate a synchronization
1172 error if it is another Exim or anything else that does the synchronization
1173 check. A QUIT command is no longer sent after a timeout.
1175 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1176 during host lookups.
1178 74. The defer_ok option of callout verification was not working if it was used
1179 when verifying addresses in header lines, that is, for this case:
1181 verify = header_sender/callout=defer_ok
1183 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1184 those file descriptors could be used for SMTP connections. If anything
1185 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1186 could be sent to the SMTP client, causing chaos. The daemon now opens
1187 stdin, stdout, and stderr to /dev/null when it puts itself into the
1190 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1191 log by default. The user can override this with suitable Perl magic.
1193 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1194 the log message when discard triggers, was not working for the DATA ACL or
1195 for the non-SMTP ACL.
1197 78. Error message wording change in sieve.c.
1199 79. If smtp_accept_max_per_host was set, the number of connections could be
1200 restricted to fewer than expected, because the daemon was trying to set up
1201 a new connection before checking whether the processes handling previous
1202 connections had finished. The check for completed processes is now done
1203 earlier. On busy systems, this bug wouldn't be noticed because something
1204 else would have woken the daemon, and it would have reaped the completed
1207 80. If a message was submitted locally by a user whose login name contained one
1208 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1209 It caused a spool format error. I have fixed the spool reading code. A
1210 related problem was that the "from" clause in the Received: line became
1211 illegal because of the space(s). It is now covered by ${quote_local_part.
1213 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1216 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1218 83. Patch from Sieve maintainer to fix unterminated string problem in
1219 "vacation" handling.
1221 84. Some minor changes to the Linux configuration files to help with other
1222 OS variants using glibc.
1224 85. One more patch for Sieve to update vacation handling to latest spec.
1227 ----------------------------------------------------
1228 See the note above about the 4.44 and 4.50 releases.
1229 ----------------------------------------------------
1235 1. Change 4.43/35 introduced a bug that caused file counts to be
1236 incorrectly computed when quota_filecount was set in an appendfile
1239 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1240 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1243 3. Give more explanation in the error message when the command for a transport
1244 filter fails to execute.
1246 4. There are several places where Exim runs a non-Exim command in a
1247 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1248 was being done only for the command run by the queryprogram router. It is
1249 now done for all such subprocesses. The other cases are: ${run, transport
1250 filters, and the commands run by the lmtp and pipe transports.
1252 5. Some older OS have a limit of 256 on the maximum number of file
1253 descriptors. Exim was using setrlimit() to set 1000 as a large value
1254 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1255 systems. I've change it so that if it can't get 1000, it tries for 256.
1257 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1258 was an oversight, and furthermore, ever since the addition of extra
1259 controls (e.g. 4.43/32), the checks on when to allow different forms of
1260 "control" were broken. There should now be diagnostics for all cases when a
1261 control that does not make sense is encountered.
1263 7. $recipients is now available in the predata ACL (oversight).
1265 8. Tidy the search cache before the fork to do a delivery from a message
1266 received from the command line. Otherwise the child will trigger a lookup
1267 failure and thereby defer the delivery if it tries to use (for example) a
1268 cached ldap connection that the parent has called unbind on.
1270 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1271 of $address_data from the recipient verification was clobbered by the
1272 sender verification.
1274 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1275 was its contents. (It was OK if the option was not defined at all.)
1277 11. A "Completed" log line is now written for messages that are removed from
1278 the spool by the -Mrm option.
1280 12. $host_address is now set to the target address during the checking of
1281 ignore_target_hosts.
1283 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1284 being passed; this would have caused $sender_host_name to have been used if
1285 matching the list had actually called for a host name (not very likely,
1286 since this list is usually IP addresses). A host name is now passed as
1289 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1290 code that specifies a non-restarting handler (typically sigaction() in
1291 modern systems) in an attempt to fix a rare and obscure crash bug.
1293 15. Narrowed the window for a race in the daemon that could cause it to ignore
1294 SIGCHLD signals. This is not a major problem, because they are used only to
1295 wake it up if nothing else does.
1297 16. A malformed maildirsize file could cause Exim to calculate negative values
1298 for the mailbox size or file count. Odd effects could occur as a result.
1299 The maildirsize information is now recalculated if the size or filecount
1302 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1303 support for a long time. Removed HAVE_SYS_VFS_H.
1305 18. Updated exipick to current release from John Jetmore.
1307 19. Allow an empty sender to be matched against a lookup in an address list.
1308 Previously the only cases considered were a regular expression, or an
1311 20. Exim went into a mad DNS lookup loop when doing a callout where the
1312 host was specified on the transport, if the DNS lookup yielded more than
1315 21. The RFC2047 encoding function was originally intended for short strings
1316 such as real names; it was not keeping to the 75-character limit for
1317 encoded words that the RFC imposes. It now respects the limit, and
1318 generates multiple encoded words if necessary. To be on the safe side, I
1319 have increased the buffer size for the ${rfc2047: expansion operator from
1322 22. Failure to deliver a bounce message always caused it to be frozen, even if
1323 there was an errors_to setting on the router. The errors_to setting is now
1326 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1327 canonical form (fully expanded) before being placed in
1328 $sender_host_address.
1330 24. Updated eximstats to version 1.33
1332 25. Include certificate and key file names in error message when GnuTLS fails
1333 to set them up, because the GnuTLS error message doesn't include the name
1334 of the failing file when there is a problem reading it.
1336 26. Expand error message when OpenSSL has problems setting up cert/key files.
1339 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1340 (this can affect the format of dates).
1342 28. exim_tidydb, when checking for the continued existence of a message for
1343 which it has found a message-specific retry record, was not finding
1344 messages that were in split spool directories. Consequently, it was
1345 deleting retry records that should have stayed in existence.
1347 29. eximstats updated to version 1.35
1348 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1349 1.35 - bugfix such that pie charts by volume are generated correctly
1351 30. The SPA authentication driver was not abandoning authentication and moving
1352 on to the next authenticator when an expansion was forced to fail,
1353 contradicting the general specification for all authenticators. Instead it
1354 was generating a temporary error. It now behaves as specified.
1356 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1357 (the order specifies the preference for clients). The order is now AES256,
1358 AES128, 3DES, ARCFOUR128.
1360 31. Small patch to Sieve code - explicitly set From: when generating an
1363 32. Exim crashed if a remote delivery caused a very long error message to be
1364 recorded - for instance if somebody sent an entire SpamAssassin report back
1365 as a large number of 550 error lines. This bug was coincidentally fixed by
1366 increasing the size of one of Exim's internal buffers (big_buffer) that
1367 happened as part of the Exiscan merge. However, to be on the safe side, I
1368 have made the code more robust (and fixed the comments that describe what
1371 33. Some experimental protocols are using DNS PTR records for new purposes. The
1372 keys for these records are domain names, not reversed IP addresses. The
1373 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1374 leaves it alone. Component reversal etc. now happens only for IP addresses.
1377 34. The host_aton() function is supposed to be passed a string that is known
1378 to be a valid IP address. However, in the case of IPv6 addresses, it was
1379 not checking this. This is a hostage to fortune. Exim now panics and dies
1380 if the condition is not met. A case was found where this could be provoked
1381 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1382 components; fortuitously, this particular loophole had already been fixed
1383 by change 4.50/55 or 4.44/33 above.
1385 If there are any other similar loopholes, the new check in host_aton()
1386 itself should stop them being exploited. The report I received stated that
1387 data on the command line could provoke the exploit when Exim was running as
1388 exim, but did not say which command line option was involved. All I could
1389 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1390 running as the user.
1393 35. There was a buffer overflow vulnerability in the SPA authentication code
1394 (which came originally from the Samba project). I have added a test to the
1395 spa_base64_to_bits() function which I hope fixes it.
1398 36. The daemon start-up calls getloadavg() while still root for those OS that
1399 need the first call to be done as root, but it missed one case: when
1400 deliver_queue_load_max is set with deliver_drop_privilege. This is
1401 necessary for the benefit of the queue runner, because there is no re-exec
1402 when deliver_drop_privilege is set.
1404 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1405 was in use, was not putting the data itself into the right store pool;
1406 consequently, it could be overwritten for a subsequent message in the same
1407 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1410 38. Sometimes the final signoff response after QUIT could fail to get
1411 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1412 < 0 before doing a fflush(). This bug looks as though it goes back to the
1413 introduction of TLS in release 3.20, but "sometimes" must have been rare
1414 because the tests only now provoked it.
1420 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1421 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1422 Now it returns an integer. A number of calls were still expecting a T/F
1423 return. Fortuitously, in all cases, the tests worked in OK situations,
1424 which is the norm. However, things would have gone wrong on any write
1425 failures on the smtp file descriptor. This function is used when sending
1426 messages over SMTP and also when doing verify callouts.
1428 2. When Exim is called to do synchronous delivery of a locally submitted
1429 message (the -odf or -odi options), it no longer closes stderr before doing
1432 3. Implemented the mua_wrapper option.
1434 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1436 5. Implemented the functions header_remove(), header_testname(),
1437 header_add_at_position(), and receive_remove_recipient(), and exported them
1440 6. If an ACL "warn" statement specified the addition of headers, Exim already
1441 inserted X-ACL-Warn: at the start if there was no header name. However, it
1442 was not making this test for the second and subsequent header lines if
1443 there were newlines in the string. This meant that an invalid header could
1444 be inserted if Exim was badly configured.
1446 7. Allow an ACL "warn" statement to add header lines at the start or after all
1447 the Received: headers, as well as at the end.
1449 8. Added the rcpt_4xx retry error code.
1451 9. Added postmaster_mailfrom=xxx to callout verification option.
1453 10. Added mailfrom=xxxx to the callout verification option, for verify=
1456 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1457 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1459 12. Inserted some casts to stop certain compilers warning when using pointer
1460 differences as field lengths or precisions in printf-type calls (mostly
1461 affecting debugging statements).
1463 13. Added optional readline() support for -be (dynamically loaded).
1465 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1466 same clock tick as a message's arrival, so that its received time was the
1467 same as the "first fail" time on the retry record, and that message
1468 remained on the queue past the ultimate address timeout, every queue runner
1469 would try a delivery (because it was past the ultimate address timeout) but
1470 after another failure, the ultimate address timeout, which should have then
1471 bounced the address, did not kick in. This was a "< instead of <=" error;
1472 in most cases the first failure would have been in the next clock tick
1473 after the received time, and all would be well.
1475 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1476 being recognized when the domain list was tested by the match_domain
1477 condition in an expansion string.
1479 16. Added the ${str2b64: operator.
1481 17. Exim was always calling setrlimit() to set a large limit for the number of
1482 processes, without checking whether the existing limit was already
1483 adequate. (It did check for the limit on file descriptors.) Furthermore,
1484 errors from getrlimit() and setrlimit() were being ignored. Now they are
1485 logged to the main and panic logs, but Exim does carry on, to try to do its
1486 job under whatever limits there are.
1488 18. Imported PCRE 5.0.
1490 19. Trivial typo in log message " temporarily refused connection" (the leading
1493 20. If the log selector return_path_on_delivery was set and an address was
1494 redirected to /dev/null, the delivery process crashed because it assumed
1495 that a return path would always be set for a "successful" delivery. In this
1496 case, the whole delivery is bypassed as an optimization, and therefore no
1499 21. Internal re-arrangement: the function for sending a challenge and reading
1500 a response while authentication was assuming a zero-terminated challenge
1501 string. It's now changed to take a pointer and a length, to allow for
1502 binary data in such strings.
1504 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1506 23. Exim was not respecting finduser_retries when seeking the login of the
1507 uid under which it was called; it was always trying 10 times. (The default
1508 setting of finduser_retries is zero.) Also, it was sleeping after the final
1509 failure, which is pointless.
1511 24. Implemented tls_on_connect_ports.
1513 25. Implemented acl_smtp_predata.
1515 26. If the domain in control=submission is set empty, Exim assumes that the
1516 authenticated id is a complete email address when it generates From: or
1517 Sender: header lines.
1519 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1520 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1521 chown and chgrp in /bin and hostname in /usr/bin.
1523 28. Exim was keeping the "process log" file open after each use, just as it
1524 does for the main log. This opens the possibility of it remaining open for
1525 long periods when the USR1 signal hits a daemon. Occasional processlog
1526 errors were reported, that could have been caused by this. Anyway, it seems
1527 much more sensible not to leave this file open at all, so that is what now
1530 29. The long-running daemon process does not normally write to the log once it
1531 has entered its main loop, and it closes the log before doing so. This is
1532 so that log files can straightforwardly be renamed and moved. However,
1533 there are a couple of unusual error situations where the daemon does write
1534 log entries, and I had neglected to close the log afterwards.
1536 30. The text of an SMTP error response that was received during a remote
1537 delivery was being truncated at 512 bytes. This is too short for some of
1538 the long messages that one sometimes sees. I've increased the limit to
1541 31. It is now possible to make retry rules that apply only when a message has a
1542 specific sender, in particular, an empty sender.
1544 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1545 it possible to be selective about when SMTP synchronization is enforced.
1547 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1549 32. Implemented hosts_connection_nolog.
1551 33. Added an ACL for QUIT.
1553 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1556 35. Added mailbox_size and mailbox_filecount to appendfile.
1558 36. Added control = no_multiline_responses to ACLs.
1560 37. There was a bug in the logic of the code that waits for the clock to tick
1561 in the case where the clock went backwards by a substantial amount such
1562 that the microsecond fraction of "now" was more than the microsecond
1563 fraction of "then" (but the whole seconds number was less).
1565 38. Added support for the libradius Radius client library this is found on
1566 FreeBSD (previously only the radiusclient library was supported).
1572 1. When certain lookups returned multiple values in the form name=value, the
1573 quoting of the values was not always being done properly. Specifically:
1574 (a) If the value started with a double quote, but contained no whitespace,
1576 (b) If the value contained whitespace other than a space character (i.e.
1577 tabs or newlines or carriage returns) it was not quoted.
1578 This fix has been applied to the mysql and pgsql lookups by writing a
1579 separate quoting function and calling it from the lookup code. The fix
1580 should probably also be applied to nisplus, ibase and oracle lookups, but
1581 since I cannot test any of those, I have not disturbed their existing code.
1583 2. A hit in the callout cache for a specific address caused a log line with no
1584 reason for rejecting RCPT. Now it says "Previous (cached) callout
1585 verification failure".
1587 3. There was an off-by-one bug in the queryprogram router. An over-long
1588 return line was truncated at 256 instead of 255 characters, thereby
1589 overflowing its buffer with the terminating zero. As well as fixing this, I
1590 have increased the buffer size to 1024 (and made a note to document this).
1592 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1593 when Exim is waiting for an SMTP response from a remote server, Exim
1594 restarts its select() call on the socket, thereby resetting its timeout.
1595 This is not a problem when such interrupts are rare. Somebody set up a cron
1596 job to run exiwhat every 2 minutes, which is less than the normal select()
1597 timeout (5 or 10 minutes). This meant that the select() timeout never
1598 kicked in because it was always reset. I have fixed this by comparing the
1599 time when an interrupt arrives with the time at the start of the first call
1600 to select(). If more time than the timeout has elapsed, the interrupt is
1601 treated as a timeout.
1603 5. Some internal re-factoring in preparation for the addition of Sieve
1604 extensions (by MH). In particular, the "personal" test is moved to a
1605 separate function, and given an option for scanning Cc: and Bcc: (which is
1606 not set for Exim filters).
1608 6. When Exim created an email address using the login of the caller as the
1609 local part (e.g. when creating a From: or Sender: header line), it was not
1610 quoting the local part when it contained special characters such as @.
1612 7. Installed new OpenBSD configuration files.
1614 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1615 try to make them clearer.
1617 9. Callout options, other than the timeout value, were being ignored when
1618 verifying sender addresses in header lines. For example, when using
1620 verify = header_sender/callout=no_cache
1622 the cache was (incorrectly) being used.
1624 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1625 only the Cygwin environment.
1627 11. When return_path_on_delivery was set as a log selector, if different remote
1628 addresses in the same message used different return paths and parallel
1629 remote delivery occurred, the wrong values would sometimes be logged.
1630 (Whenever a remote delivery process finished, the return path value from
1631 the most recently started remote delivery process was logged.)
1633 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1634 lines when AUTH and/or TLS are in use. This is the "received protocol"
1635 field. Exim used to use "asmtp" for authenticated SMTP, without any
1636 indication (in the protocol name) for TLS use. Now it follows the RFC and
1637 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1638 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1639 names appear in log lines as well as in Received: header lines.
1641 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1642 extensions, and comparison tests, and to fix some bugs.
1644 14. Changes to the "personal" filter test:
1646 (1) The test was buggy in that it was just doing the equivalent of
1647 "contains" tests on header lines. For example, if a user's address was
1648 anne@some.where, the "personal" test would incorrectly be true for
1650 To: susanne@some.where
1652 This test is now done by extracting each address from the header in turn,
1653 and checking the entire address. Other tests that are part of "personal"
1654 are now done using regular expressions (for example, to check local parts
1655 of addresses in From: header lines).
1657 (2) The list of non-personal local parts in From: addresses has been
1658 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1659 taken from the Sieve specification recommendations.
1661 (3) If the message contains any header line starting with "List-" it is
1662 treated as non-personal.
1664 (4) The test for "circular" in the Subject: header line has been removed
1665 because it now seems ill-conceived.
1667 15. Minor typos in src/EDITME comments corrected.
1669 16. Installed latest exipick from John Jetmore.
1671 17. If headers_add on a router specified a text string that was too long for
1672 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1673 of string_sprintf() is now avoided.
1675 18. $message_body_size was not set (it was always zero) when running the DATA
1676 ACL and the local_scan() function.
1678 19. For the "mail" command in an Exim filter, no default was being set for
1679 the once_repeat time, causing a random time value to be used if "once" was
1680 specified. (If the value happened to be <= 0, no repeat happened.) The
1681 default is now 0s, meaning "never repeat". The "vacation" command was OK
1682 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1683 (I found it when inspecting the code).
1685 20. There is now an overall timeout for performing a callout verification. It
1686 defaults to 4 times the callout timeout, which applies to individual SMTP
1687 commands during the callout. The overall timeout applies when there is more
1688 than one host that can be tried. The timeout is checked before trying the
1689 next host. This prevents very long delays if there are a large number of
1690 hosts and all are timing out (e.g. when the network connections are timing
1691 out). The value of the overall timeout can be changed by specifying an
1692 additional sub-option for "callout", called "maxwait". For example:
1694 verify = sender/callout=5s,maxwait=20s
1696 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1697 to the end before writing, but this should make it even safer).
1699 22. Exim was forgetting that it had advertised PIPELINING for the second and
1700 subsequent messages on an SMTP connection. It was also not resetting its
1701 memory on STARTTLS and an internal HELO.
1703 23. When Exim logs an SMTP synchronization error within a session, it now
1704 records whether PIPELINING has been advertised or not.
1706 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1707 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1708 rather than long int.
1710 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1712 26. Added the never_mail option to autoreply.
1718 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1719 crash if the getsockname() call failed; this can happen if a connection is
1720 closed very soon after it is established. The problem was simply in the
1721 order in which certain operations were done, causing Exim to try to write
1722 to the SMTP stream before it had set up the file descriptor. The bug has
1723 been fixed by making things happen in the correct order.
1729 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1730 before the connection was closed, thus losing the rejection response.
1732 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1733 some early Solaris releases, but causes trouble in current releases where
1734 socklen_t is defined.
1736 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1739 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1742 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1743 that releasing the top of it at the end releases what was used for sub-
1744 expansions (unless the block got too big). However, discard this block if
1745 the first thing is a variable or header, so that we can use its block when
1746 it is dynamic (useful for very large $message_headers, for example).
1748 6. Lookups now cache *every* query, not just the most recent. A new, separate
1749 store pool is used for this. It can be recovered when all lookup caches are
1750 flushed. Lookups now release memory at the end of their result strings.
1751 This has involved some general refactoring of the lookup sources.
1753 7. Some code has been added to the store_xxx() functions to reduce the amount
1754 of flapping under certain conditions.
1756 8. log_incoming_interface used to affect only the <= reception log lines. Now
1757 it causes the local interface and port to be added to several more SMTP log
1758 lines, for example "SMTP connection from", and rejection lines.
1760 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1762 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1764 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1765 could be overwritten at the end of the current message (or the start of a
1766 new message if it was set in a HELO ACL). The value is now preserved for
1767 the duration of the SMTP connection.
1769 12. If a transport had a headers_rewrite setting, and a matching header line
1770 contained an unqualified address, that address was qualified, even if it
1771 did not match any rewriting rules. The underlying bug was that the values
1772 of the flags that permit the existence of unqualified sender and recipient
1773 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1774 non-local messages, and by -bnq for local messages) were not being
1775 preserved with the message after it was received.
1777 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1778 "next input=" as part of the text comprising the host identity instead of
1779 the correct text. The code was using the same buffer for two different
1780 strings. However, depending on which order the printing function evaluated
1781 its arguments, the bug did not always show up. Under Linux, for example, my
1782 test suite worked just fine.
1784 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1785 doesn't work with some older versions of Perl. It has been changed to "my",
1786 which in any case is probably the better facility to use.
1788 15. A really picky compiler found some instances of statements for creating
1789 error messages that either had too many or two few arguments for the format
1792 16. The size of the buffer for calls to the DNS resolver has been increased
1793 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1794 for addresses that have a lot of PTR records. This alleviates a problem; it
1795 does not fully solve it.
1797 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1798 buffer now truncates the list and logs the incident, which is the same
1799 action as happens when Exim is looking up a host name and its aliases.
1800 Previously in this situation something unpredictable would happen;
1801 sometimes it was "internal error: store_reset failed".
1803 18. If a server dropped the connection unexpectedly when an Exim client was
1804 using GnuTLS and trying to read a response, the client delivery process
1805 crashed while trying to generate an error log message.
1807 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1808 string, for example:
1810 warn message = H1: something\nH2: something
1812 the text was added as a single header line from Exim's point of view
1813 though it ended up OK in the delivered message. However, searching for the
1814 second and subsequent header lines using $h_h2: did not work. This has been
1815 fixed. Similarly, if a system filter added multiple headers in this way,
1816 the routers could not see them.
1818 20. Expanded the error message when iplsearch is called with an invalid key to
1819 suggest using net-iplsearch in a host list.
1821 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1822 ACLs are no longer actually imposed (and a message to that effect is
1825 22. If a "gecos" field in a passwd entry contained escaped characters, in
1826 particular, if it contained a \" sequence, Exim got it wrong when building
1827 a From: or a Sender: header from that name. A second bug also caused
1828 incorrect handling when an unquoted " was present following a character
1829 that needed quoting.
1831 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1832 was not being matched caselessly.
1834 24. Arranged for all hyphens in the exim.8 source to be escaped with
1837 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1838 properly. Recipient callout cache records were still being keyed to include
1839 the sender, even when use_sender was set false. This led to far more
1840 callouts that were necessary. The sender is no longer included in the key
1841 when use_sender is false.
1843 26. Added "control = submission" modifier to ACLs.
1845 27. Added the ${base62d: operator to decode base 62 numbers.
1847 28. dnsdb lookups can now access SRV records.
1849 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1850 the configuration file.
1852 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1853 (-v) mode. This makes the output for a verbose queue run more intelligible.
1855 31. Added a use_postmaster feature to recipient callouts.
1857 32. Added the $body_zerocount variable, containing the number of binary zero
1858 bytes in the message body.
1860 33. The time of last modification of the "new" subdirectory is now used as the
1861 "mailbox time last read" when there is a quota error for a maildir
1864 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1866 35. Added +ignore_unknown as a special item in host lists.
1868 36. Code for decoding IPv6 addresses in host lists is now included, even if
1869 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1870 address was recognized as an IP address, but was then not correctly decoded
1871 into binary, causing unexpected and incorrect effects when compared with
1878 1. Very minor rewording of debugging text in manualroute to say "list of
1879 hosts" instead of "hostlist".
1881 2. If verify=header_syntax was set, and a header line with an unqualified
1882 address (no domain) and a large number of spaces between the end of the
1883 name and the colon was received, the reception process suffered a buffer
1884 overflow, and (when I tested it) crashed. This was caused by some obsolete
1885 code that should have been removed. The fix is to remove it!
1887 3. When running in the test harness, delay a bit after writing a bounce
1888 message to get a bit more predictability in the log output.
1890 4. Added a call to search_tidyup() just before forking a reception process. In
1891 theory, someone could use a lookup in the expansion of smtp_accept_max_
1892 per_host which, without the tidyup, could leave open a database connection.
1894 5. Added the variables $recipient_data and $sender_data which get set from a
1895 lookup success in an ACL "recipients" or "senders" condition, or a router
1896 "senders" option, similar to $domain_data and $local_part_data.
1898 6. Moved the writing of debug_print from before to after the "senders" test
1901 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1902 problems for message scanning, either using a data ACL, or using
1903 local_scan() because the Received: header was not generated till after they
1904 were called (in order to set the time as the time of reception completion).
1905 I have revised the way this works. The header is now generated after the
1906 body is received, but before the ACL or local_scan() are called. After they
1907 are run, the timestamp in the header is updated.
1913 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1914 before starting a queue runner without re-exec. This happened only when
1915 deliver_drop_privilege was set or when the Exim user was set to root. The
1916 effect of the bug was that timeouts during subsequent deliveries caused
1917 crashes instead of being properly handled. The handler is now left at its
1918 default (and expected) setting.
1920 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1921 message, again when deliver_drop_privilege is set or Exim is run as root.
1922 The bug described in (1) was not present in this case, but the tidying up
1923 of the other signals was missing. I have made the two cases consistent.
1925 3. The ignore_target_hosts setting on a manualroute router was being ignored
1926 for hosts that were looked up using the /MX notation.
1928 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1931 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1932 operated on the sender address. After changing the $sender_address to <>
1933 for the sender address verify, Exim was re-instated it as the original
1934 (before rewriting) address, but remembering that it had rewritten it, so it
1935 wasn't rewriting it again. This bug also had the effect of breaking the
1936 sender address verification caching when the sender address was rewritten.
1938 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1939 This has been changed so that if the ip literal address matches
1940 ignore_target_hosts, the router declines.
1942 7. Added expansion conditions match_domain, match_address, and match_local_
1943 part (NOT match_host).
1945 8. The placeholder for the Received: header didn't have a length field set.
1947 9. Added code to Exim itself and to exim_lock to test for a specific race
1948 condition that could lead to file corruption when using MBX delivery. The
1949 issue is with the lockfile that is created in /tmp. If this file is removed
1950 after a process has opened it but before that process has acquired a lock,
1951 there is the potential for a second process to recreate the file and also
1952 acquire a lock. This could lead to two Exim processes writing to the file
1953 at the same time. The added code performs the same test as UW imapd; it
1954 checks after acquiring the lock that its file descriptor still refers to
1955 the same named file.
1957 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1958 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1961 11. Added the smtp_active_hostname option. If used, this will typically be made
1962 to depend on the incoming interface address. Because $interface_address is
1963 not set up until the daemon has forked a reception process, error responses
1964 that can happen earlier (such as "too many connections") no longer contain
1967 12. If an expansion in a condition on a "warn" statement fails because a lookup
1968 defers, the "warn" statement is abandoned, and the next ACL statement is
1969 processed. Previously this caused the whole ACL to be aborted.
1971 13. Added the iplsearch lookup type.
1973 14. Added ident_timeout as a log selector.
1975 15. Added tls_certificate_verified as a log selector.
1977 16. Added a global option tls_require_ciphers (compare the smtp transport
1978 option of the same name). This controls incoming TLS connections.
1980 17. I finally figured out how to make tls_require_ciphers do a similar thing
1981 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1982 before starting the TLS session.
1984 18. Tabs are now shown as \t in -bP output.
1986 19. If the log selector return_path_on_delivery was set, Exim crashed when
1987 bouncing a message because it had too many Received: header lines.
1989 20. If two routers both had headers_remove settings, and the first one included
1990 a superfluous trailing colon, the final name in the first list and the
1991 first name in the second list were incorrectly joined into one item (with a
1992 colon in the middle).
1998 1. Added -C and -D options to the exinext utility, mainly to make it easier
1999 to include in the automated testing, but these could be helpful when
2000 multiple configurations are in use.
2002 2. The exinext utility was not formatting the output nicely when there was
2003 an alternate port involved in the retry record key, nor when there was a
2004 message id as well (for retries that were specific to a specific message
2005 and a specific host). It was also confused by IPv6 addresses, because of
2006 the additional colons they contain. I have fixed the IPv4 problem, and
2007 patched it up to do a reasonable job for IPv6.
2009 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2010 delivery, the log line now contains "pipelined" if PIPELINING was used.
2012 4. An SMTP transport process used to panic and die if the bind() call to set
2013 an explicit outgoing interface failed. This has been changed; it is now
2014 treated in the same way as a connect() failure.
2016 5. A reference to $sender_host_name in the part of a conditional expansion
2017 that was being skipped was still causing a DNS lookup. This no longer
2020 6. The def: expansion condition was not recognizing references to header lines
2021 that used bh_ and bheader_.
2023 7. Added the _cache feature to named lists.
2025 8. The code for checking quota_filecount in the appendfile transport was
2026 allowing one more file than it should have been.
2028 9. For compatibility with Sendmail, the command line option
2036 and sets the incoming protocol and host name (for trusted callers). The
2037 host name and its colon can be omitted when only the protocol is to be set.
2038 Note the Exim already has two private options, -pd and -ps, that refer to
2039 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2040 "s", but I don't think that's a major issue.
2042 10. A number of refactoring changes to the code, none of which should affect
2045 (a) The number of logging options was getting close to filling up the
2046 32-bit word that was used as a bit map. I have split them into two classes:
2047 those that are passed in the argument to log_write(), and those that are
2048 only ever tested independently outside of that function. These are now in
2049 separate 32-bit words, so there is plenty of room for expansion again.
2050 There is no change in the user interface or the logging behaviour.
2052 (b) When building, for example, log lines, the code previously used a
2053 macro that called string_cat() twice, in order to add two strings. This is
2054 not really sufficiently general. Furthermore, there was one instance where
2055 it was actually wrong because one of the argument was used twice, and in
2056 one call a function was used. (As it happened, calling the function twice
2057 did not affect the overall behaviour.) The macro has been replaced by a
2058 function that can join an arbitrary number of extra strings onto a growing
2061 (c) The code for expansion conditions now uses a table and a binary chop
2062 instead of a serial search (which was left over from when there were very
2063 few conditions). Also, it now recognizes conditions like "pam" even when
2064 the relevant support is not compiled in: a suitably worded error message is
2065 given if an attempt is made to use such a condition.
2067 11. Added ${time_interval:xxxxx}.
2069 12. A bug was causing one of the ddress fields not to be passed back correctly
2070 from remote delivery subprocesses. The field in question was not being
2071 subsequently used, so this caused to problems in practice.
2073 13. Added new log selectors queue_time and deliver_time.
2075 14. Might have fixed a bug in maildirsizefile handling that threw up
2076 "unexpected character" debug warnings, and recalculated the data
2077 unnecessarily. In any case, I expanded the warning message to give more
2080 15. Added the message "Restricted characters in address" to the statements in
2081 the default ACL that block characters like @ and % in local parts.
2083 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2084 Three changes have been made:
2086 (a) There was a serious bug; a negative response to MAIL caused the whole
2087 recipient domain to be cached as invalid, thereby blocking all messages
2088 to all local parts at the same domain, from all senders. This bug has
2089 been fixed. The domain is no longer cached after a negative response to
2090 MAIL if the sender used is not empty.
2092 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2095 (c) A new callout option, "use_sender" has been added for people who want
2096 the modified behaviour.
2102 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2105 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2106 indeed breaks things for older releases.
2108 3. Added additional logging to the case where there is a problem reading data
2109 from a filter that is running in a subprocess using a pipe, in order to
2110 try to track down a specific problem.
2112 4. Testing facility fudge: when running in the test harness and attempting
2113 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2114 getting "No route to host". Convert this to a timeout.
2116 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2119 6. Some OS don't have socklen_t but use size_t instead. This affects the
2120 fifth argument of getsockopt() amongst other things. This is now
2121 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2122 can be set for individual OS. I have set it for SunOS5, OSF1, and
2123 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2124 some earlier ones do not.
2126 7. Change 4.30/15 was not doing the test caselessly.
2128 8. The standard form for an IPv6 address literal was being rejected by address
2129 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2130 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2131 this, as well as the form without the "IPv6" on the front (but only when
2132 address literals are enabled, of course).
2134 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2136 10. Exim crashed if a message with an empty sender address specified by -f
2137 encountered a router with an errors_to setting. This could be provoked only
2138 by a command such as
2142 where an empty string was supplied; "<>" did not hit this bug.
2144 11. Installed PCRE release 4.5.
2146 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2147 remained set. It is now erased.
2149 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2150 times from message ids (which are base 36 rather than the normal 62).
2152 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2153 were being counted as actual protocol errors, and logged if the log
2154 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2155 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2156 and DATA following a set of rejected RCPTs do not count as protocol errors.
2157 In other words, Exim assumes they were pipelined, though this may not
2158 actually be the case. Of course, in all cases the client gets an
2159 appropriate error code.
2161 15. If a lookup fails in an ACL condition, a message about the failure may
2162 be available; it is used if testing the ACL cannot continue, because most
2163 such messages specify what the cause of the deferral is. However, some
2164 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2165 that caused an old message to be retained and used if a later statement
2166 caused a defer, replacing the real cause of the deferral.
2168 16. If an IP address had so many PTR records that the DNS lookup buffer
2169 was not large enough to hold them, Exim could crash while trying to process
2170 the truncated data. It now detects and logs this case.
2172 17. Further to 4.21/58, another change has been made: if (and only if) the
2173 first line of a message (the first header line) ends with CRLF, a bare LF
2174 in a subsequent header line has a space inserted after it, so as not to
2175 terminate the header.
2177 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2178 unnecessarily, used atoi() instead of strtol(), and didn't check the
2179 termination when getting file sizes from file names by regex.
2181 19. Completely re-implemented the support for maildirsize files, in the light
2182 of a number of problems with the previous contributed implementation
2183 (4.30/29). In particular:
2185 . If the quota is zero, the maildirsize file is maintained, but no quota is
2188 . If the maildir directory does not exist, it is created before any attempt
2189 to write a maildirsize file.
2191 . The quota value in the file is just a cache; if the quota is changed in
2192 the transport, the new value overrides.
2194 . A regular expression is available for excluding directories from the
2197 20. The autoreply transport checks the characters in options that define the
2198 message's headers; it allows continued headers, but it was checking with
2199 isspace() after an embedded newline instead of explicitly looking for a
2202 21. If all the "regular" hosts to which an address was routed had passed their
2203 expiry times, and had not reached their retry times, the address was
2204 bounced, even if fallback hosts were defined. Now Exim should go on to try
2207 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2208 equivalent code in the SMTP transport. Some hosts send humungous responses
2209 to HELO/EHLO, more than 1024 it seems.
2211 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2212 gives compiler warnings in some environments. I've now done it "properly",
2215 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2216 (because of problems with the built-in one) was declared to return uschar *
2217 instead of char *, causing compiler failure.
2219 25. Fixed a file descriptor leak when processing alias/forward files.
2221 26. Fixed a minor format string issue in dbfn.c.
2223 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2225 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2226 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2229 29. When a sender address is verified, it is cached, to save repeating the test
2230 when there is more than one recipient in a message. However, when the
2231 verification involves a callout, it is possible for different callout
2232 options to be set for different recipients. It is too complicated to keep
2233 track of this in the cache, so now Exim always runs a verification when a
2234 callout is required, relying on the callout cache for the optimization.
2235 The overhead is duplication of the address routing, but this should not be
2238 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2239 to be verified with callout=postmaster, and the main callout worked but the
2240 postmaster check failed, the verification correctly failed. However, if a
2241 subsequent RCPT command asked for sender verification *without* the
2242 postmaster check, incorrect caching caused this verification also to fail,
2245 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2246 it was not caching the DNS options (qualify_single, search_parents) that
2247 were used when the lookup failed. A subsequent lookup with different
2248 options therefore always gave the same answer, though there were cases
2249 where it should not have. (Example: a "domains = !$mx_any" option on a
2250 dnslookup router: the "domains" option is always processed without any
2251 widening, but the router might have qualify_single set.) Now Exim uses the
2252 cached value only when the same options are set.
2254 32. Added John Jetmore's "exipick" utility to the distribution.
2256 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2257 than a timeout (e.g. a certificate is required, and is not provided), an
2258 Exim server now closes the connection immediately. Previously it waited for
2259 the client to close - but if the client is SSL, it seems that they each
2260 wait for each other, leading to a delay before one of them times out.
2262 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2263 maintained 0.8.x compatibility because I don't think many are using it, and
2264 it is clearly obsolete.
2266 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2269 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2270 client certificate was expired. A simple patch fixes this, though I don't
2271 understand the full logic of why the verify callback is called multiple
2274 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2275 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2276 which causes problems with some clients (such as the Certicom SSL Plus
2277 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2278 disables the coutermeasure allowing Eudora to connect."
2280 38. Exim was not checking that a write() to a log file succeeded. This could
2281 lead to Bad Things if a log got too big, in particular if it hit a file
2282 size limit. Exim now panics and dies if it cannot write to a log file, just
2283 as it does if it cannot open a log file.
2285 39. Modified OS/Makefile-Linux so that it now contains
2287 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2289 The two -D definitions ensure that Exim is compiled with large file
2290 support, which makes it possible to handle log files that are bigger than
2293 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2294 instance) a domain was checked against a named list that involved a lookup,
2295 causing $domain_data to be set, then another domain was checked against the
2296 same list, then the first domain was re-checked, the value of $domain_data
2297 after the final check could be wrong. In particular, if the second check
2298 failed, it could be set empty. This bug probably also applied to
2301 41. The strip_trailing_dot option was not being applied to the address given
2302 with the -f command-line option.
2304 42. The code for reading a message's header from the spool was incrementing
2305 $received_count, but never initializing it. This meant that the value was
2306 incorrect (doubled) while delivering a message in the same process in which
2307 it was received. In the most common configuration of Exim, this never
2308 happens - a fresh exec is done - but it can happen when
2309 deliver_drop_privilege is set.
2311 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2312 it now includes up to 150 characters of the unexpected data in the log
2315 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2316 and building data strings. The size of both of these buffers was 10 000
2317 bytes - far larger than anybody would *ever* want, thought I. Needless to
2318 say, somebody hit the limit. I have increased the maximum line length to
2319 20 000 and the maximum data length of concatenated lines to 100 000. I have
2320 also fixed two bugs, because there was no checking on these buffers. Tsk,
2321 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2322 buffer is too small.
2324 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2325 lsearch lookups. Now it does.
2327 46. When parsing a route_list item in a manualroute router, a fixed-length
2328 buffer was used for the list of hosts. I made this 1024 bytes long,
2329 thinking that nobody would ever have a list of hosts that long. Wrong.
2330 Somebody had a whole pile of complicated expansion conditions, and the
2331 string was silently truncated, leading to an expansion error. It turns out
2332 that it is easier to change to an unlimited length (owing to other changes
2333 that have happened since this code was originally written) than to build
2334 structure for giving a limitation error. The length of the item that
2335 expands into the list of hosts is now unlimited.
2337 47. The lsearch lookup could not handle data where the length of text line was
2338 more than 4095 characters. Such lines were truncated, leading to shortened
2339 data being returned. It should now handle lines of any length.
2341 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2342 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2345 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2346 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2347 no longer happens in comment lines. A list of replacements is now placed
2348 at the head of all of the source files, except those whose only change is
2349 to replace PERL_COMMAND in the very first #! line.
2351 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2352 messages on the queue, with a bottom-up merge sort, using code contributed
2353 by Michael Haardt. This should make operations like -bp somewhat faster on
2354 large queues. It won't affect queue runners, except when queue_run_in_order
2357 51. Installed eximstats 1.31 in the distribution.
2359 52. Added support for SRV lookups to the dnslookup router.
2361 53. If an ACL referred to $message_body or $message_body_end, the value was not
2362 reset for any messages that followed in the same SMTP session.
2364 54. The store-handling optimization for building very long strings was not
2365 differentiating between the different store pools. I don't think this
2366 actually made any difference in practice, but I've tidied it.
2368 55. While running the routers to verify a sender address, $sender_address
2369 was still set to the sender address. This is wrong, because when routing to
2370 send a bounce to the sender, it would be empty. Therefore, I have changed
2371 it so that, while verifying a sender address, $sender_address is set to <>.
2372 (There is no change to what happens when verifying a recipient address.)
2374 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2375 target A or AAAA records (if not already returned) without resetting the
2376 qualify_single or search_parents options of the DNS resolver. These are
2377 inappropriate in this case because the targets of MX and SRV records must
2378 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2379 target that, when qualified, matched something in the local domain. These
2380 two options are now turned off when doing these lookups.
2382 57. It seems that at least some releases of Reiserfs (which does not have the
2383 concept of a fixed number of inodes) returns zero and not -1 for the
2384 number of available inodes. This interacted badly with check_spool_inodes,
2385 which assumed that -1 was the "no such thing" setting. What I have done is
2386 to check that the total number of inodes is greater than zero before doing
2387 the test of how many are available.
2389 58. When a "warn" ACL statement has a log_message modifier, the message is
2390 remembered, and not repeated. This is to avoid a lot of repetition when a
2391 message has many recipients that cause the same warning to be written.
2392 Howewer, Exim was preserving the list of already written lines for an
2393 entire SMTP session, which doesn't seem right. The memory is now reset if a
2394 new message is started.
2396 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2397 debugging output unless log_rewrite was also set.
2399 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2400 of (int)(handle) when we know that handle contains (void *)(-1).
2402 61. The Exim daemon panic-logs an error return when it closes the incoming
2403 connection. However "connection reset by peer" seems to be common, and
2404 isn't really an error worthy of noting specially, so that particular error
2407 62. When Exim is trying to find all the local interfaces, it used to panic and
2408 die if the ioctl to get the interface flags failed. However, it seems that
2409 on at least one OS (Solaris 9) it is possible to have an interface that is
2410 included in the list of interfaces, but for which you get a failure error
2411 for this call. This happens when the interface is not "plumbed" into a
2412 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2413 failure of the "get flags" call assumes that the interface is down.
2415 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2416 makes life easier for people who are doing arithmetic on fields extracted
2417 from dates, where you often get leading zeros that should not be
2418 interpreted as octal.
2420 64. Added qualify_domain to the redirect router, to override the global
2423 65. If a pathologically long header line contained very many addresses (the
2424 report of this problem mentioned 10 000) and each of them was rewritten,
2425 Exim could use up a very large amount of memory. (It kept on making new
2426 copies of the header line as it rewrote, and never released the old ones.)
2427 At the expense of a bit more processing, the header rewriting function has
2428 been changed so that it no longer eats memory in this way.
2430 66. The generation of the Received: header has been moved from the time that a
2431 message starts to be received, to the time that it finishes. The timestamp
2432 in the Received: header should now be very close to that of the <= log
2433 line. There are two side-effects of this change:
2435 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2436 logged header lines no longer include the local Received: line, because
2437 it has not yet been created. The same applies to a copy of the message
2438 that is returned to a non-SMTP sender when a message is rejected.
2440 (b) When a filter file is tested using -bf, no additional Received: header
2441 is added to the test message. After some thought, I decided that this
2444 This change does not affect the value of $received_for. It is still set
2445 after address rewriting, but before local_scan() is called.
2447 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2449 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2450 gave an unhelpful panic error message, and a defer error. I have managed to
2451 change this behaviour so that it now rejects any supplied certificate,
2452 which seems right, as the list of acceptable certificates is empty.
2454 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2455 gave an unhelpful defer error. I have not managed to make this reject any
2456 supplied certificates, but the error message it gives is "no certificate
2457 supplied", which is not helpful.
2459 70. exigrep's output now also includes lines that are not associated with any
2460 message, but which match the given pattern. Implemented by a patch from
2461 Martin Sluka, which also tidied up the Perl a bit.
2463 71. Recipient callout verification, like sender verification, was using <> in
2464 the MAIL FROM command. This isn't really the right thing, since the actual
2465 sender may affect whether the remote host accepts the recipient or not. I
2466 have changed it to use the actual sender in the callout; this means that
2467 the cache record is now keyed on a recipient/sender pair, not just the
2468 recipient address. There doesn't seem to be a real danger of callout loops,
2469 since a callout by the remote host to check the sender would use <>.
2470 [SEE ABOVE: changed after hitting problems.]
2472 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2473 temporary errors. However, in the case of such a code being given after
2474 the end of a data transmission (i.e. after ".") Exim was failing to write
2475 a retry record for the message. (Yes, there was some broken host that was
2476 actually sending 8xx at this point.)
2478 73. An unknown lookup type in a host list could cause Exim to panic-die when
2479 the list was checked. (An example that provoked this was putting <; in the
2480 middle of a list instead of at the start.) If this happened during a DATA
2481 ACL check, a -D file could be left lying around. This kind of configuration
2482 error no longer causes Exim to die; instead it causes a defer errror. The
2483 incident is still logged to the main and panic logs.
2485 74. Buglet left over from Exim 3 conversion. The message "too many messages
2486 in one connection" was written to the rejectlog but not the mainlog, except
2487 when address rewriting (yes!) was being logged.
2489 75. Added write_rejectlog option.
2491 76. When a system filter was run not as root (that is, when system_filter_user
2492 was set), the values of the $n variables were not being returned to the
2493 main process; thus, they were not subsequently available in the $sn
2496 77. Added +return_path_on_delivery log selector.
2498 78. A connection timeout was being treated differently from recipients deferred
2499 when testing hosts_max_try with a message that was older than the host's
2500 retry timeout. (The host should not be counted, thus allowing all hosts to
2501 be tried at least once before bouncing.) This may have been the cause of an
2502 occasionally reported bug whereby a message would remain on the queue
2503 longer than the retry timeout, but would be bounced if a delivery was
2504 forced. I say "may" because I never totally pinned down the problem;
2505 setting up timeout/retry tests is difficult. See also the next item.
2507 79. The ultimate address timeout was not being applied to errors that involved
2508 a combination of host plus message (for example, a timeout on a MAIL
2509 command). When an address resolved to a number of possible hosts, and they
2510 were not all tried for each delivery (e.g. because of hosts_max_try), a
2511 message could remain on the queue longer than the retry timeout.
2513 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2516 81. Fixed an obscure SMTP outgoing bug which required at least the following
2517 conditions: (a) there was another message waiting for the same server;
2518 (b) the server returned 5xx to all RCPT commands in the first message so
2519 that the message was not completed; (c) the server dropped the connection
2520 or gave a negative response to the RSET that Exim sends to abort the
2521 transaction. The observed case was a dropped connection after DATA that had
2522 been sent in pipelining mode. That is, the server had advertised PIPELINING
2523 but was not implementing it correctly. The effect of the bug was incorrect
2524 behaviour, such as trying another host, and this could lead to a crash.
2530 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2531 and daemon.c were passed as pointers to ints; they should have been
2532 pointers to socklen_t variables (which are typically unsigned ints).
2534 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2537 3. Fixed a really odd bug that affected only the testing scheme; patching a
2538 certain fixed string in the binary changed the value of another string that
2539 happened to be identical to the end of the original first string.
2541 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2542 name", it returns that address as the IP address. On some operating
2543 systems (e.g. Solaris), it also passes back the IP address string as the
2544 "host name". However, on others (e.g. Linux), it passes back an empty
2545 string. Exim wasn't checking for this, and was changing the host name to an
2546 empty string, assuming it had been canonicized.
2548 5. Although rare, it is permitted to have more than one PTR record for a given
2549 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2550 all the names associated with an address, because they do in Solaris.
2551 However, it seems that they do not in Linux for data that comes from the
2552 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2553 I found this out when I moved to a new Linux workstation and tried to run
2554 the Exim test suite.
2556 To get round this problem I have changed the code so that it now does its
2557 own call to the DNS to look up PTR records when searching for a host name.
2558 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2559 addresses that are only in /etc/hosts are still found.
2561 This behaviour is, however, controlled by an option called host_lookup_
2562 order, which defaults to "bydns:byaddr". If people want to use the other
2563 order, or indeed, just use one or the other means of lookup, they can
2564 specify it in this variable.
2566 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2567 some operating systems, this comes back from gethostbyaddr() as an empty
2568 string, and this is what Exim used to test for. However, it seems that in
2569 other systems, "." is yielded. Exim now tests for this case too.
2571 7. The values of check_spool_space and check_log_space are now held internally
2572 as a number of kilobytes instead of an absolute number of bytes. If a
2573 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2574 kilobyte. This means that much larger values can be stored.
2576 8. Exim monitor: an attempt to get the action menu when not actually pointing
2577 at a message produces an empty menu entitled "No message selected". This
2578 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2579 no entries in it ("Shell widget menu has zero width and/or height"). So I
2580 have added a single, blank menu entry in this case.
2582 9. Added ${quote_local_part.
2584 10. MIME decoding is now applied to the contents of Subject: header lines when
2587 11. Now that a reference to $sender_host_address automatically causes a reverse
2588 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2589 host lookup before query-style lookups in lists that might use this
2590 variable. This has therefore been abolished, and the "net-" prefix is no
2591 longer necessary for query-style lookups.
2593 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2594 have been a typo for LFLAGS, so it has been changed.
2596 13. The install script calls Exim with "-C /dev/null" in order to find the
2597 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2598 to be output. Howeve, since Exim outputs its version number before the
2599 error, it didn't break the script. It just looked ugly. I fixed this by
2600 always allowing "-C /dev/null" if the caller is root.
2602 14. Ignore overlarge ACL variable number when reading spool file - insurance
2603 against a later release with more variables having written the file.
2605 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2606 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2607 this, as well as the form without the "IPv6" on the front.
2609 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2610 OS/Makefile-Darwin file.
2612 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2613 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2615 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2616 variable. However, some OS use ints for inodes. I've added cast to long int
2617 to get rid of the compiler warning.
2619 19. I had forgotten to lock out "/../" in configuration file names when
2620 ALT_CONFIG_PREFIX was set.
2622 20. Routers used for verification do not need to specify transports. However,
2623 if such a router generated a host list, and callout was configured, Exim
2624 crashed, because it could not find a port number from the (non-existent)
2625 transport. It now assumes port 25 in this circumstance.
2627 21. Added the -t option to exigrep.
2629 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2630 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2631 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2632 them is set, LOOKUP_LSEARCH is forced.
2634 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2636 24. Added sender and host information to the "rejected by local_scan()" log
2637 line; previously there was no indication of these.
2639 25. Added .include_if_exists.
2641 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2642 Linux. It turns out that not all file systems support this. Apparently some
2643 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2644 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2645 not supported on the file descriptor, is now ignored when Exim is trying to
2646 sync a directory. This applies only to Linux.
2648 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2650 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2651 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2654 29. Added support for maildirsize files from supplied patch (modified a bit).
2656 30. The use of :fail: followed by an empty string could lead Exim to respond to
2657 sender verification failures with (e.g.):
2659 550 Verification failed for <xxx>
2660 550 Sender verify failed
2662 where the first response line was missing the '-' that indicates it is not
2663 the final line of the response.
2665 31. The loop for finding the name of the user that called Exim had a hardwired
2666 limit of 10; it now uses the value of finduser_retries, which is used for
2667 all other user lookups.
2669 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2672 33. Exim was neglecting to zero errno before one call of strtol() when
2673 expanding a string and expecting an integer value. On some systems this
2674 resulted in spurious "integer overflow" errors. Also, it was casting the
2675 result into an int without checking.
2677 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2678 did not work. The code looks as if it has *never* worked, though it appears
2679 to have been documented since at least releast 1.62. I have made it work.
2681 35. The "timeout_DNS" error in retry rules, also documented since at least
2682 1.62, also never worked. As it isn't clear exactly what this means, and
2683 clearly it isn't a major issue, I have abolished the feature by treating it
2684 as "timeout", and writing a warning to the main and panic logs.
2686 36. The display of retry rules for -brt wasn't always showing the error code
2689 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2690 timeout_connect_A, timeout_connect_MX.
2692 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2693 to the empty sender.
2695 39. The daemon was not analysing the content of -oX till after it had closed
2696 stderr and disconnected from the controlling terminal. This meant that any
2697 syntax errors were only noted on the panic log, and the return code from
2698 the command was 0. By re-arranging the code a little, I've made the
2699 decoding happen first, so such errors now appear on stderr, and the return
2700 code is 1. However, the actual setting up of the sockets still happens in
2701 the disconnected process, so errors there are still only recorded on the
2704 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2705 connections (as happens on some IP stacks) was logged at start up time as
2706 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2707 it from "IPv6 and IPv4", which means that two separate sockets are being
2710 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2711 says whether AF_INET or AF_INET6 was passed as an argument.
2713 42. Exiwhat output was messed up when time zones were included in log
2716 43. Exiwhat now gives more information about the daemon's listening ports,
2717 and whether -tls-on-connect was used.
2719 44. The "port" option of the smtp transport is now expanded.
2721 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2722 silently ignored. Now an error message is written to the main and panic
2725 46. There's a new ACL modifier called "logwrite" which writes to a log file
2726 as soon as it is encountered.
2728 47. Added $local_user_uid and $local_user_gid at routing time.
2730 48. Exim crashed when trying to verify a sender address that was being
2733 49. Exim was recognizing only a space character after ".include". It now also
2734 recognizes a tab character.
2736 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2737 extracting the relevant information from the specification. The man page no
2738 longer contains scrambled data for the -d option, and I've added a section
2739 at the front about calling Exim under different names.
2741 51. Added "extra_headers" argument to the "mail" command in filter files.
2743 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2746 53. Installed eximstats 1.29.
2748 54. Added transport_filter_timeout as a generic transport option.
2750 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2751 Cc: header lines. This was required by RFC 822, but it not required by RFC
2754 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2755 incoming messages that did not have them. Now it does so only if the
2756 message originates locally, that is, if there is no associated remote host
2757 address. When Resent- header lines are present, this applies to the Resent-
2758 lines rather than the non-Resent- lines.
2760 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2761 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2763 58. Messages for configuration errors now include the name of the main
2764 configuration file - useful now that there may be more than one file in a
2765 list (.included file names were always shown).
2767 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2768 for those rare installations that do not start the daemon as root or run it
2769 setuid root. I've cut out the call to initgroups() if the daemon is not
2772 60. The Exim user and group can now be bound into the binary as text strings
2773 that are looked up at the start of Exim's processing.
2775 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2777 62. Added $mailstore_basename variable.
2779 63. Installed patch to sieve.c from Michael Haardt.
2781 64. When Exim failed to open the panic log after failing to open the main log,
2782 the original message it was trying to log was written to stderr and debug
2783 output, but if they were not available (the usual case in production), it
2784 was lost. Now it is written to syslog before the two lines that record the
2785 failures to open the logs.
2787 65. Users' Exim filters run in subprocesses under the user's uid. It is
2788 possible for a "deliver" command or an alias in a "personal" command to
2789 provoke an address rewrite. If logging of address rewriting is configured,
2790 this fails because the process is not running as root or exim. There may be
2791 a better way of dealing with this, but for the moment (because 4.30 needs
2792 to be released), I have disabled address rewrite logging when running a
2793 filter in a non-root, non-exim process.
2799 1. The buildconfig auxiliary program wasn't quoting the value set for
2800 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2801 not defined. This bug was masked in 4.22 by the effect that was fixed in
2804 2. Some messages that were rejected after a message id was allocated were
2805 shown as "incomplete" by exigrep. It no longer does this for messages that
2806 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2808 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2809 have allow_domain_literals set, the ID did not get logged in the <= line.
2810 Domain literals are now always recognized in Message-ID: header lines.
2812 4. The first argument for a ${extract expansion item is the key name or field
2813 number. Leading and trailing spaces in this item were not being ignored,
2814 causing some misleading effects.
2816 5. When deliver_drop_privilege was set, single queue runner processes started
2817 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2818 same command in the process it spins off) were not dropping privilege.
2820 6. When the daemon running as "exim" started a queue runner, it always
2821 re-executed Exim in the spun-off process. This is a waste of effort when
2822 deliver_drop_privilege is set. The new process now just calls the
2823 queue-runner function directly.
2829 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2832 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2833 ignored. Though the use of -oP was forcing the writing of a pid file, it
2834 was always written to the default place.
2836 3. If the message "no IP address found for host xxxx" is generated during
2837 incoming verification, it is now followed by identification of the incoming
2838 connection (so you can more easily find what provoked it).
2840 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2842 5. Added some features to "harden" Exim a bit more against certain attacks:
2844 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2845 be put in Local/Makefile. This is like the never_users runtime option,
2846 but it cannot be overridden. The default setting is "root".
2848 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2849 prefix string with which any file named in a -C command line option
2852 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2853 is retained for -C and -D only if the caller of Exim is root. Without
2854 it, the exim user may also use -C and -D and retain privilege.
2856 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2857 command line option is disabled.
2859 6. Macro names set by the -D option must start with an upper case letter, just
2860 like macro names defined in the configuration file.
2862 7. Added "dereference=" facility to LDAP.
2864 8. Two instances of the typo "uknown" in the source files are fixed.
2866 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2867 the Configure-Makefile script screwed up while processing it.
2869 10. Incorporated PCRE 4.4.
2871 11. The SMTP synchronization check was not operating right at the start of an
2872 SMTP session. For example, it could not catch a HELO sent before the client
2873 waited for the greeting. There is now a check for outstanding input at the
2874 point when the greeting is written. Because of the duplex, asynchronous
2875 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2876 way, but not yet received, when the check is performed.
2878 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2879 on TCP/IP sockets, because this apparently causes some broken clients to
2882 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2883 unchanged) from the Cygwin maintainer.
2885 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2886 was supported instead of testing for mbx. Effectively a typo.
2888 15. The spa authenticator server code was not checking that the input it
2889 received was valid base64.
2891 16. The debug output line for the "set" modifier in ACLs was not showing the
2892 name of the variable that was being set.
2894 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2896 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2897 lookup on its own. Something else was needed to trigger the lookup. For
2898 example, a match in host_lookup or the need for a host name in a host list.
2899 Now, if $sender_host_name is referenced and the host name has not yet been
2900 looked up, a lookup is performed. If the lookup fails, the variable remains
2901 empty, and $host_lookup_failed is set to "1".
2903 19. Added "eqi" as a case-independent comparison operator.
2905 20. The saslauthd authentication condition could segfault if neither service
2906 nor realm was specified.
2908 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2909 error message that was logged was misleading, and incoming SMTP
2910 connections were dropped. The message is now more accurate, and temporary
2911 errors are given to SMTP connections.
2913 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2914 (except RSET) with a 421 error, until QUIT is received. However, it was
2915 failing to send a response to QUIT.
2917 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2918 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2919 ACL is now run after the helo_try_verify_hosts code.
2921 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2922 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2923 case-independent, so other case variants are also recognized). Apparently
2924 some systems use these upper case variants.
2926 25. If more than two messages were waiting for the same host, and a transport
2927 filter was specified for the transport, Exim sent two messages over the
2928 same TCP/IP connection, and then failed with "socket operation on non-
2929 socket" when it tried to send the third.
2931 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2933 27. The extern definition of crypt16() in expand.c was not being excluded when
2934 the OS had its own crypt16() function.
2936 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2937 as a preferred synonym for return_size_limit, both as an option and as an
2940 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2942 30. Changed the default configuration ACL to relax the local part checking rule
2943 for addresses that are not in any local domains. For these addresses,
2944 slashes and pipe symbols are allowed within local parts, but the sequence
2945 /../ is explicitly forbidden.
2947 31. SPA server authentication was not clearing the challenge buffer before
2950 32. log_message in a "warn" ACL statement was writing to the reject log as
2951 well as to the main log, which contradicts the documentation and doesn't
2952 seem right (because no rejection is happening). So I have stopped it.
2954 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2955 However, I am unable to do any testing of this.
2957 34. Fixed an infelicity in the appendfile transport. When checking directories
2958 for a mailbox, to see if any needed to be created, it was accidentally
2959 using path names with one or more superfluous leading slashes; tracing
2960 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2962 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2963 contents are added to the log line that is written for every discarded
2964 recipient. (Previously a log_message setting was ignored.)
2966 36. The ${quote: operator now quotes the string if it is empty.
2968 37. The install script runs exim in order to find its version number. If for
2969 some reason other than non-existence or emptiness, which it checks, it
2970 could not run './exim', it was installing it with an empty version number,
2971 i.e. as "exim-". This error state is now caught, and the installation is
2974 38. An argument was missing from the function that creates an error message
2975 when Exim fails to connect to the socket for saslauthd authentication.
2976 This could cause Exim to crash, or give a corrupted message.
2978 39. Added isip, isip4, and isip6 to ${if conditions.
2980 40. The ACL variables $acl_xx are now saved with the message, and can be
2981 accessed later in routers, transports, and filters.
2983 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2984 strings in the file are not string-expanded.
2986 42. If a MAIL command specified a SIZE value that was too large to fit into an
2987 int variable, the check against message_size_limit failed. Such values are
2988 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2989 day this will have to be increased, but I don't think I want to be around
2990 when emails are that large.
2997 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2998 iconv() is not standard in FreeBSD.
3000 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3001 IPv6 enabled. The observed symptom was a segmentation fault on return from
3002 the function os_common_find_running_interfaces() in src/os.c.
3004 3. In the check_special_case() function in daemon.c I had used "errno" as an
3005 argument name, which causes warnings on some systems. This was basically a
3006 typo, since it was named "eno" in the comments!
3008 4. The code that waits for the clock to tick (at a resolution of some fraction
3009 of a second) so as to ensure message-id uniqueness was always waiting for
3010 at least one whole tick, when it could have waited for less. [This is
3011 almost certainly not relevant at current processor speeds, where it is
3012 unlikely to ever wait at all. But we try to future-proof.]
3014 5. The function that sleeps for a time interval that includes fractions of a
3015 second contained a race. It did not block SIGALRM between setting the
3016 timer, and suspending (a couple of lines later). If the interval was short
3017 and the sigsuspend() was delayed until after it had expired, the suspension
3018 never ended. On busy systems this could lead to processes getting stuck for
3021 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3022 process, before it forks any delivery processes. The open lookup caching
3023 mechanism meant that the open file or database connection was passed into
3024 the delivery process. The problem was that delivery processes always tidy
3025 up cached lookup data. This could cause a problem for the next delivery
3026 process started by the queue runner, because the external queue runner
3027 process does not know about the closure. So the next delivery process
3028 still has data in the lookup cache. In the case of a file lookup, there was
3029 no problem because closing a file descriptor in a subprocess doesn't affect
3030 the parent. However, if the lookup was caching a connection to a database,
3031 the connection was closed, and the second delivery process was likely to
3032 see errors such as "PGSQL: query failed: server closed the connection
3033 unexpectedly". The problem has been fixed by closing all cached lookups
3034 in a queue runner before running a delivery process.
3036 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3037 seem to have the "const" qualifier which it has on other OS. I've
3040 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3041 *of the same type* (client or server) with the same public name that an
3042 error should be diagnosed.
3044 9. When Exim looked up a host name for an IP address, but failed to find the
3045 original IP address when looking up the host name (a safety check), it
3046 output the message "<ip address> does not match any IP for NULL", which was
3047 confusing, to say the least. The bug was that the host name should have
3048 appeared instead of "NULL".
3050 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3051 user that is built into the binary, and the -C or -D options is used, root
3052 privilege is dropped before the configuration file is read. In addition,
3053 logging is switched to stderr instead of the normal log files. If the
3054 configuration then re-defines the Exim user, the unprivileged environment
3055 is probably not what is expected, so Exim logs a panic warning message (but
3058 However, if deliver_drop_privilege is set, the unprivileged state may well
3059 be exactly what is intended, so the warning has been cut out in that case,
3060 and Exim is allowed to try to write to its normal log files.
3066 1. smtp_return_error_details was not giving details for temporary sender
3067 or receiver verification errors.
3069 2. Diagnose a configuration error if two authenticators have the same public
3072 3. Exim used not to create the message log file for a message until the first
3073 delivery attempt. This could be confusing when incoming messages were held
3074 for policy or load reasons. The message log file is now created at the time
3075 the message is received, and an initial "Received" line is written to it.
3077 4. The automatically generated man page for command line options had a minor
3078 bug that caused no ill effects; however, a more serious problem was that
3079 the procedure for building the man page automatically didn't always
3080 operate. Consequently, release 4.20 contains an out-of-date version. This
3081 shouldn't happen again.
3083 5. When building Exim with embedded Perl support, the script that builds the
3084 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3085 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3087 6. The freeze_tell option was not being used for messages that were frozen on
3088 arrival, either by an ACL or by local_scan().
3090 7. Added the smtp_incomplete_transaction log selector.
3092 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3093 was accepting AUTH without a new EHLO.
3095 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3096 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3099 10. Logging of TCP/IP connections (when configured) now happens in the main
3100 daemon process instead of the child process, so that the TCP/IP connection
3101 count is more accurate (but it can never be perfect).
3103 11. The use of "drop" in a nested ACL was not being handled correctly in the
3104 outer ACL. Now, if condition failure induced by the nested "drop" causes
3105 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3106 or "require"), the connection is dropped.
3108 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3109 that yield "discard" can now be used with an "accept" or a "discard" verb,
3110 but an error is generated for any others (because I can't see a useful way
3111 to define what should happen).
3113 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3114 are now processed in the same way as lines in the Exim configuration file.
3115 In particular, continuation lines are supported.
3117 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3119 15. Added -ti meaning -t -i.
3121 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3122 domains, and warn by logging if others are found.
3124 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3125 returned from GIFCONF when Exim is trying to find the list of interfaces on
3126 a host. The code in os.c has been modified to copy each ifreq to an aligned
3127 structure in all cases.
3129 Also, in some cases, the returned ifreq's were being copied to a 'struct
3130 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3131 means the last couple of bytes of an IPv6 address could be chopped if the
3132 ifreq contained only a normal sockaddr (14 bytes storage).
3134 18. Named domain lists were not supported in the hosts_treat_as_local option.
3135 An entry such as +xxxx was not recognized, and was treated as a literal
3138 19. Ensure that header lines added by a DATA ACL are included in the reject log
3139 if the ACL subsequently rejects the message.
3141 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3142 MD5 (which is deprecated).
3144 21. When testing a filter file using -bf, Exim was writing a message when it
3145 took the sender from a "From " line in the message, but it was not doing so
3146 when it took $return_path from a Return-Path: header line. It now does.
3148 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3149 with a valid header line field name (a series of printing characters
3150 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3152 23. Changed "disc" in the source to "disk" to conform to the documentation and
3153 the book and for uniformity.
3155 24. Ignore Sendmail's -Ooption=value command line item.
3157 25. When execve() failed while trying to run a command in a pipe transport,
3158 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3159 could be confused with a return value of 69 from the command itself. This
3160 has been changed to 127, the value the shell returns if it is asked to run
3161 a non-existent command. The wording for the related log line suggests a
3162 non-existent command as the problem.
3164 26. If received_header_text expands to an empty string, do not add a Received:
3165 header line to the message. (Well, it adds a token one on the spool, but
3166 marks it "old" so that it doesn't get used or transmitted.)
3168 27. Installed eximstats 1.28 (addition of -nt option).
3170 28. There was no check for failure on the call to getsockname() in the daemon
3171 code. This can fail if there is a shortage of resources on the system, with
3172 ENOMEM, for example. A temporary error is now given on failure.
3174 29. Contrary to the C standard, it seems that in some environments, the
3175 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3176 program. Exim now does this explicitly; it affects the formatting of
3177 timestamps using strftime().
3179 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3180 complaints. I've now initialized all the variables, to avoid this.
3182 32. Header lines added by a system filter were not being "seen" during
3183 transport-time rewrites.
3185 33. The info_callback() function passed to OpenSSL is set up with type void
3186 (*)(SSL *, int, int), as described somewhere. However, when calling the
3187 function (actually a macro) that sets it up, the type void(*)() is
3188 expected. I've put in a cast to prevent warnings from picky compilers.
3190 34. If a DNS black list lookup found a CNAME record, but there were no A
3191 records associated with the domain it pointed at, Exim crashed.
3193 35. If a DNS black list lookup returned more than one A record, Exim ignored
3194 all but the first. It now scans all returned addresses if a particular IP
3195 value is being sought. In this situation, the contents of the
3196 $dnslist_value variable are a list of all the addresses, separated by a
3199 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3200 to accept a host name and all its aliases if the forward lookup for any of
3201 them yielded the IP address of the incoming connection. Now it accepts only
3202 those names whose forward lookup yields the correct IP address. Any other
3203 names are discarded. This closes a loophole whereby a rogue DNS
3204 administrator could create reverse DNS records to break through a
3205 wildcarded host restriction in an ACL.
3207 37. If a user filter or a system filter that ran in a subprocess used any of
3208 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3209 the wrong values were passed to the pipe command ($thisaddress had the
3210 value of $0, $0 had the value of $1, etc). This bug was introduced by
3211 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3213 38. Improved the line breaking for long SMTP error messages from ACLs.
3214 Previously, if there was no break point between 40 and 75 characters, Exim
3215 left the rest of the message alone. Two changes have been made: (a) I've
3216 reduced the minimum length to 35 characters; (b) if it can't find a break
3217 point between 35 and 75 characters, it looks ahead and uses the first one
3218 that it finds. This may give the occasional overlong line, but at least the
3219 remaining text gets split now.
3221 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3222 file descriptors might be low, and that setting 1000 would always raise it.
3223 It turns out that in some environments, the limit is already over 1000 and
3224 that lowering it causes trouble. So now Exim takes care not to decrease it.
3226 40. When delivering a message, the value of $return_path is set to $sender_
3227 address at the start of routing (routers may change the value). By an
3228 oversight, this default was not being set up when an address was tested by
3229 -bt or -bv, which affected the outcome if any router or filter referred to
3232 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3233 log only when "message" or "log_message" are set. However, if one of the
3234 conditions was an address verification, or a call to a nested ACL, the
3235 messages generated by the underlying test were being passed through. This
3236 no longer happens. The underlying message is available in $acl_verify_
3237 message for both "message" and "log_message" expansions, so it can be
3238 passed through if needed.
3240 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3241 new expansion $bh_ to give the encoded byte string without charset
3242 translation. Translation happens only if iconv() is available; HAVE_ICONV
3243 indicates this at build time. HEADERS_CHARSET gives the charset to
3244 translate to; headers_charset can change it in the configuration, and
3245 "headers charset" can change it in an individual filter file.
3247 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3248 that creates RFC 2047 encoded "words" labels them as that charset instead
3249 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3250 expansion operator; (ii) when Exim creates a From: line for a local
3251 message; (iii) when a header line is rewritten to include a "phrase" part.
3253 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3254 buggy, causing it to skip the first lines of messages whose message ID
3255 ended in 'D'. This would not have bitten before Exim release 4.14, because
3256 message IDs were unlikely to end in 'D' before then. The effect was to have
3257 incorrect size information for certain domains.
3259 45. #include "config.h" was missing at the start of the crypt16.c module. This
3260 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3263 46. If there was a timeout during a "random" callout check, Exim treated it as
3264 a failure of the random address, and carried on sending RSET and the real
3265 address. If the delay was just some slowness somewhere, the response to the
3266 original RCPT would be taken as a response to RSET and so on, causing
3267 mayhem of various kinds.
3269 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3270 when I implemented it. It didn't allow for the fact that some option values
3271 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3272 the right test for positive values.
3274 48. Domain names in DNS records are case-independent. Exim always looks them up
3275 in lower case. Some resolvers return domain names in exactly the case they
3276 appear in the zone file, that is, they may contain uppercase letters. Not
3277 all resolvers do this - some return always lower case. Exim was treating a
3278 change of case by a resolver as a change of domain, similar to a widening
3279 of a domain abbreviation. This triggered its re-routing code and so it was
3280 trying to route what was effectively the same domain again. This normally
3281 caused routing to fail (because the router wouldn't handle the domain
3282 twice). Now Exim checks for this case specially, and just changes the
3283 casing of the domain that it ultimately uses when it transmits the message
3286 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3289 50. If a filter generated a file delivery with a non-absolute name (possible if
3290 no home directory exists for the router), the forbid_file option was not
3293 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3294 the existing equality matching.
3296 52. Exim was using ints instead of ino_t variables in some places where it was
3297 dealing with inode numbers.
3299 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3300 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3301 called TMPDIR, and if it finds it is different, it changes its value.
3303 54. The smtp_printf() function is now made available to local_scan() so
3304 additional output lines can be written before returning. There is also an
3305 smtp_fflush() function to enable the detection of a dropped connection.
3306 The variables smtp_input and smtp_batched_input are exported to
3309 55. Changed the default runtime configuration: the message "Unknown user"
3310 has been removed from the ACL, and instead placed on the localuser router,
3311 using the cannot_route_message feature. This means that any verification
3312 failures that generate their own messages won't get overridden. Similarly,
3313 the "Unrouteable address" message that was in the ACL for unverifiable
3314 relay addresses has also been removed.
3316 56. Added hosts_avoid_esmtp to the smtp transport.
3318 57. The exicyclog script was not checking for the esoteric option
3319 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3320 will work only if exicyclog is run under the appropriate euid.
3322 58. Following a discussion on the list, the rules by which Exim recognises line
3323 endings on incoming messages have been changed. The -dropcr and drop_cr
3324 options are now no-ops, retained only for backwards compatibility. The
3325 following line terminators are recognized: LF CRLF CR. However, special
3326 processing applies to CR:
3328 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3329 nor a local message in the state where . is a terminator.
3331 (ii) If a bare CR is encountered in a header line, an extra space is added
3332 after the line terminator so as not to end the header. The reasoning
3333 behind this is that bare CRs in header lines are most likely either
3334 to be mistakes, or people trying to play silly games.
3336 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3337 was being incorrectly given as 18 bytes larger than it should have been.
3338 This is a VOB (very old bug).
3340 60. This may never have affected anything current, but just in case it has:
3341 When the local host is found other than at the start of a list of hosts,
3342 the local host, those with the same MX, and any that follow, are discarded.
3343 When the list in question was part of a longer list of hosts, the following
3344 hosts (not currently being processed) were also being discarded. This no
3345 longer happens. I'm not sure if this situation could ever has previously
3348 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3351 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3352 header. This is something that is recommended in a new Internet Draft, and
3353 is something that is documented as being done by Sendmail. There are two
3354 possible values. For messages generated by the autoreply transport, Exim
3357 Auto-Submitted: auto-replied
3359 whereas for all other generated messages (e.g. bounces) it adds
3361 Auto-Submitted: auto-generated
3363 63. The "personal" condition in filters now includes a test for the
3364 Auto-Submitted: header. If it contains the string "auto-" the message it
3365 not considered personal.
3367 64. Added rcpt_include_affixes as a generic transport option.
3369 65. Added queue_only_override (default true).
3371 66. Added the syslog_duplication option.
3373 67. If what should have been the first header line of a message consisted of
3374 a space followed by a colon, Exim was mis-interpreting it as a header line.
3375 It isn't of course - it is syntactically invalid and should therefore be
3376 treated as the start of the message body. The misbehaviour could have
3377 caused a number of strange effects, including loss of data in subsequent
3378 header lines, and spool format errors.
3380 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3381 client host had authenticated. This control can now be exercised by an ACL
3382 for more flexibility.
3384 69. By default, callouts do not happen when testing with -bh. There is now a
3385 variant, -bhc, which does actually run the callout code, including
3386 consulting and updating the callout cache.
3388 70. Added support for saslauthd authentication, courtesy of Alexander
3391 71. If statvfs() failed on the spool or log directories while checking their
3392 size for availability, Exim confusingly gave the error "space shortage".
3393 Furthermore, in debugging mode it crashed with a floating point exception.
3394 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3395 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3396 serious problem, Exim now writes to the main and panic logs when this
3397 happens, with details of the failure. It then refuses to accept the
3398 incoming message, giving the message "spool directory problem" or "log
3399 directory problem" with a 421 code for SMTP messages.
3401 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3402 0, 1, and 2 exist, because some OS complain for execs without them (see
3403 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3404 descriptors. However, the code omitted to check that the open succeeded,
3405 causing mysterious errors if for some reason the permissions on /dev/null
3406 got screwed. Now Exim writes a message to the main and panic logs, and
3407 bombs out if it can't open /dev/null.
3409 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3410 interact so that it is all more flexible. It is supposed to remain
3411 backwards compatible. Also added extra_local_interfaces.
3413 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3414 to bomb out with an assertion failure - to the client this appears as a
3415 connection drop. This problem occurs in the part of the code that was taken
3416 from the Samba project. Fortunately, the assertion is in a very simple
3417 function, so I have fixed this by reproducing the function inline in the
3418 one place where it is called, and arranging for authentication to fail
3419 instead of killing the process with assert().
3421 75. The SPA client code was not working when the server requested OEM rather
3422 than Unicode encoding.
3424 76. Added code to make require_files with a specific uid setting more usable in
3425 the case where statting the file as root fails - usually a non-root-mounted
3426 NFS file system. When this happens and the failure is EACCES, Exim now
3427 forks a subprocess and does the per-uid checking as the relevant uid.
3429 77. Added process_log_path.
3431 78. If log_file_path was not explicitly set, a setting of check_log_space or
3432 check_log_inodes was ignored.
3434 79. If a space check for the spool or log partitions fails, the incident is now
3435 logged. Of course, in the latter case the data may get lost...
3437 80. Added the %p formatting code to string_format() so that it can be used to
3438 print addresses in debug_print(). Adjusted all the address printing in the
3439 debugging in store.c to use %p rather than %d.
3441 81. There was a concern that a line of code in smtp_in.c could overflow a
3442 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3443 initially expressed, the concern was not well-founded, because trailing
3444 spaces are removed early. However, if the trailing spaces were followed by
3445 a NULL, they did not get removed, so the overflow was possible. Two fixes
3448 (a) I re-wrote the offending code in a cleaner fashion.
3449 (b) If an incoming SMTP command contains a NULL character, it is rejected
3452 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3453 runs initgroups(), so that if the Exim user is in any additional groups,
3454 they will be used during message reception.
3460 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob