1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.255 2005/11/14 10:46:27 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
18 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
37 PH/03 For some reason, the default setting of PATH when running a command from
38 a pipe transport was just "/usr/bin". I have changed it to
41 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
42 anything to be listed in the output from -bV.
44 PH/05 When a filter generated an autoreply, the entire To: header line was
45 quoted in the delivery log line, like this:
47 => >A.N.Other <ano@some.domain> <original@ddress> ...
49 This has been changed so that it extracts the operative address. There
50 may be more than one such address. If so, they are comma-separated, like
53 => >ano@some.domain,ona@other.domain <original@ddress> ...
55 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
56 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
57 not being looked up in the rDNS to get a host name, Exim was showing the
58 IP address twice in Received: lines, even though the IP addresses were
59 identical. For example:
61 Received: from [1.2.3.4] (helo=[1.2.3.4])
63 However, if the real host name was known, it was omitting the HELO data
64 if it matched the actual IP address. This has been tidied up so that it
65 doesn't show the same IP address twice.
67 PH/07 When both +timestamp and +memory debugging was on, the value given by
68 $tod_xxx expansions could be wrong, because the tod_stamp() function was
69 called by the debug printing, thereby overwriting the timestamp buffer.
70 Debugging no longer uses the tod_stamp() function when +timestamp is set.
76 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
77 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
80 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
81 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
83 PH/03 Typo: missing ".o" in src/pcre/Makefile.
85 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
86 header line, restrict the check to what is listed in RFCs 2369 and 2929.
87 Also, for "Auto-Submitted", treat anything other than "no" as
88 non-personal, in accordance with RFC 3834. (Previously it treated
89 anything starting "auto-" as non-personal.)
91 TF/01 The control=submission/name=... option had a problem with syntax
92 errors if the name included a slash character. The /name= option
93 now slurps the rest of the string, so it can include any characters
94 but it must come last in the list of options (after /sender_retain
97 PH/05 Some modifications to the interface to the fake nameserver for the new
104 TK/01 Added the "success_on_redirect" address verification option. See
105 NewStuff for rationale and an example.
107 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
109 PH/02 Patch to exigrep to allow it to work on syslog lines.
111 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
112 fread() to skip over the body file's header line, because in Cygwin the
113 header line is locked and is inaccessible.
115 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
116 co-exist for some time) to make it clear that it is the Exim ID that is
117 referenced, not the Message-ID: header line.
119 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
120 string_format() function, because snprintf() does not exist on all
123 PH/06 The use of forbid_filter_existstest now also locks out the use of the
124 ${stat: expansion item.
126 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
127 protocol synchronization error", to keep the pedants happy.
129 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
130 well as for IRIX systems, when gcc is being used. See the host.c source
133 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
135 PH/10 Named domain lists were not working if used in a queue_smtp_domains
138 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
139 transport and to the smtp transport in LMTP mode.
141 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
143 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
144 run a filter in a subprocess. This could lead to confusion in subsequent
145 lookups in the parent process. There should also be a search_tidyup() at
146 the end of the subprocess.
148 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
149 only if the host matched helo_try_verify_hosts, which caused the
150 verification to occur when the EHLO/HELO command was issued. The ACL just
151 tested the remembered result. Now, if a previous verification attempt has
152 not happened, "verify = helo" does it there and then.
154 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
156 TK/03 Fix log output including CR from clamd.
158 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
159 exist provoked a memory error which could cause a segfault.
161 PH/15 Installed PCRE 6.2
163 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
165 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
166 of the problem. Specifically, suggested +O2 rather than +O1 for the
169 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
171 PH/20 If a delivery was routed to a non-standard port by means of an SRV
172 record, the port was not correctly logged when the outgoing_port log
173 selector was set (it logged the transort's default port).
175 PH/21 Added support for host-specific ports to manualroute, queryprogram,
176 fallback_hosts, and "hosts" in the smtp transport.
178 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
179 host errors such as "Connection refused".
181 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
182 authentication with radiusclient 0.4.9:
184 - Error returned from rc_read_config was caught wrongly
185 - Username/password not passed on to radius server due to wrong length.
187 The presumption is that some radiusclient API changes for 4.51/PH/17
188 were not taken care of correctly. The code is still untested by me (my
189 Linux distribution still has 0.3.2 of radiusclient), but it was
190 contributed by a Radius user.
192 PH/24 When doing a callout, the value of $domain wasn't set correctly when
193 expanding the "port" option of the smtp transport.
195 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
196 while reading a MIME header. Thanks to Tom Hughes for a patch.
198 PH/24 Include config.h inside local_scan.h so that configuration settings are
201 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
202 that in an ACL for RCPT (for example), you can examine exactly what was
205 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
206 commands, but it was not correctly comparing the address with the actual
207 client host address. Thus, it would show the EHLO address in Received:
208 header lines when this was not necessary.
210 PH/27 Added the % operator to ${eval:}.
212 PH/28 Exim tries to create and chdir to its spool directory when it starts;
213 it should be ignoring failures (because with -C, for example, it has lost
214 privilege). It wasn't ignoring creation failures other than "already
217 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
220 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
221 because an input line was too long, either on its own, or by virtue of
222 too many continuations, the temporary file was not being removed, and the
223 return code was incorrect.
225 PH/31 Missing "BOOL" in function definition in filtertest.c.
227 PH/32 Applied Sieve patches from the maintainer.
229 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
231 PH/33 Added "verify = not_blind".
233 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
234 Local/Makefile (with some defaults set). These are used in built scripts
235 such as exicyclog, but they have never been used in the exim_install
236 script (though there are many overriding facilities there). I have
237 arranged that the exim_install script now takes note of these two
240 PH/35 Installed configuration files for Dragonfly.
242 PH/36 When a locally submitted message by a trusted user did not contain a
243 From: header, and the sender address was obtained from -f or from an SMTP
244 MAIL command, and the trusted user did not use -F to supply a sender
245 name, $originator_name was incorrectly used when constructing a From:
246 header. Furthermore, $originator_name was used for submission mode
247 messages from external hosts without From: headers in a similar way,
248 which is clearly wrong.
250 PH/37 Added control=suppress_local_fixups.
252 PH/38 When log_selector = +received_sender was set, and the addition of the
253 sender made the log line's construction buffer exactly full, or one byte
254 less than full, an overflow happened when the terminating "\n" was
257 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
258 when the result of a list match is failure because a DNS lookup failed.
260 PH/40 RM_COMMAND is now used in the building process.
262 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
263 the "build-* directories that it finds.
265 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
266 address in a domain literal was a prefix of an interface address.
268 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
269 when verifying a sender address, unless rewrite_headers is false.
271 PH/44 Wrote a long comment about why errors_to addresses are verified as
272 recipients, not senders.
274 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
275 the ratelimit ACL was added.
277 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
279 PH/46 Added extra information about PostgreSQL errors to the error string.
281 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
282 avoiding the need to install special zones in a real server. This is
283 backwards compatible; if it can't find the fake resolver, it drops back.
284 Thus, both old and new test suites can be run.
286 TF/02 Added util/ratelimit.pl
288 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
289 clock is set back in time.
291 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
292 Candler <B.Candler@pobox.com>.
294 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
295 OK for addresses that are the result of redirections.
297 PH/48 A number of further additions for the benefit of the new test suite,
298 including a fake gethostbyname() that interfaces to the fake DNS resolver
301 TF/06 The fix for widen_domains has also been applied to qualify_single and
302 search_parents which are the other dnslookup options that can cause
305 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
308 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
310 TF/07 Exim produced the error message "an SRV record indicated no SMTP
311 service" if it encountered an MX record with an empty target hostname.
312 The message is now "an MX or SRV record indicated no SMTP service".
314 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
315 if the DNS of the sending site is misconfigured. This is quite a
316 common situation. This change restores the behaviour of treating a
317 helo verification defer as a failure.
319 PH/51 If self=fail was set on a router, the bounce message did not include the
320 actual error message.
326 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
328 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
329 command itself ended in error, the underlying message about the transport
330 filter timeout was being overwritten with the pipe command error. Now the
331 underlying error message should be appended to the second error message.
333 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
335 PH/02 Reduce the amount of output that "make" produces by default. Full output
336 can still be requested.
338 PH/03 The warning log line about a condition test deferring for a "warn" verb
339 was being output only once per connection, rather than after each
340 occurrence (because it was using the same function as for successful
341 "warn" verbs). This seems wrong, so I have changed it.
343 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
344 it should not have, which might have caused a crash in the right
345 circumstances, but probably never did.
347 PH/04 Installed a modified version of Tony Finch's patch to make submission
348 mode fix the return path as well as the Sender: header line, and to
349 add a /name= option so that you can make the user's friendly name appear
352 TF/03 Added the control = fakedefer ACL modifier.
354 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
355 Mark Lowes for thorough testing.
357 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
359 TK/03 Merged latest SRS patch from Miles Wilton.
361 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
362 with the definition in sysexits.h (which is #included earlier).
363 Fortunately, Exim does not actually use EX_OK. The code used to try to
364 preserve the sysexits.h value, by assumimg that macro definitions were
365 scanned for macro replacements. I have been disabused of this notion,
366 so now the code just undefines EX_OK before #including unistd.h.
368 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
369 in the smtp transport. When a block could not be written in a single
370 write() function, the timeout was being re-applied to each part-write.
371 This seems wrong - if the receiver was accepting one byte at a time it
372 would take for ever. The timeout is now adjusted when this happens. It
373 doesn't have to be particularly precise.
375 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
376 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
378 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
379 without a domain if the check to <postmaster@domain> fails.
381 SC/01 Eximstats: added -xls and the ability to specify output files
382 (patch written by Frank Heydlauf).
384 SC/02 Eximstats: use FileHandles for outputing results.
386 SC/03 Eximstats: allow any combination of xls, txt, and html output.
388 SC/04 Eximstats: fixed display of large numbers with -nvr option
390 SC/05 Eximstats: fixed merging of reports with empty tables.
392 SC/06 Eximstats: added the -include_original_destination flag
394 SC/07 Eximstats: removed tabs and trailing whitespace.
396 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
398 TK/06 MBOX spool code: Add real "From " MBOX separator line
399 so the .eml file is really in mbox format (even though
400 most programs do not really care). Patch from Alex Miller.
402 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
403 The latter is generated from $received_to and is only set if the
404 message has one envelope recipient. SA can use these headers,
405 obviously out-of-the-box. Patch from Alex Miller.
407 PH/08 The ${def test on a variable was returning false if the variable's
408 value was "0", contrary to what the specification has always said!
409 The result should be true unless the variable is empty.
411 PH/09 The syntax error of a character other than { following "${if
412 def:variable_name" (after optional whitespace) was not being diagnosed.
413 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
414 accidental colon was present, for example, could give incorrect results.
416 PH/10 Tidied the code in a number of places where the st_size field of a stat()
417 result is used (not including appendfile, where other changes are about
420 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
421 This involved changing a lot of size variables from int to off_t. It
422 should work with maildirs and everything.
424 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
425 spamd dying while we are connected to it.
427 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
430 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
431 function for BDB 4.3.
433 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
435 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
436 the sources for PCRE have been re-arranged and more files are now
439 PH/15 The code I had for printing potentially long long variables in PH/11
440 above was not the best (it lost precision). The length of off_t variables
441 is now inspected at build time, and an appropriate printing format (%ld
442 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
443 to be "long long int" or "long int". This is needed for the internal
444 formatting function string_vformat().
446 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
447 the configuration file to be ":syslog", then the script "guesses" where
448 the logs files are, rather than using the compiled in default. In our
449 case the guess is not the same as the compiled default, so the script
450 suddenly stopped working when I started to use syslog. The patch checks
451 to see if log_file_path is "". If so, it attempts to read it from exim
452 with no configuration file to get the compiled in version, before it
453 falls back to the previous guessing code."
455 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
456 implementing BATV in an Exim configuration. See NewStuff for the gory
459 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
460 Makefile that are specific to HP-UX.
462 PH/18 If the "use_postmaster" option was set for a recipient callout together
463 with the "random" option, the postmaster address was used as the MAIL
464 FROM address for the random test, but not for the subsequent recipient
465 test. It is now used for both.
467 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
468 patch removes a few documentation additions to RFC 3028, because the
469 latest draft now contains them. It adds the new en;ascii-case comparator
470 and a new error check for 8bit text in MIME parts. Comparator and
471 require names are now matched exactly. I enabled the subaddress
472 extension, but it is not well tested yet (read: it works for me)."
474 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
475 rework some of the code of TK/09 above to avoid the hardwired use of
476 "%lld" and "long long". Replaced the call to snprintf() with a call to
479 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
480 records point to non-existent hosts", "retry timeout exceeded", and
481 "retry time not reached for any host after a long failure period".
483 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
484 experimental DomainKeys support:
486 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
487 (2) On an error such as an illegally used "control", the wrong name for
488 the control was given.
490 These problems did NOT occur unless DomainKeys support was compiled.
492 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
494 PH/24 Added ${if match_ip condition.
496 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
497 databases so that it will be absolutely obvious if a crash occurs in the
498 DB library. This is a regular occurrence (often caused by mis-matched
501 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
502 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
503 picked up on a user's system that detects such things. There doesn't seem
504 to be a gcc warning option for this - only an attribute that has to be
505 put on the function's prototype. It seems that in Fedora Core 4 they have
506 set this on a number of new functions. No doubt there will be more in due
509 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
510 specify a transport. However, if an address that was verified by such a
511 router was the subject of a callout, Exim crashed because it tried to
512 read the rcpt_include_affixes from the non-existent transport. Now it
513 just assumes that the setting of that option is false. This bug was
514 introduced by 4.51/PH/31.
516 PH/28 Changed -d+all to exclude +memory, because that information is very
517 rarely of interest, but it makes the output a lot bigger. People tend to
518 do -d+all out of habit.
520 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
521 code in os-type was giving problems when libc.so lives in lib64, like on
524 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
525 aren't the modern standard, and it seems that some systems' include files
526 don't always have them. Exim was already checking for some of the newer
527 ones like T_AAAA, and defining it itself. I've added checks for all the
528 record types that Exim uses.
530 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
531 not automatically generating a new one, as it is supposed to. This
532 prevented TLS from working. If the file did exist, but contained invalid
533 data, a new version was generated, as expected. It was only the case of a
534 non-existent file that was broken.
536 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
537 with a change in libdomainkeys > 0.64.
539 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
540 from DNS. If the selector record carries the flag, it now has
541 precedence over the domain-wide flag.
543 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
545 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
546 the use of a port name that isn't defined in /etc/services) occurred, the
547 message was deferred as in a normal delivery, and thus remained on the
548 spool, instead of being failed because of the mua_wrapper setting. This
549 is now fixed, and I tidied up some of the mua_wrapper messages at the
552 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
553 the messages in a hash of arrays rather than using individual hashes.
554 This is a bit cleaner and results in dramatic memory savings, albeit
555 at a slight CPU cost.
557 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
558 as requested by Marc Sherman.
560 SC/10 Eximstats: added histograms for user specified patterns as requested
563 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
565 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
566 fopen() in the content-scanning modules that did not already have it.
572 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
573 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
575 TK/02 Fix ACL "control" statement not being available in MIME ACL.
577 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
579 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
580 to test Sieve filters that use "vacation".
582 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
583 that changes the way the GnuTLS parameters are stored in the cache file.
584 The new format can be generated externally. For backward compatibility,
585 if the data in the cache doesn't make sense, Exim assumes it has read an
586 old-format file, and it generates new data and writes a new file. This
587 means that you can't go back to an older release without removing the
590 PH/03 A redirect router that has both "unseen" and "one_time" set does not
591 work if there are any delivery delays because "one_time" forces the
592 parent to be marked "delivered", so its unseen clone is never tried
593 again. For this reason, Exim now forbids the simultaneous setting of
596 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
597 redirected to themselves ("homonym" addresses). Read the long ChangeLog
598 entry if you want to know the details. The fix, however, neglected to
599 consider the case when local delivery batching is involved. The test for
600 "previously delivered" was not happening when checking to see if an
601 address could be batched with a previous (undelivered) one; under
602 certain circumstances this could lead to multiple deliveries to the same
605 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
606 in its include files, and this causes problems building Exim.
608 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
609 header_syntax) but Exim was just ignoring anything given after a slash.
610 In particular, this caused confusion with an attempt to use "verify =
611 reverse_host_lookup/defer_ok". An error is now given when options are
612 supplied for verify items that do not have them. (Maybe reverse_host_
613 lookup should have a defer_ok option, but that's a different point.)
615 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
616 defined by RFC 821) to 2048, because there were problems with some AUTH
617 commands, and RFC 1869 says the size should be increased for extended
618 SMTP commands that take arguments.
620 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
623 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
624 "unknown" error; now it says that the functionality isn't in the binary.
626 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
627 an address' error message when a string expansion fails (syntax or
628 whatever). Otherwise the password may appear in the log. Following change
629 PH/42 below, there is no longer a chance of it appearing in a bounce
632 PH/11 Installed exipick version 20050225.0 from John Jetmore.
634 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
635 of its addresses was ever tried. (Bugzilla bug #2.)
637 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
638 the result incorrectly in the debug output. (It correctly added a newline
639 to what was transported.)
641 TF/01 Added $received_time.
643 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
644 commented out examples of how to interface to a virus scanner and to
645 SpamAssassin. Also added commented examples of av_scanner and
646 spamd_address settings.
648 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
649 and controls are allowed in which ACLs. There were a couple of minor
650 errors. Some of the entries in the conditions table (which is a table of
651 where they are NOT allowed) were getting very unwieldy; rewrote them as a
652 negation of where the condition IS allowed.
654 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
656 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
657 header file does not have a version number, so I've had to invent a new
658 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
659 API. The code is untested by me (my Linux distribution still has 0.3.2 of
660 radiusclient), but it was contributed by a Radius user.
662 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
663 files or directories, for OpenSSL.
665 PH/19 When an Exim process that is running as root has to create an Exim log
666 file, it does so in a subprocess that runs as exim:exim so as to get the
667 ownership right at creation (otherwise, other Exim processes might see
668 the file with the wrong ownership). There was no test for failure of this
669 fork() call, which would lead to the process getting stuck as it waited
670 for a non-existent subprocess. Forks do occasionally fail when resources
671 run out. I reviewed all the other calls to fork(); they all seem to check
674 PH/20 When checking for unexpected SMTP input at connect time (before writing
675 the banner), Exim was not dealing correctly with a non-positive return
676 from the read() function. If the client had disconnected by this time,
677 the result was a log entry for a synchronization error with an empty
678 string after "input=" when read() returned zero. If read() returned -1
679 (an event I could not check), uninitialized data bytes were printed.
680 There were reports of junk text (parts of files, etc) appearing after
683 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
685 PH/22 Added support for macro redefinition, and (re)definition in between
686 driver and ACL definitions.
688 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
689 forgetting to use the resulting value; it was using the unexpanded value.
691 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
692 hadn't been configured. The fix is from Juergen Kreileder, who
693 understands it better than I do:
695 "Here's what I see happening with three configured cyrus_sasl
696 authenticators configured (plain, login, cram-md5):
698 On startup auth_cyrus_sasl_init() gets called for each of these.
699 This means three calls to sasl_listmech() without a specified mech_list.
700 => SASL tests which mechs of all available mechs actually work
701 => three warnings about OTP not working
702 => the returned list contains: plain, login, cram-md5, digest-md5, ...
704 With the patch, sasl_listmech() also gets called three times. But now
705 SASL's mech_list option is set to the server_mech specified in the the
706 authenticator. Or in other words, the answer from sasl_listmech()
707 gets limited to just the mech you're testing for (which is different
709 => the return list contains just 'plain' or 'login', 'cram-md5' or
710 nothing depending on the value of ob->server_mech.
712 I've just tested the patch: Authentication still works fine,
713 unavailable mechs specified in the exim configuration are still
714 caught, and the auth.log warnings about OTP are gone."
716 PH/25 When debugging is enabled, the contents of the command line are added
717 to the debugging output, even when log_selector=+arguments is not
720 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
721 answer is "GNU", and only if the return is "GNU/something" is the answer
724 PH/27 $acl_verify_message is now set immediately after the failure of a
725 verification in an ACL, and so is available in subsequent modifiers. In
726 particular, the message can be preserved by coding like this:
728 warn !verify = sender
729 set acl_m0 = $acl_verify_message
731 Previously, $acl_verify_message was set only while expanding "message"
732 and "log_message" when a very denied access.
734 PH/28 Modified OS/os.c-Linux with
736 -#ifndef OS_LOAD_AVERAGE
737 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
739 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
740 nomenclature these days.)
742 PH/29 Installed patch from the Sieve maintainer that adds the options
743 sieve_useraddress and sieve_subaddress to the redirect router.
745 PH/30 In these circumstances:
746 . Two addresses routed to the same list of hosts;
747 . First host does not offer TLS;
748 . First host accepts first address;
749 . First host gives temporary error to second address;
750 . Second host offers TLS and a TLS session is established;
751 . Second host accepts second address.
752 Exim incorrectly logged both deliveries with the TLS parameters (cipher
753 and peerdn, if requested) that were in fact used only for the second
756 PH/31 When doing a callout as part of verifying an address, Exim was not paying
757 attention to any local part prefix or suffix that was matched by the
758 router that accepted the address. It now behaves in the same way as it
759 does for delivery: the affixes are removed from the local part unless
760 rcpt_include_affixes is set on the transport.
762 PH/32 Add the sender address, as F=<...>, to the log line when logging a
763 timeout during the DATA phase of an incoming message.
765 PH/33 Sieve envelope tests were broken for match types other than :is. I have
766 applied a patch sanctioned by the Sieve maintainer.
768 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
769 the uid or gid is negative. A case of a negative gid caused this to be
770 noticed. The fix allows for either to be negative.
772 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
773 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
776 PH/36 The change PH/12 above was broken. Fixed it.
778 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
779 the grounds that routing the same address twice would always produce the
780 same answer. This might have been true once, but it is certainly no
781 longer true now. Routing a child address may depend on the previous
782 routing that produced that child. Some complicated redirection strategies
783 went wrong when messages had multiple recipients, and made Exim's
784 behaviour dependent on the order in which the addresses were given.
786 I have moved the duplicate checking until after the routing is complete.
787 Exim scans the addresses that are assigned to local and remote
788 transports, and removes any duplicates. This means that more work will be
789 done, as duplicates will always all be routed, but duplicates are
790 presumably rare, so I don't expect this is of any significance.
792 For deliveries to pipes, files, and autoreplies, the duplicate checking
793 still happens during the routing process, since they are not going to be
796 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
797 It corrects a timeout issue with spamd. This is Ian's comment: "The
798 background is that sometimes spamd either never reads data from a
799 connection it has accepted, or it never writes response data. The exiscan
800 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
801 blindly assumes that writes won't block so it may never time out."
803 PH/39 Allow G after quota size as well as K and M.
805 PH/40 The value set for $authenticated_id in an authenticator may not contain
806 binary zeroes or newlines because the value is written to log lines and
807 to spool files. There was no check on this. Now the value is run through
808 the string_printing() function so that such characters are converted to
809 printable escape sequences.
811 PH/41 $message_linecount is a new variable that contains the total number of
812 lines in the message. Compare $body_linecount, which is the count for the
815 PH/42 Exim no longer gives details of delivery errors for specific addresses in
816 bounce and delay warning messages, except in certain special cases, which
819 (a) An SMTP error message from a remote host;
820 (b) A message specified in a :fail: redirection;
821 (c) A message specified in a "fail" command in a system filter;
822 (d) A message specified in a FAIL return from the queryprogram router;
823 (e) A message specified by the cannot_route_message router option.
825 In these cases only, Exim does include the error details in bounce and
826 warning messages. There are also a few cases where bland messages such
827 as "unrouteable address" or "local delivery error" are given.
829 PH/43 $value is now also set for the "else" part of a ${run expansion.
831 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
832 being worked on, but at least Exim now implements the latest version to
835 PH/45 In a pipe transport, although a timeout while waiting for the pipe
836 process to complete was treated as a delivery failure, a timeout while
837 writing the message to the pipe was logged, but erroneously treated as a
838 successful delivery. Such timeouts include transport filter timeouts. For
839 consistency with the overall process timeout, these timeouts are now
840 treated as errors, giving rise to delivery failures by default. However,
841 there is now a new Boolean option for the pipe transport called
842 timeout_defer, which, if set TRUE, converts the failures into defers for
843 both kinds of timeout. A transport filter timeout is now identified in
846 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
847 systems where "make" and "gmake" are different, calling "gmake" at top
848 level broke things. I've arranged for the value of $(MAKE) to be passed
849 from the Makefile to this script so that it can call the same version of
853 A note about Exim versions 4.44 and 4.50
854 ----------------------------------------
856 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
857 changes of various kinds. As a consequence, a big documentation update was
858 needed. This delayed the release for rather longer than seemed good, especially
859 in the light of a couple of (minor) security issues. Therefore, the changes
860 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
861 release. So 4.44 and 4.50 are in effect two different branches that both start
864 I have left the 4.50 change log unchanged; it contains all the changes since
865 4.43. The change log for 4.44 is below; many of its items are identical to
866 those for 4.50. This seems to be the most sensible way to preserve the
867 historical information.
873 1. Minor wording change to the doc/README.SIEVE file.
875 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
876 computation of the current number of files was incorrect.
878 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
879 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
882 4. Give more explanation in the error message when the command for a transport
883 filter fails to execute.
885 5. There are several places where Exim runs a non-Exim command in a
886 subprocess. The SIGUSR1 signal should be disabled for these processes. This
887 was being done only for the command run by the queryprogram router. It is
888 now done for all such subprocesses. The other cases are: ${run, transport
889 filters, and the commands run by the lmtp and pipe transports.
891 6. Added CONFIGURE_GROUP build-time option.
893 7. Some older OS have a limit of 256 on the maximum number of file
894 descriptors. Exim was using setrlimit() to set 1000 as a large value
895 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
896 systems. I've change it so that if it can't get 1000, it tries for 256.
898 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
899 was an oversight, and furthermore, ever since the addition of extra
900 controls (e.g. 4.43/32), the checks on when to allow different forms of
901 "control" were broken. There should now be diagnostics for all cases when a
902 control that does not make sense is encountered.
904 9. Added the /retain_sender option to "control=submission".
906 10. $recipients is now available in the predata ACL (oversight).
908 11. Tidy the search cache before the fork to do a delivery from a message
909 received from the command line. Otherwise the child will trigger a lookup
910 failure and thereby defer the delivery if it tries to use (for example) a
911 cached ldap connection that the parent has called unbind on.
913 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
914 of $address_data from the recipient verification was clobbered by the
917 13. The value of address_data from a sender verification is now available in
918 $sender_address_data in subsequent conditions in the ACL statement.
920 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
922 15. Added a new option "connect=<time>" to callout options, to set a different
925 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
926 was its contents. (It was OK if the option was not defined at all.)
928 17. A "Completed" log line is now written for messages that are removed from
929 the spool by the -Mrm option.
931 18. New variables $sender_verify_failure and $recipient_verify_failure contain
932 information about exactly what failed.
934 19. Added -dd to debug only the daemon process.
936 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
937 handles timeouts, both on the server side and network timeouts. Renamed the
938 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
940 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
941 It is now set to "smtps".
943 22. $host_address is now set to the target address during the checking of
946 23. When checking ignore_target_hosts for an ipliteral router, no host name was
947 being passed; this would have caused $sender_host_name to have been used if
948 matching the list had actually called for a host name (not very likely,
949 since this list is usually IP addresses). A host name is now passed as
952 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
953 code that specifies a non-restarting handler (typically sigaction() in
954 modern systems) in an attempt to fix a rare and obscure crash bug.
956 25. Narrowed the window for a race in the daemon that could cause it to ignore
957 SIGCHLD signals. This is not a major problem, because they are used only to
958 wake it up if nothing else does.
960 26. A malformed maildirsize file could cause Exim to calculate negative values
961 for the mailbox size or file count. Odd effects could occur as a result.
962 The maildirsize information is now recalculated if the size or filecount
965 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
966 support for a long time. Removed HAVE_SYS_VFS_H.
968 28. Installed the latest version of exipick from John Jetmore.
970 29. In an address list, if the pattern was not a regular expression, an empty
971 subject address (from a bounce message) matched only if the pattern was an
972 empty string. Non-empty patterns were not even tested. This was the wrong
973 because it is perfectly reasonable to use an empty address as part of a
974 database query. An empty address is now tested by patterns that are
975 lookups. However, all the other forms of pattern expect the subject to
976 contain a local part and a domain, and therefore, for them, an empty
977 address still always fails if the pattern is not itself empty.
979 30. Exim went into a mad DNS loop when attempting to do a callout where the
980 host was specified on an smtp transport, and looking it up yielded more
983 31. Re-factored the code for checking spool and log partition space into a
984 function that finds that data and another that does the check. The former
985 is then used to implement four new variables: $spool_space, $log_space,
986 $spool_inodes, and $log_inodes.
988 32. The RFC2047 encoding function was originally intended for short strings
989 such as real names; it was not keeping to the 75-character limit for
990 encoded words that the RFC imposes. It now respects the limit, and
991 generates multiple encoded words if necessary. To be on the safe side, I
992 have increased the buffer size for the ${rfc2047: expansion operator from
995 33. It is now permitted to omit both strings after an "if" condition; if the
996 condition is true, the result is "true". As before, when the second string
997 is omitted, a false condition yields an empty string. This makes it less
998 cumbersome to write custom ACL and router conditions.
1000 34. Failure to deliver a bounce message always caused it to be frozen, even if
1001 there was an errors_to setting on the router. The errors_to setting is now
1004 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1005 canonical form (fully expanded) before being placed in
1006 $sender_host_address.
1008 36. The table in the code that translates DNS record types into text (T_A to
1009 "A" for instance) was missing entries for NS and CNAME. It is just possible
1010 that this could have caused confusion if both these types were looked up
1011 for the same domain, because the text type is used as part of Exim's
1012 per-process caching. But the chance of anyone hitting this buglet seems
1015 37. The dnsdb lookup has been extended in a number of ways.
1017 (1) There is a new type, "zns", which walks up the domain tree until it
1018 finds some nameserver records. It should be used with care.
1020 (2) There is a new type, "mxh", which is like "mx" except that it returns
1021 just the host names, not the priorities.
1023 (3) It is now possible to give a list of domains (or IP addresses) to be
1024 looked up. The behaviour when one of the lookups defers can be
1025 controlled by a keyword.
1027 (4) It is now possible to specify the separator character for use when
1028 multiple records are returned.
1030 38. The dnslists ACL condition has been extended: it is now possible to supply
1031 a list of IP addresses and/or domains to be looked up in a particular DNS
1034 39. Added log_selector=+queue_time_overall.
1036 40. When running the queue in the test harness, wait just a tad after forking a
1037 delivery process, to get repeatability of debugging output.
1039 41. Include certificate and key file names in error message when GnuTLS fails
1040 to set them up, because the GnuTLS error message doesn't include the name
1041 of the failing file when there is a problem reading it.
1043 42. Allow both -bf and -bF in the same test run.
1045 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1047 44. The "Exiscan patch" is now merged into the mainline Exim source.
1049 45. Sometimes the final signoff response after QUIT could fail to get
1050 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1051 < 0 before doing a fflush(). This bug looks as though it goes back to the
1052 introduction of TLS in release 3.20, but "sometimes" must have been rare
1053 because the tests only now provoked it.
1055 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1056 (this can affect the format of dates).
1058 47. exim_tidydb, when checking for the continued existence of a message for
1059 which it has found a message-specific retry record, was not finding
1060 messages that were in split spool directories. Consequently, it was
1061 deleting retry records that should have stayed in existence.
1063 48. Steve fixed some bugs in eximstats.
1065 49. The SPA authentication driver was not abandoning authentication and moving
1066 on to the next authenticator when an expansion was forced to fail,
1067 contradicting the general specification for all authenticators. Instead it
1068 was generating a temporary error. It now behaves as specified.
1070 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1071 (the order specifies the preference for clients). The order is now AES256,
1072 AES128, 3DES, ARCFOUR128.
1074 51. Small patch to Sieve code - explicitly set From: when generating an
1077 52. Exim crashed if a remote delivery caused a very long error message to be
1078 recorded - for instance if somebody sent an entire SpamAssassin report back
1079 as a large number of 550 error lines. This bug was coincidentally fixed by
1080 increasing the size of one of Exim's internal buffers (big_buffer) that
1081 happened as part of the Exiscan merge. However, to be on the safe side, I
1082 have made the code more robust (and fixed the comments that describe what
1085 53. Now that there can be additional text after "Completed" in log lines (if
1086 the queue_time_overall log selector is set), a one-byte patch to exigrep
1087 was needed to allow it to recognize "Completed" as not the last thing in
1090 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1091 patch that reportedly fixes this has been added. I am not expert enough to
1092 create a test for it. This is what the patch creator wrote:
1094 "I found a little strange behaviour of ldap code when working with
1095 Windows 2003 AD Domain, where users was placed in more than one
1096 Organization Units. When I tried to give exim partial DN, the exit code
1097 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1098 But simultaneously result of request was absolutely normal ldap result,
1099 so I produce this patch..."
1101 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1102 REFERENCE, so I have modified the code to exclude the patch when that macro
1105 55. Some experimental protocols are using DNS PTR records for new purposes. The
1106 keys for these records are domain names, not reversed IP addresses. The
1107 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1108 leaves it alone. Component reversal etc. now happens only for IP addresses.
1111 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1113 57. Double the size of the debug message buffer (to 2048) so that more of very
1114 long debug lines gets shown.
1116 58. The exicyclog utility now does better if the number of log files to keep
1117 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1119 59. Two changes related to the smtp_active_hostname option:
1121 (1) $smtp_active_hostname is now available as a variable.
1122 (2) The default for smtp_banner uses $smtp_active_hostname instead
1123 of $primary_hostname.
1125 60. The host_aton() function is supposed to be passed a string that is known
1126 to be a valid IP address. However, in the case of IPv6 addresses, it was
1127 not checking this. This is a hostage to fortune. Exim now panics and dies
1128 if the condition is not met. A case was found where this could be provoked
1129 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1130 components; fortuitously, this particular loophole had already been fixed
1131 by change 4.50/55 above.
1133 If there are any other similar loopholes, the new check in host_aton()
1134 itself should stop them being exploited. The report I received stated that
1135 data on the command line could provoke the exploit when Exim was running as
1136 exim, but did not say which command line option was involved. All I could
1137 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1138 running as the user.
1141 61. There was a buffer overflow vulnerability in the SPA authentication code
1142 (which came originally from the Samba project). I have added a test to the
1143 spa_base64_to_bits() function which I hope fixes it.
1146 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1147 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1149 63. The daemon start-up calls getloadavg() while still root for those OS that
1150 need the first call to be done as root, but it missed one case: when
1151 deliver_queue_load_max is set with deliver_drop_privilege. This is
1152 necessary for the benefit of the queue runner, because there is no re-exec
1153 when deliver_drop_privilege is set.
1155 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1156 This has been fixed.
1158 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1159 was in use, was not putting the data itself into the right store pool;
1160 consequently, it could be overwritten for a subsequent message in the same
1161 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1164 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1166 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1167 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1168 the function were treating the return as a boolean value, which happened to
1169 work because 0=false and not-0=true, but is not correct code.
1171 68. The host_aton() function was not handling scoped IPv6 addresses (those
1172 with, for example, "%eth0" on the end) correctly.
1174 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1175 negated items (that is, ~something) in unsigned ints. Some compilers
1176 apparently mutter when there is no cast.
1178 70. If an address verification called from an ACL failed, and did not produce a
1179 user-specific message (i.e. there was only a "system" message), nothing was
1180 put in $acl_verify_message. In this situation, it now puts the system
1183 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1184 session; change 4.31/43 added the unwanted input to the log line - except
1185 that it did not do this in the start of session case. It now does.
1187 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1188 This is wrong and can cause the other end to generate a synchronization
1189 error if it is another Exim or anything else that does the synchronization
1190 check. A QUIT command is no longer sent after a timeout.
1192 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1193 during host lookups.
1195 74. The defer_ok option of callout verification was not working if it was used
1196 when verifying addresses in header lines, that is, for this case:
1198 verify = header_sender/callout=defer_ok
1200 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1201 those file descriptors could be used for SMTP connections. If anything
1202 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1203 could be sent to the SMTP client, causing chaos. The daemon now opens
1204 stdin, stdout, and stderr to /dev/null when it puts itself into the
1207 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1208 log by default. The user can override this with suitable Perl magic.
1210 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1211 the log message when discard triggers, was not working for the DATA ACL or
1212 for the non-SMTP ACL.
1214 78. Error message wording change in sieve.c.
1216 79. If smtp_accept_max_per_host was set, the number of connections could be
1217 restricted to fewer than expected, because the daemon was trying to set up
1218 a new connection before checking whether the processes handling previous
1219 connections had finished. The check for completed processes is now done
1220 earlier. On busy systems, this bug wouldn't be noticed because something
1221 else would have woken the daemon, and it would have reaped the completed
1224 80. If a message was submitted locally by a user whose login name contained one
1225 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1226 It caused a spool format error. I have fixed the spool reading code. A
1227 related problem was that the "from" clause in the Received: line became
1228 illegal because of the space(s). It is now covered by ${quote_local_part.
1230 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1233 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1235 83. Patch from Sieve maintainer to fix unterminated string problem in
1236 "vacation" handling.
1238 84. Some minor changes to the Linux configuration files to help with other
1239 OS variants using glibc.
1241 85. One more patch for Sieve to update vacation handling to latest spec.
1244 ----------------------------------------------------
1245 See the note above about the 4.44 and 4.50 releases.
1246 ----------------------------------------------------
1252 1. Change 4.43/35 introduced a bug that caused file counts to be
1253 incorrectly computed when quota_filecount was set in an appendfile
1256 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1257 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1260 3. Give more explanation in the error message when the command for a transport
1261 filter fails to execute.
1263 4. There are several places where Exim runs a non-Exim command in a
1264 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1265 was being done only for the command run by the queryprogram router. It is
1266 now done for all such subprocesses. The other cases are: ${run, transport
1267 filters, and the commands run by the lmtp and pipe transports.
1269 5. Some older OS have a limit of 256 on the maximum number of file
1270 descriptors. Exim was using setrlimit() to set 1000 as a large value
1271 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1272 systems. I've change it so that if it can't get 1000, it tries for 256.
1274 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1275 was an oversight, and furthermore, ever since the addition of extra
1276 controls (e.g. 4.43/32), the checks on when to allow different forms of
1277 "control" were broken. There should now be diagnostics for all cases when a
1278 control that does not make sense is encountered.
1280 7. $recipients is now available in the predata ACL (oversight).
1282 8. Tidy the search cache before the fork to do a delivery from a message
1283 received from the command line. Otherwise the child will trigger a lookup
1284 failure and thereby defer the delivery if it tries to use (for example) a
1285 cached ldap connection that the parent has called unbind on.
1287 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1288 of $address_data from the recipient verification was clobbered by the
1289 sender verification.
1291 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1292 was its contents. (It was OK if the option was not defined at all.)
1294 11. A "Completed" log line is now written for messages that are removed from
1295 the spool by the -Mrm option.
1297 12. $host_address is now set to the target address during the checking of
1298 ignore_target_hosts.
1300 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1301 being passed; this would have caused $sender_host_name to have been used if
1302 matching the list had actually called for a host name (not very likely,
1303 since this list is usually IP addresses). A host name is now passed as
1306 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1307 code that specifies a non-restarting handler (typically sigaction() in
1308 modern systems) in an attempt to fix a rare and obscure crash bug.
1310 15. Narrowed the window for a race in the daemon that could cause it to ignore
1311 SIGCHLD signals. This is not a major problem, because they are used only to
1312 wake it up if nothing else does.
1314 16. A malformed maildirsize file could cause Exim to calculate negative values
1315 for the mailbox size or file count. Odd effects could occur as a result.
1316 The maildirsize information is now recalculated if the size or filecount
1319 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1320 support for a long time. Removed HAVE_SYS_VFS_H.
1322 18. Updated exipick to current release from John Jetmore.
1324 19. Allow an empty sender to be matched against a lookup in an address list.
1325 Previously the only cases considered were a regular expression, or an
1328 20. Exim went into a mad DNS lookup loop when doing a callout where the
1329 host was specified on the transport, if the DNS lookup yielded more than
1332 21. The RFC2047 encoding function was originally intended for short strings
1333 such as real names; it was not keeping to the 75-character limit for
1334 encoded words that the RFC imposes. It now respects the limit, and
1335 generates multiple encoded words if necessary. To be on the safe side, I
1336 have increased the buffer size for the ${rfc2047: expansion operator from
1339 22. Failure to deliver a bounce message always caused it to be frozen, even if
1340 there was an errors_to setting on the router. The errors_to setting is now
1343 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1344 canonical form (fully expanded) before being placed in
1345 $sender_host_address.
1347 24. Updated eximstats to version 1.33
1349 25. Include certificate and key file names in error message when GnuTLS fails
1350 to set them up, because the GnuTLS error message doesn't include the name
1351 of the failing file when there is a problem reading it.
1353 26. Expand error message when OpenSSL has problems setting up cert/key files.
1356 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1357 (this can affect the format of dates).
1359 28. exim_tidydb, when checking for the continued existence of a message for
1360 which it has found a message-specific retry record, was not finding
1361 messages that were in split spool directories. Consequently, it was
1362 deleting retry records that should have stayed in existence.
1364 29. eximstats updated to version 1.35
1365 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1366 1.35 - bugfix such that pie charts by volume are generated correctly
1368 30. The SPA authentication driver was not abandoning authentication and moving
1369 on to the next authenticator when an expansion was forced to fail,
1370 contradicting the general specification for all authenticators. Instead it
1371 was generating a temporary error. It now behaves as specified.
1373 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1374 (the order specifies the preference for clients). The order is now AES256,
1375 AES128, 3DES, ARCFOUR128.
1377 31. Small patch to Sieve code - explicitly set From: when generating an
1380 32. Exim crashed if a remote delivery caused a very long error message to be
1381 recorded - for instance if somebody sent an entire SpamAssassin report back
1382 as a large number of 550 error lines. This bug was coincidentally fixed by
1383 increasing the size of one of Exim's internal buffers (big_buffer) that
1384 happened as part of the Exiscan merge. However, to be on the safe side, I
1385 have made the code more robust (and fixed the comments that describe what
1388 33. Some experimental protocols are using DNS PTR records for new purposes. The
1389 keys for these records are domain names, not reversed IP addresses. The
1390 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1391 leaves it alone. Component reversal etc. now happens only for IP addresses.
1394 34. The host_aton() function is supposed to be passed a string that is known
1395 to be a valid IP address. However, in the case of IPv6 addresses, it was
1396 not checking this. This is a hostage to fortune. Exim now panics and dies
1397 if the condition is not met. A case was found where this could be provoked
1398 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1399 components; fortuitously, this particular loophole had already been fixed
1400 by change 4.50/55 or 4.44/33 above.
1402 If there are any other similar loopholes, the new check in host_aton()
1403 itself should stop them being exploited. The report I received stated that
1404 data on the command line could provoke the exploit when Exim was running as
1405 exim, but did not say which command line option was involved. All I could
1406 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1407 running as the user.
1410 35. There was a buffer overflow vulnerability in the SPA authentication code
1411 (which came originally from the Samba project). I have added a test to the
1412 spa_base64_to_bits() function which I hope fixes it.
1415 36. The daemon start-up calls getloadavg() while still root for those OS that
1416 need the first call to be done as root, but it missed one case: when
1417 deliver_queue_load_max is set with deliver_drop_privilege. This is
1418 necessary for the benefit of the queue runner, because there is no re-exec
1419 when deliver_drop_privilege is set.
1421 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1422 was in use, was not putting the data itself into the right store pool;
1423 consequently, it could be overwritten for a subsequent message in the same
1424 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1427 38. Sometimes the final signoff response after QUIT could fail to get
1428 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1429 < 0 before doing a fflush(). This bug looks as though it goes back to the
1430 introduction of TLS in release 3.20, but "sometimes" must have been rare
1431 because the tests only now provoked it.
1437 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1438 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1439 Now it returns an integer. A number of calls were still expecting a T/F
1440 return. Fortuitously, in all cases, the tests worked in OK situations,
1441 which is the norm. However, things would have gone wrong on any write
1442 failures on the smtp file descriptor. This function is used when sending
1443 messages over SMTP and also when doing verify callouts.
1445 2. When Exim is called to do synchronous delivery of a locally submitted
1446 message (the -odf or -odi options), it no longer closes stderr before doing
1449 3. Implemented the mua_wrapper option.
1451 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1453 5. Implemented the functions header_remove(), header_testname(),
1454 header_add_at_position(), and receive_remove_recipient(), and exported them
1457 6. If an ACL "warn" statement specified the addition of headers, Exim already
1458 inserted X-ACL-Warn: at the start if there was no header name. However, it
1459 was not making this test for the second and subsequent header lines if
1460 there were newlines in the string. This meant that an invalid header could
1461 be inserted if Exim was badly configured.
1463 7. Allow an ACL "warn" statement to add header lines at the start or after all
1464 the Received: headers, as well as at the end.
1466 8. Added the rcpt_4xx retry error code.
1468 9. Added postmaster_mailfrom=xxx to callout verification option.
1470 10. Added mailfrom=xxxx to the callout verification option, for verify=
1473 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1474 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1476 12. Inserted some casts to stop certain compilers warning when using pointer
1477 differences as field lengths or precisions in printf-type calls (mostly
1478 affecting debugging statements).
1480 13. Added optional readline() support for -be (dynamically loaded).
1482 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1483 same clock tick as a message's arrival, so that its received time was the
1484 same as the "first fail" time on the retry record, and that message
1485 remained on the queue past the ultimate address timeout, every queue runner
1486 would try a delivery (because it was past the ultimate address timeout) but
1487 after another failure, the ultimate address timeout, which should have then
1488 bounced the address, did not kick in. This was a "< instead of <=" error;
1489 in most cases the first failure would have been in the next clock tick
1490 after the received time, and all would be well.
1492 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1493 being recognized when the domain list was tested by the match_domain
1494 condition in an expansion string.
1496 16. Added the ${str2b64: operator.
1498 17. Exim was always calling setrlimit() to set a large limit for the number of
1499 processes, without checking whether the existing limit was already
1500 adequate. (It did check for the limit on file descriptors.) Furthermore,
1501 errors from getrlimit() and setrlimit() were being ignored. Now they are
1502 logged to the main and panic logs, but Exim does carry on, to try to do its
1503 job under whatever limits there are.
1505 18. Imported PCRE 5.0.
1507 19. Trivial typo in log message " temporarily refused connection" (the leading
1510 20. If the log selector return_path_on_delivery was set and an address was
1511 redirected to /dev/null, the delivery process crashed because it assumed
1512 that a return path would always be set for a "successful" delivery. In this
1513 case, the whole delivery is bypassed as an optimization, and therefore no
1516 21. Internal re-arrangement: the function for sending a challenge and reading
1517 a response while authentication was assuming a zero-terminated challenge
1518 string. It's now changed to take a pointer and a length, to allow for
1519 binary data in such strings.
1521 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1523 23. Exim was not respecting finduser_retries when seeking the login of the
1524 uid under which it was called; it was always trying 10 times. (The default
1525 setting of finduser_retries is zero.) Also, it was sleeping after the final
1526 failure, which is pointless.
1528 24. Implemented tls_on_connect_ports.
1530 25. Implemented acl_smtp_predata.
1532 26. If the domain in control=submission is set empty, Exim assumes that the
1533 authenticated id is a complete email address when it generates From: or
1534 Sender: header lines.
1536 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1537 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1538 chown and chgrp in /bin and hostname in /usr/bin.
1540 28. Exim was keeping the "process log" file open after each use, just as it
1541 does for the main log. This opens the possibility of it remaining open for
1542 long periods when the USR1 signal hits a daemon. Occasional processlog
1543 errors were reported, that could have been caused by this. Anyway, it seems
1544 much more sensible not to leave this file open at all, so that is what now
1547 29. The long-running daemon process does not normally write to the log once it
1548 has entered its main loop, and it closes the log before doing so. This is
1549 so that log files can straightforwardly be renamed and moved. However,
1550 there are a couple of unusual error situations where the daemon does write
1551 log entries, and I had neglected to close the log afterwards.
1553 30. The text of an SMTP error response that was received during a remote
1554 delivery was being truncated at 512 bytes. This is too short for some of
1555 the long messages that one sometimes sees. I've increased the limit to
1558 31. It is now possible to make retry rules that apply only when a message has a
1559 specific sender, in particular, an empty sender.
1561 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1562 it possible to be selective about when SMTP synchronization is enforced.
1564 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1566 32. Implemented hosts_connection_nolog.
1568 33. Added an ACL for QUIT.
1570 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1573 35. Added mailbox_size and mailbox_filecount to appendfile.
1575 36. Added control = no_multiline_responses to ACLs.
1577 37. There was a bug in the logic of the code that waits for the clock to tick
1578 in the case where the clock went backwards by a substantial amount such
1579 that the microsecond fraction of "now" was more than the microsecond
1580 fraction of "then" (but the whole seconds number was less).
1582 38. Added support for the libradius Radius client library this is found on
1583 FreeBSD (previously only the radiusclient library was supported).
1589 1. When certain lookups returned multiple values in the form name=value, the
1590 quoting of the values was not always being done properly. Specifically:
1591 (a) If the value started with a double quote, but contained no whitespace,
1593 (b) If the value contained whitespace other than a space character (i.e.
1594 tabs or newlines or carriage returns) it was not quoted.
1595 This fix has been applied to the mysql and pgsql lookups by writing a
1596 separate quoting function and calling it from the lookup code. The fix
1597 should probably also be applied to nisplus, ibase and oracle lookups, but
1598 since I cannot test any of those, I have not disturbed their existing code.
1600 2. A hit in the callout cache for a specific address caused a log line with no
1601 reason for rejecting RCPT. Now it says "Previous (cached) callout
1602 verification failure".
1604 3. There was an off-by-one bug in the queryprogram router. An over-long
1605 return line was truncated at 256 instead of 255 characters, thereby
1606 overflowing its buffer with the terminating zero. As well as fixing this, I
1607 have increased the buffer size to 1024 (and made a note to document this).
1609 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1610 when Exim is waiting for an SMTP response from a remote server, Exim
1611 restarts its select() call on the socket, thereby resetting its timeout.
1612 This is not a problem when such interrupts are rare. Somebody set up a cron
1613 job to run exiwhat every 2 minutes, which is less than the normal select()
1614 timeout (5 or 10 minutes). This meant that the select() timeout never
1615 kicked in because it was always reset. I have fixed this by comparing the
1616 time when an interrupt arrives with the time at the start of the first call
1617 to select(). If more time than the timeout has elapsed, the interrupt is
1618 treated as a timeout.
1620 5. Some internal re-factoring in preparation for the addition of Sieve
1621 extensions (by MH). In particular, the "personal" test is moved to a
1622 separate function, and given an option for scanning Cc: and Bcc: (which is
1623 not set for Exim filters).
1625 6. When Exim created an email address using the login of the caller as the
1626 local part (e.g. when creating a From: or Sender: header line), it was not
1627 quoting the local part when it contained special characters such as @.
1629 7. Installed new OpenBSD configuration files.
1631 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1632 try to make them clearer.
1634 9. Callout options, other than the timeout value, were being ignored when
1635 verifying sender addresses in header lines. For example, when using
1637 verify = header_sender/callout=no_cache
1639 the cache was (incorrectly) being used.
1641 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1642 only the Cygwin environment.
1644 11. When return_path_on_delivery was set as a log selector, if different remote
1645 addresses in the same message used different return paths and parallel
1646 remote delivery occurred, the wrong values would sometimes be logged.
1647 (Whenever a remote delivery process finished, the return path value from
1648 the most recently started remote delivery process was logged.)
1650 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1651 lines when AUTH and/or TLS are in use. This is the "received protocol"
1652 field. Exim used to use "asmtp" for authenticated SMTP, without any
1653 indication (in the protocol name) for TLS use. Now it follows the RFC and
1654 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1655 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1656 names appear in log lines as well as in Received: header lines.
1658 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1659 extensions, and comparison tests, and to fix some bugs.
1661 14. Changes to the "personal" filter test:
1663 (1) The test was buggy in that it was just doing the equivalent of
1664 "contains" tests on header lines. For example, if a user's address was
1665 anne@some.where, the "personal" test would incorrectly be true for
1667 To: susanne@some.where
1669 This test is now done by extracting each address from the header in turn,
1670 and checking the entire address. Other tests that are part of "personal"
1671 are now done using regular expressions (for example, to check local parts
1672 of addresses in From: header lines).
1674 (2) The list of non-personal local parts in From: addresses has been
1675 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1676 taken from the Sieve specification recommendations.
1678 (3) If the message contains any header line starting with "List-" it is
1679 treated as non-personal.
1681 (4) The test for "circular" in the Subject: header line has been removed
1682 because it now seems ill-conceived.
1684 15. Minor typos in src/EDITME comments corrected.
1686 16. Installed latest exipick from John Jetmore.
1688 17. If headers_add on a router specified a text string that was too long for
1689 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1690 of string_sprintf() is now avoided.
1692 18. $message_body_size was not set (it was always zero) when running the DATA
1693 ACL and the local_scan() function.
1695 19. For the "mail" command in an Exim filter, no default was being set for
1696 the once_repeat time, causing a random time value to be used if "once" was
1697 specified. (If the value happened to be <= 0, no repeat happened.) The
1698 default is now 0s, meaning "never repeat". The "vacation" command was OK
1699 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1700 (I found it when inspecting the code).
1702 20. There is now an overall timeout for performing a callout verification. It
1703 defaults to 4 times the callout timeout, which applies to individual SMTP
1704 commands during the callout. The overall timeout applies when there is more
1705 than one host that can be tried. The timeout is checked before trying the
1706 next host. This prevents very long delays if there are a large number of
1707 hosts and all are timing out (e.g. when the network connections are timing
1708 out). The value of the overall timeout can be changed by specifying an
1709 additional sub-option for "callout", called "maxwait". For example:
1711 verify = sender/callout=5s,maxwait=20s
1713 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1714 to the end before writing, but this should make it even safer).
1716 22. Exim was forgetting that it had advertised PIPELINING for the second and
1717 subsequent messages on an SMTP connection. It was also not resetting its
1718 memory on STARTTLS and an internal HELO.
1720 23. When Exim logs an SMTP synchronization error within a session, it now
1721 records whether PIPELINING has been advertised or not.
1723 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1724 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1725 rather than long int.
1727 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1729 26. Added the never_mail option to autoreply.
1735 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1736 crash if the getsockname() call failed; this can happen if a connection is
1737 closed very soon after it is established. The problem was simply in the
1738 order in which certain operations were done, causing Exim to try to write
1739 to the SMTP stream before it had set up the file descriptor. The bug has
1740 been fixed by making things happen in the correct order.
1746 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1747 before the connection was closed, thus losing the rejection response.
1749 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1750 some early Solaris releases, but causes trouble in current releases where
1751 socklen_t is defined.
1753 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1756 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1759 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1760 that releasing the top of it at the end releases what was used for sub-
1761 expansions (unless the block got too big). However, discard this block if
1762 the first thing is a variable or header, so that we can use its block when
1763 it is dynamic (useful for very large $message_headers, for example).
1765 6. Lookups now cache *every* query, not just the most recent. A new, separate
1766 store pool is used for this. It can be recovered when all lookup caches are
1767 flushed. Lookups now release memory at the end of their result strings.
1768 This has involved some general refactoring of the lookup sources.
1770 7. Some code has been added to the store_xxx() functions to reduce the amount
1771 of flapping under certain conditions.
1773 8. log_incoming_interface used to affect only the <= reception log lines. Now
1774 it causes the local interface and port to be added to several more SMTP log
1775 lines, for example "SMTP connection from", and rejection lines.
1777 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1779 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1781 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1782 could be overwritten at the end of the current message (or the start of a
1783 new message if it was set in a HELO ACL). The value is now preserved for
1784 the duration of the SMTP connection.
1786 12. If a transport had a headers_rewrite setting, and a matching header line
1787 contained an unqualified address, that address was qualified, even if it
1788 did not match any rewriting rules. The underlying bug was that the values
1789 of the flags that permit the existence of unqualified sender and recipient
1790 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1791 non-local messages, and by -bnq for local messages) were not being
1792 preserved with the message after it was received.
1794 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1795 "next input=" as part of the text comprising the host identity instead of
1796 the correct text. The code was using the same buffer for two different
1797 strings. However, depending on which order the printing function evaluated
1798 its arguments, the bug did not always show up. Under Linux, for example, my
1799 test suite worked just fine.
1801 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1802 doesn't work with some older versions of Perl. It has been changed to "my",
1803 which in any case is probably the better facility to use.
1805 15. A really picky compiler found some instances of statements for creating
1806 error messages that either had too many or two few arguments for the format
1809 16. The size of the buffer for calls to the DNS resolver has been increased
1810 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1811 for addresses that have a lot of PTR records. This alleviates a problem; it
1812 does not fully solve it.
1814 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1815 buffer now truncates the list and logs the incident, which is the same
1816 action as happens when Exim is looking up a host name and its aliases.
1817 Previously in this situation something unpredictable would happen;
1818 sometimes it was "internal error: store_reset failed".
1820 18. If a server dropped the connection unexpectedly when an Exim client was
1821 using GnuTLS and trying to read a response, the client delivery process
1822 crashed while trying to generate an error log message.
1824 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1825 string, for example:
1827 warn message = H1: something\nH2: something
1829 the text was added as a single header line from Exim's point of view
1830 though it ended up OK in the delivered message. However, searching for the
1831 second and subsequent header lines using $h_h2: did not work. This has been
1832 fixed. Similarly, if a system filter added multiple headers in this way,
1833 the routers could not see them.
1835 20. Expanded the error message when iplsearch is called with an invalid key to
1836 suggest using net-iplsearch in a host list.
1838 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1839 ACLs are no longer actually imposed (and a message to that effect is
1842 22. If a "gecos" field in a passwd entry contained escaped characters, in
1843 particular, if it contained a \" sequence, Exim got it wrong when building
1844 a From: or a Sender: header from that name. A second bug also caused
1845 incorrect handling when an unquoted " was present following a character
1846 that needed quoting.
1848 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1849 was not being matched caselessly.
1851 24. Arranged for all hyphens in the exim.8 source to be escaped with
1854 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1855 properly. Recipient callout cache records were still being keyed to include
1856 the sender, even when use_sender was set false. This led to far more
1857 callouts that were necessary. The sender is no longer included in the key
1858 when use_sender is false.
1860 26. Added "control = submission" modifier to ACLs.
1862 27. Added the ${base62d: operator to decode base 62 numbers.
1864 28. dnsdb lookups can now access SRV records.
1866 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1867 the configuration file.
1869 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1870 (-v) mode. This makes the output for a verbose queue run more intelligible.
1872 31. Added a use_postmaster feature to recipient callouts.
1874 32. Added the $body_zerocount variable, containing the number of binary zero
1875 bytes in the message body.
1877 33. The time of last modification of the "new" subdirectory is now used as the
1878 "mailbox time last read" when there is a quota error for a maildir
1881 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1883 35. Added +ignore_unknown as a special item in host lists.
1885 36. Code for decoding IPv6 addresses in host lists is now included, even if
1886 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1887 address was recognized as an IP address, but was then not correctly decoded
1888 into binary, causing unexpected and incorrect effects when compared with
1895 1. Very minor rewording of debugging text in manualroute to say "list of
1896 hosts" instead of "hostlist".
1898 2. If verify=header_syntax was set, and a header line with an unqualified
1899 address (no domain) and a large number of spaces between the end of the
1900 name and the colon was received, the reception process suffered a buffer
1901 overflow, and (when I tested it) crashed. This was caused by some obsolete
1902 code that should have been removed. The fix is to remove it!
1904 3. When running in the test harness, delay a bit after writing a bounce
1905 message to get a bit more predictability in the log output.
1907 4. Added a call to search_tidyup() just before forking a reception process. In
1908 theory, someone could use a lookup in the expansion of smtp_accept_max_
1909 per_host which, without the tidyup, could leave open a database connection.
1911 5. Added the variables $recipient_data and $sender_data which get set from a
1912 lookup success in an ACL "recipients" or "senders" condition, or a router
1913 "senders" option, similar to $domain_data and $local_part_data.
1915 6. Moved the writing of debug_print from before to after the "senders" test
1918 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1919 problems for message scanning, either using a data ACL, or using
1920 local_scan() because the Received: header was not generated till after they
1921 were called (in order to set the time as the time of reception completion).
1922 I have revised the way this works. The header is now generated after the
1923 body is received, but before the ACL or local_scan() are called. After they
1924 are run, the timestamp in the header is updated.
1930 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1931 before starting a queue runner without re-exec. This happened only when
1932 deliver_drop_privilege was set or when the Exim user was set to root. The
1933 effect of the bug was that timeouts during subsequent deliveries caused
1934 crashes instead of being properly handled. The handler is now left at its
1935 default (and expected) setting.
1937 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1938 message, again when deliver_drop_privilege is set or Exim is run as root.
1939 The bug described in (1) was not present in this case, but the tidying up
1940 of the other signals was missing. I have made the two cases consistent.
1942 3. The ignore_target_hosts setting on a manualroute router was being ignored
1943 for hosts that were looked up using the /MX notation.
1945 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1948 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1949 operated on the sender address. After changing the $sender_address to <>
1950 for the sender address verify, Exim was re-instated it as the original
1951 (before rewriting) address, but remembering that it had rewritten it, so it
1952 wasn't rewriting it again. This bug also had the effect of breaking the
1953 sender address verification caching when the sender address was rewritten.
1955 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1956 This has been changed so that if the ip literal address matches
1957 ignore_target_hosts, the router declines.
1959 7. Added expansion conditions match_domain, match_address, and match_local_
1960 part (NOT match_host).
1962 8. The placeholder for the Received: header didn't have a length field set.
1964 9. Added code to Exim itself and to exim_lock to test for a specific race
1965 condition that could lead to file corruption when using MBX delivery. The
1966 issue is with the lockfile that is created in /tmp. If this file is removed
1967 after a process has opened it but before that process has acquired a lock,
1968 there is the potential for a second process to recreate the file and also
1969 acquire a lock. This could lead to two Exim processes writing to the file
1970 at the same time. The added code performs the same test as UW imapd; it
1971 checks after acquiring the lock that its file descriptor still refers to
1972 the same named file.
1974 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1975 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1978 11. Added the smtp_active_hostname option. If used, this will typically be made
1979 to depend on the incoming interface address. Because $interface_address is
1980 not set up until the daemon has forked a reception process, error responses
1981 that can happen earlier (such as "too many connections") no longer contain
1984 12. If an expansion in a condition on a "warn" statement fails because a lookup
1985 defers, the "warn" statement is abandoned, and the next ACL statement is
1986 processed. Previously this caused the whole ACL to be aborted.
1988 13. Added the iplsearch lookup type.
1990 14. Added ident_timeout as a log selector.
1992 15. Added tls_certificate_verified as a log selector.
1994 16. Added a global option tls_require_ciphers (compare the smtp transport
1995 option of the same name). This controls incoming TLS connections.
1997 17. I finally figured out how to make tls_require_ciphers do a similar thing
1998 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1999 before starting the TLS session.
2001 18. Tabs are now shown as \t in -bP output.
2003 19. If the log selector return_path_on_delivery was set, Exim crashed when
2004 bouncing a message because it had too many Received: header lines.
2006 20. If two routers both had headers_remove settings, and the first one included
2007 a superfluous trailing colon, the final name in the first list and the
2008 first name in the second list were incorrectly joined into one item (with a
2009 colon in the middle).
2015 1. Added -C and -D options to the exinext utility, mainly to make it easier
2016 to include in the automated testing, but these could be helpful when
2017 multiple configurations are in use.
2019 2. The exinext utility was not formatting the output nicely when there was
2020 an alternate port involved in the retry record key, nor when there was a
2021 message id as well (for retries that were specific to a specific message
2022 and a specific host). It was also confused by IPv6 addresses, because of
2023 the additional colons they contain. I have fixed the IPv4 problem, and
2024 patched it up to do a reasonable job for IPv6.
2026 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2027 delivery, the log line now contains "pipelined" if PIPELINING was used.
2029 4. An SMTP transport process used to panic and die if the bind() call to set
2030 an explicit outgoing interface failed. This has been changed; it is now
2031 treated in the same way as a connect() failure.
2033 5. A reference to $sender_host_name in the part of a conditional expansion
2034 that was being skipped was still causing a DNS lookup. This no longer
2037 6. The def: expansion condition was not recognizing references to header lines
2038 that used bh_ and bheader_.
2040 7. Added the _cache feature to named lists.
2042 8. The code for checking quota_filecount in the appendfile transport was
2043 allowing one more file than it should have been.
2045 9. For compatibility with Sendmail, the command line option
2053 and sets the incoming protocol and host name (for trusted callers). The
2054 host name and its colon can be omitted when only the protocol is to be set.
2055 Note the Exim already has two private options, -pd and -ps, that refer to
2056 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2057 "s", but I don't think that's a major issue.
2059 10. A number of refactoring changes to the code, none of which should affect
2062 (a) The number of logging options was getting close to filling up the
2063 32-bit word that was used as a bit map. I have split them into two classes:
2064 those that are passed in the argument to log_write(), and those that are
2065 only ever tested independently outside of that function. These are now in
2066 separate 32-bit words, so there is plenty of room for expansion again.
2067 There is no change in the user interface or the logging behaviour.
2069 (b) When building, for example, log lines, the code previously used a
2070 macro that called string_cat() twice, in order to add two strings. This is
2071 not really sufficiently general. Furthermore, there was one instance where
2072 it was actually wrong because one of the argument was used twice, and in
2073 one call a function was used. (As it happened, calling the function twice
2074 did not affect the overall behaviour.) The macro has been replaced by a
2075 function that can join an arbitrary number of extra strings onto a growing
2078 (c) The code for expansion conditions now uses a table and a binary chop
2079 instead of a serial search (which was left over from when there were very
2080 few conditions). Also, it now recognizes conditions like "pam" even when
2081 the relevant support is not compiled in: a suitably worded error message is
2082 given if an attempt is made to use such a condition.
2084 11. Added ${time_interval:xxxxx}.
2086 12. A bug was causing one of the ddress fields not to be passed back correctly
2087 from remote delivery subprocesses. The field in question was not being
2088 subsequently used, so this caused to problems in practice.
2090 13. Added new log selectors queue_time and deliver_time.
2092 14. Might have fixed a bug in maildirsizefile handling that threw up
2093 "unexpected character" debug warnings, and recalculated the data
2094 unnecessarily. In any case, I expanded the warning message to give more
2097 15. Added the message "Restricted characters in address" to the statements in
2098 the default ACL that block characters like @ and % in local parts.
2100 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2101 Three changes have been made:
2103 (a) There was a serious bug; a negative response to MAIL caused the whole
2104 recipient domain to be cached as invalid, thereby blocking all messages
2105 to all local parts at the same domain, from all senders. This bug has
2106 been fixed. The domain is no longer cached after a negative response to
2107 MAIL if the sender used is not empty.
2109 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2112 (c) A new callout option, "use_sender" has been added for people who want
2113 the modified behaviour.
2119 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2122 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2123 indeed breaks things for older releases.
2125 3. Added additional logging to the case where there is a problem reading data
2126 from a filter that is running in a subprocess using a pipe, in order to
2127 try to track down a specific problem.
2129 4. Testing facility fudge: when running in the test harness and attempting
2130 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2131 getting "No route to host". Convert this to a timeout.
2133 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2136 6. Some OS don't have socklen_t but use size_t instead. This affects the
2137 fifth argument of getsockopt() amongst other things. This is now
2138 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2139 can be set for individual OS. I have set it for SunOS5, OSF1, and
2140 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2141 some earlier ones do not.
2143 7. Change 4.30/15 was not doing the test caselessly.
2145 8. The standard form for an IPv6 address literal was being rejected by address
2146 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2147 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2148 this, as well as the form without the "IPv6" on the front (but only when
2149 address literals are enabled, of course).
2151 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2153 10. Exim crashed if a message with an empty sender address specified by -f
2154 encountered a router with an errors_to setting. This could be provoked only
2155 by a command such as
2159 where an empty string was supplied; "<>" did not hit this bug.
2161 11. Installed PCRE release 4.5.
2163 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2164 remained set. It is now erased.
2166 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2167 times from message ids (which are base 36 rather than the normal 62).
2169 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2170 were being counted as actual protocol errors, and logged if the log
2171 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2172 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2173 and DATA following a set of rejected RCPTs do not count as protocol errors.
2174 In other words, Exim assumes they were pipelined, though this may not
2175 actually be the case. Of course, in all cases the client gets an
2176 appropriate error code.
2178 15. If a lookup fails in an ACL condition, a message about the failure may
2179 be available; it is used if testing the ACL cannot continue, because most
2180 such messages specify what the cause of the deferral is. However, some
2181 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2182 that caused an old message to be retained and used if a later statement
2183 caused a defer, replacing the real cause of the deferral.
2185 16. If an IP address had so many PTR records that the DNS lookup buffer
2186 was not large enough to hold them, Exim could crash while trying to process
2187 the truncated data. It now detects and logs this case.
2189 17. Further to 4.21/58, another change has been made: if (and only if) the
2190 first line of a message (the first header line) ends with CRLF, a bare LF
2191 in a subsequent header line has a space inserted after it, so as not to
2192 terminate the header.
2194 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2195 unnecessarily, used atoi() instead of strtol(), and didn't check the
2196 termination when getting file sizes from file names by regex.
2198 19. Completely re-implemented the support for maildirsize files, in the light
2199 of a number of problems with the previous contributed implementation
2200 (4.30/29). In particular:
2202 . If the quota is zero, the maildirsize file is maintained, but no quota is
2205 . If the maildir directory does not exist, it is created before any attempt
2206 to write a maildirsize file.
2208 . The quota value in the file is just a cache; if the quota is changed in
2209 the transport, the new value overrides.
2211 . A regular expression is available for excluding directories from the
2214 20. The autoreply transport checks the characters in options that define the
2215 message's headers; it allows continued headers, but it was checking with
2216 isspace() after an embedded newline instead of explicitly looking for a
2219 21. If all the "regular" hosts to which an address was routed had passed their
2220 expiry times, and had not reached their retry times, the address was
2221 bounced, even if fallback hosts were defined. Now Exim should go on to try
2224 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2225 equivalent code in the SMTP transport. Some hosts send humungous responses
2226 to HELO/EHLO, more than 1024 it seems.
2228 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2229 gives compiler warnings in some environments. I've now done it "properly",
2232 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2233 (because of problems with the built-in one) was declared to return uschar *
2234 instead of char *, causing compiler failure.
2236 25. Fixed a file descriptor leak when processing alias/forward files.
2238 26. Fixed a minor format string issue in dbfn.c.
2240 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2242 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2243 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2246 29. When a sender address is verified, it is cached, to save repeating the test
2247 when there is more than one recipient in a message. However, when the
2248 verification involves a callout, it is possible for different callout
2249 options to be set for different recipients. It is too complicated to keep
2250 track of this in the cache, so now Exim always runs a verification when a
2251 callout is required, relying on the callout cache for the optimization.
2252 The overhead is duplication of the address routing, but this should not be
2255 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2256 to be verified with callout=postmaster, and the main callout worked but the
2257 postmaster check failed, the verification correctly failed. However, if a
2258 subsequent RCPT command asked for sender verification *without* the
2259 postmaster check, incorrect caching caused this verification also to fail,
2262 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2263 it was not caching the DNS options (qualify_single, search_parents) that
2264 were used when the lookup failed. A subsequent lookup with different
2265 options therefore always gave the same answer, though there were cases
2266 where it should not have. (Example: a "domains = !$mx_any" option on a
2267 dnslookup router: the "domains" option is always processed without any
2268 widening, but the router might have qualify_single set.) Now Exim uses the
2269 cached value only when the same options are set.
2271 32. Added John Jetmore's "exipick" utility to the distribution.
2273 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2274 than a timeout (e.g. a certificate is required, and is not provided), an
2275 Exim server now closes the connection immediately. Previously it waited for
2276 the client to close - but if the client is SSL, it seems that they each
2277 wait for each other, leading to a delay before one of them times out.
2279 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2280 maintained 0.8.x compatibility because I don't think many are using it, and
2281 it is clearly obsolete.
2283 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2286 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2287 client certificate was expired. A simple patch fixes this, though I don't
2288 understand the full logic of why the verify callback is called multiple
2291 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2292 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2293 which causes problems with some clients (such as the Certicom SSL Plus
2294 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2295 disables the coutermeasure allowing Eudora to connect."
2297 38. Exim was not checking that a write() to a log file succeeded. This could
2298 lead to Bad Things if a log got too big, in particular if it hit a file
2299 size limit. Exim now panics and dies if it cannot write to a log file, just
2300 as it does if it cannot open a log file.
2302 39. Modified OS/Makefile-Linux so that it now contains
2304 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2306 The two -D definitions ensure that Exim is compiled with large file
2307 support, which makes it possible to handle log files that are bigger than
2310 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2311 instance) a domain was checked against a named list that involved a lookup,
2312 causing $domain_data to be set, then another domain was checked against the
2313 same list, then the first domain was re-checked, the value of $domain_data
2314 after the final check could be wrong. In particular, if the second check
2315 failed, it could be set empty. This bug probably also applied to
2318 41. The strip_trailing_dot option was not being applied to the address given
2319 with the -f command-line option.
2321 42. The code for reading a message's header from the spool was incrementing
2322 $received_count, but never initializing it. This meant that the value was
2323 incorrect (doubled) while delivering a message in the same process in which
2324 it was received. In the most common configuration of Exim, this never
2325 happens - a fresh exec is done - but it can happen when
2326 deliver_drop_privilege is set.
2328 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2329 it now includes up to 150 characters of the unexpected data in the log
2332 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2333 and building data strings. The size of both of these buffers was 10 000
2334 bytes - far larger than anybody would *ever* want, thought I. Needless to
2335 say, somebody hit the limit. I have increased the maximum line length to
2336 20 000 and the maximum data length of concatenated lines to 100 000. I have
2337 also fixed two bugs, because there was no checking on these buffers. Tsk,
2338 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2339 buffer is too small.
2341 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2342 lsearch lookups. Now it does.
2344 46. When parsing a route_list item in a manualroute router, a fixed-length
2345 buffer was used for the list of hosts. I made this 1024 bytes long,
2346 thinking that nobody would ever have a list of hosts that long. Wrong.
2347 Somebody had a whole pile of complicated expansion conditions, and the
2348 string was silently truncated, leading to an expansion error. It turns out
2349 that it is easier to change to an unlimited length (owing to other changes
2350 that have happened since this code was originally written) than to build
2351 structure for giving a limitation error. The length of the item that
2352 expands into the list of hosts is now unlimited.
2354 47. The lsearch lookup could not handle data where the length of text line was
2355 more than 4095 characters. Such lines were truncated, leading to shortened
2356 data being returned. It should now handle lines of any length.
2358 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2359 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2362 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2363 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2364 no longer happens in comment lines. A list of replacements is now placed
2365 at the head of all of the source files, except those whose only change is
2366 to replace PERL_COMMAND in the very first #! line.
2368 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2369 messages on the queue, with a bottom-up merge sort, using code contributed
2370 by Michael Haardt. This should make operations like -bp somewhat faster on
2371 large queues. It won't affect queue runners, except when queue_run_in_order
2374 51. Installed eximstats 1.31 in the distribution.
2376 52. Added support for SRV lookups to the dnslookup router.
2378 53. If an ACL referred to $message_body or $message_body_end, the value was not
2379 reset for any messages that followed in the same SMTP session.
2381 54. The store-handling optimization for building very long strings was not
2382 differentiating between the different store pools. I don't think this
2383 actually made any difference in practice, but I've tidied it.
2385 55. While running the routers to verify a sender address, $sender_address
2386 was still set to the sender address. This is wrong, because when routing to
2387 send a bounce to the sender, it would be empty. Therefore, I have changed
2388 it so that, while verifying a sender address, $sender_address is set to <>.
2389 (There is no change to what happens when verifying a recipient address.)
2391 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2392 target A or AAAA records (if not already returned) without resetting the
2393 qualify_single or search_parents options of the DNS resolver. These are
2394 inappropriate in this case because the targets of MX and SRV records must
2395 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2396 target that, when qualified, matched something in the local domain. These
2397 two options are now turned off when doing these lookups.
2399 57. It seems that at least some releases of Reiserfs (which does not have the
2400 concept of a fixed number of inodes) returns zero and not -1 for the
2401 number of available inodes. This interacted badly with check_spool_inodes,
2402 which assumed that -1 was the "no such thing" setting. What I have done is
2403 to check that the total number of inodes is greater than zero before doing
2404 the test of how many are available.
2406 58. When a "warn" ACL statement has a log_message modifier, the message is
2407 remembered, and not repeated. This is to avoid a lot of repetition when a
2408 message has many recipients that cause the same warning to be written.
2409 Howewer, Exim was preserving the list of already written lines for an
2410 entire SMTP session, which doesn't seem right. The memory is now reset if a
2411 new message is started.
2413 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2414 debugging output unless log_rewrite was also set.
2416 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2417 of (int)(handle) when we know that handle contains (void *)(-1).
2419 61. The Exim daemon panic-logs an error return when it closes the incoming
2420 connection. However "connection reset by peer" seems to be common, and
2421 isn't really an error worthy of noting specially, so that particular error
2424 62. When Exim is trying to find all the local interfaces, it used to panic and
2425 die if the ioctl to get the interface flags failed. However, it seems that
2426 on at least one OS (Solaris 9) it is possible to have an interface that is
2427 included in the list of interfaces, but for which you get a failure error
2428 for this call. This happens when the interface is not "plumbed" into a
2429 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2430 failure of the "get flags" call assumes that the interface is down.
2432 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2433 makes life easier for people who are doing arithmetic on fields extracted
2434 from dates, where you often get leading zeros that should not be
2435 interpreted as octal.
2437 64. Added qualify_domain to the redirect router, to override the global
2440 65. If a pathologically long header line contained very many addresses (the
2441 report of this problem mentioned 10 000) and each of them was rewritten,
2442 Exim could use up a very large amount of memory. (It kept on making new
2443 copies of the header line as it rewrote, and never released the old ones.)
2444 At the expense of a bit more processing, the header rewriting function has
2445 been changed so that it no longer eats memory in this way.
2447 66. The generation of the Received: header has been moved from the time that a
2448 message starts to be received, to the time that it finishes. The timestamp
2449 in the Received: header should now be very close to that of the <= log
2450 line. There are two side-effects of this change:
2452 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2453 logged header lines no longer include the local Received: line, because
2454 it has not yet been created. The same applies to a copy of the message
2455 that is returned to a non-SMTP sender when a message is rejected.
2457 (b) When a filter file is tested using -bf, no additional Received: header
2458 is added to the test message. After some thought, I decided that this
2461 This change does not affect the value of $received_for. It is still set
2462 after address rewriting, but before local_scan() is called.
2464 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2466 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2467 gave an unhelpful panic error message, and a defer error. I have managed to
2468 change this behaviour so that it now rejects any supplied certificate,
2469 which seems right, as the list of acceptable certificates is empty.
2471 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2472 gave an unhelpful defer error. I have not managed to make this reject any
2473 supplied certificates, but the error message it gives is "no certificate
2474 supplied", which is not helpful.
2476 70. exigrep's output now also includes lines that are not associated with any
2477 message, but which match the given pattern. Implemented by a patch from
2478 Martin Sluka, which also tidied up the Perl a bit.
2480 71. Recipient callout verification, like sender verification, was using <> in
2481 the MAIL FROM command. This isn't really the right thing, since the actual
2482 sender may affect whether the remote host accepts the recipient or not. I
2483 have changed it to use the actual sender in the callout; this means that
2484 the cache record is now keyed on a recipient/sender pair, not just the
2485 recipient address. There doesn't seem to be a real danger of callout loops,
2486 since a callout by the remote host to check the sender would use <>.
2487 [SEE ABOVE: changed after hitting problems.]
2489 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2490 temporary errors. However, in the case of such a code being given after
2491 the end of a data transmission (i.e. after ".") Exim was failing to write
2492 a retry record for the message. (Yes, there was some broken host that was
2493 actually sending 8xx at this point.)
2495 73. An unknown lookup type in a host list could cause Exim to panic-die when
2496 the list was checked. (An example that provoked this was putting <; in the
2497 middle of a list instead of at the start.) If this happened during a DATA
2498 ACL check, a -D file could be left lying around. This kind of configuration
2499 error no longer causes Exim to die; instead it causes a defer errror. The
2500 incident is still logged to the main and panic logs.
2502 74. Buglet left over from Exim 3 conversion. The message "too many messages
2503 in one connection" was written to the rejectlog but not the mainlog, except
2504 when address rewriting (yes!) was being logged.
2506 75. Added write_rejectlog option.
2508 76. When a system filter was run not as root (that is, when system_filter_user
2509 was set), the values of the $n variables were not being returned to the
2510 main process; thus, they were not subsequently available in the $sn
2513 77. Added +return_path_on_delivery log selector.
2515 78. A connection timeout was being treated differently from recipients deferred
2516 when testing hosts_max_try with a message that was older than the host's
2517 retry timeout. (The host should not be counted, thus allowing all hosts to
2518 be tried at least once before bouncing.) This may have been the cause of an
2519 occasionally reported bug whereby a message would remain on the queue
2520 longer than the retry timeout, but would be bounced if a delivery was
2521 forced. I say "may" because I never totally pinned down the problem;
2522 setting up timeout/retry tests is difficult. See also the next item.
2524 79. The ultimate address timeout was not being applied to errors that involved
2525 a combination of host plus message (for example, a timeout on a MAIL
2526 command). When an address resolved to a number of possible hosts, and they
2527 were not all tried for each delivery (e.g. because of hosts_max_try), a
2528 message could remain on the queue longer than the retry timeout.
2530 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2533 81. Fixed an obscure SMTP outgoing bug which required at least the following
2534 conditions: (a) there was another message waiting for the same server;
2535 (b) the server returned 5xx to all RCPT commands in the first message so
2536 that the message was not completed; (c) the server dropped the connection
2537 or gave a negative response to the RSET that Exim sends to abort the
2538 transaction. The observed case was a dropped connection after DATA that had
2539 been sent in pipelining mode. That is, the server had advertised PIPELINING
2540 but was not implementing it correctly. The effect of the bug was incorrect
2541 behaviour, such as trying another host, and this could lead to a crash.
2547 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2548 and daemon.c were passed as pointers to ints; they should have been
2549 pointers to socklen_t variables (which are typically unsigned ints).
2551 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2554 3. Fixed a really odd bug that affected only the testing scheme; patching a
2555 certain fixed string in the binary changed the value of another string that
2556 happened to be identical to the end of the original first string.
2558 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2559 name", it returns that address as the IP address. On some operating
2560 systems (e.g. Solaris), it also passes back the IP address string as the
2561 "host name". However, on others (e.g. Linux), it passes back an empty
2562 string. Exim wasn't checking for this, and was changing the host name to an
2563 empty string, assuming it had been canonicized.
2565 5. Although rare, it is permitted to have more than one PTR record for a given
2566 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2567 all the names associated with an address, because they do in Solaris.
2568 However, it seems that they do not in Linux for data that comes from the
2569 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2570 I found this out when I moved to a new Linux workstation and tried to run
2571 the Exim test suite.
2573 To get round this problem I have changed the code so that it now does its
2574 own call to the DNS to look up PTR records when searching for a host name.
2575 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2576 addresses that are only in /etc/hosts are still found.
2578 This behaviour is, however, controlled by an option called host_lookup_
2579 order, which defaults to "bydns:byaddr". If people want to use the other
2580 order, or indeed, just use one or the other means of lookup, they can
2581 specify it in this variable.
2583 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2584 some operating systems, this comes back from gethostbyaddr() as an empty
2585 string, and this is what Exim used to test for. However, it seems that in
2586 other systems, "." is yielded. Exim now tests for this case too.
2588 7. The values of check_spool_space and check_log_space are now held internally
2589 as a number of kilobytes instead of an absolute number of bytes. If a
2590 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2591 kilobyte. This means that much larger values can be stored.
2593 8. Exim monitor: an attempt to get the action menu when not actually pointing
2594 at a message produces an empty menu entitled "No message selected". This
2595 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2596 no entries in it ("Shell widget menu has zero width and/or height"). So I
2597 have added a single, blank menu entry in this case.
2599 9. Added ${quote_local_part.
2601 10. MIME decoding is now applied to the contents of Subject: header lines when
2604 11. Now that a reference to $sender_host_address automatically causes a reverse
2605 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2606 host lookup before query-style lookups in lists that might use this
2607 variable. This has therefore been abolished, and the "net-" prefix is no
2608 longer necessary for query-style lookups.
2610 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2611 have been a typo for LFLAGS, so it has been changed.
2613 13. The install script calls Exim with "-C /dev/null" in order to find the
2614 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2615 to be output. Howeve, since Exim outputs its version number before the
2616 error, it didn't break the script. It just looked ugly. I fixed this by
2617 always allowing "-C /dev/null" if the caller is root.
2619 14. Ignore overlarge ACL variable number when reading spool file - insurance
2620 against a later release with more variables having written the file.
2622 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2623 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2624 this, as well as the form without the "IPv6" on the front.
2626 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2627 OS/Makefile-Darwin file.
2629 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2630 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2632 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2633 variable. However, some OS use ints for inodes. I've added cast to long int
2634 to get rid of the compiler warning.
2636 19. I had forgotten to lock out "/../" in configuration file names when
2637 ALT_CONFIG_PREFIX was set.
2639 20. Routers used for verification do not need to specify transports. However,
2640 if such a router generated a host list, and callout was configured, Exim
2641 crashed, because it could not find a port number from the (non-existent)
2642 transport. It now assumes port 25 in this circumstance.
2644 21. Added the -t option to exigrep.
2646 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2647 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2648 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2649 them is set, LOOKUP_LSEARCH is forced.
2651 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2653 24. Added sender and host information to the "rejected by local_scan()" log
2654 line; previously there was no indication of these.
2656 25. Added .include_if_exists.
2658 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2659 Linux. It turns out that not all file systems support this. Apparently some
2660 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2661 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2662 not supported on the file descriptor, is now ignored when Exim is trying to
2663 sync a directory. This applies only to Linux.
2665 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2667 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2668 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2671 29. Added support for maildirsize files from supplied patch (modified a bit).
2673 30. The use of :fail: followed by an empty string could lead Exim to respond to
2674 sender verification failures with (e.g.):
2676 550 Verification failed for <xxx>
2677 550 Sender verify failed
2679 where the first response line was missing the '-' that indicates it is not
2680 the final line of the response.
2682 31. The loop for finding the name of the user that called Exim had a hardwired
2683 limit of 10; it now uses the value of finduser_retries, which is used for
2684 all other user lookups.
2686 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2689 33. Exim was neglecting to zero errno before one call of strtol() when
2690 expanding a string and expecting an integer value. On some systems this
2691 resulted in spurious "integer overflow" errors. Also, it was casting the
2692 result into an int without checking.
2694 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2695 did not work. The code looks as if it has *never* worked, though it appears
2696 to have been documented since at least releast 1.62. I have made it work.
2698 35. The "timeout_DNS" error in retry rules, also documented since at least
2699 1.62, also never worked. As it isn't clear exactly what this means, and
2700 clearly it isn't a major issue, I have abolished the feature by treating it
2701 as "timeout", and writing a warning to the main and panic logs.
2703 36. The display of retry rules for -brt wasn't always showing the error code
2706 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2707 timeout_connect_A, timeout_connect_MX.
2709 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2710 to the empty sender.
2712 39. The daemon was not analysing the content of -oX till after it had closed
2713 stderr and disconnected from the controlling terminal. This meant that any
2714 syntax errors were only noted on the panic log, and the return code from
2715 the command was 0. By re-arranging the code a little, I've made the
2716 decoding happen first, so such errors now appear on stderr, and the return
2717 code is 1. However, the actual setting up of the sockets still happens in
2718 the disconnected process, so errors there are still only recorded on the
2721 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2722 connections (as happens on some IP stacks) was logged at start up time as
2723 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2724 it from "IPv6 and IPv4", which means that two separate sockets are being
2727 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2728 says whether AF_INET or AF_INET6 was passed as an argument.
2730 42. Exiwhat output was messed up when time zones were included in log
2733 43. Exiwhat now gives more information about the daemon's listening ports,
2734 and whether -tls-on-connect was used.
2736 44. The "port" option of the smtp transport is now expanded.
2738 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2739 silently ignored. Now an error message is written to the main and panic
2742 46. There's a new ACL modifier called "logwrite" which writes to a log file
2743 as soon as it is encountered.
2745 47. Added $local_user_uid and $local_user_gid at routing time.
2747 48. Exim crashed when trying to verify a sender address that was being
2750 49. Exim was recognizing only a space character after ".include". It now also
2751 recognizes a tab character.
2753 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2754 extracting the relevant information from the specification. The man page no
2755 longer contains scrambled data for the -d option, and I've added a section
2756 at the front about calling Exim under different names.
2758 51. Added "extra_headers" argument to the "mail" command in filter files.
2760 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2763 53. Installed eximstats 1.29.
2765 54. Added transport_filter_timeout as a generic transport option.
2767 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2768 Cc: header lines. This was required by RFC 822, but it not required by RFC
2771 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2772 incoming messages that did not have them. Now it does so only if the
2773 message originates locally, that is, if there is no associated remote host
2774 address. When Resent- header lines are present, this applies to the Resent-
2775 lines rather than the non-Resent- lines.
2777 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2778 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2780 58. Messages for configuration errors now include the name of the main
2781 configuration file - useful now that there may be more than one file in a
2782 list (.included file names were always shown).
2784 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2785 for those rare installations that do not start the daemon as root or run it
2786 setuid root. I've cut out the call to initgroups() if the daemon is not
2789 60. The Exim user and group can now be bound into the binary as text strings
2790 that are looked up at the start of Exim's processing.
2792 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2794 62. Added $mailstore_basename variable.
2796 63. Installed patch to sieve.c from Michael Haardt.
2798 64. When Exim failed to open the panic log after failing to open the main log,
2799 the original message it was trying to log was written to stderr and debug
2800 output, but if they were not available (the usual case in production), it
2801 was lost. Now it is written to syslog before the two lines that record the
2802 failures to open the logs.
2804 65. Users' Exim filters run in subprocesses under the user's uid. It is
2805 possible for a "deliver" command or an alias in a "personal" command to
2806 provoke an address rewrite. If logging of address rewriting is configured,
2807 this fails because the process is not running as root or exim. There may be
2808 a better way of dealing with this, but for the moment (because 4.30 needs
2809 to be released), I have disabled address rewrite logging when running a
2810 filter in a non-root, non-exim process.
2816 1. The buildconfig auxiliary program wasn't quoting the value set for
2817 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2818 not defined. This bug was masked in 4.22 by the effect that was fixed in
2821 2. Some messages that were rejected after a message id was allocated were
2822 shown as "incomplete" by exigrep. It no longer does this for messages that
2823 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2825 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2826 have allow_domain_literals set, the ID did not get logged in the <= line.
2827 Domain literals are now always recognized in Message-ID: header lines.
2829 4. The first argument for a ${extract expansion item is the key name or field
2830 number. Leading and trailing spaces in this item were not being ignored,
2831 causing some misleading effects.
2833 5. When deliver_drop_privilege was set, single queue runner processes started
2834 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2835 same command in the process it spins off) were not dropping privilege.
2837 6. When the daemon running as "exim" started a queue runner, it always
2838 re-executed Exim in the spun-off process. This is a waste of effort when
2839 deliver_drop_privilege is set. The new process now just calls the
2840 queue-runner function directly.
2846 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2849 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2850 ignored. Though the use of -oP was forcing the writing of a pid file, it
2851 was always written to the default place.
2853 3. If the message "no IP address found for host xxxx" is generated during
2854 incoming verification, it is now followed by identification of the incoming
2855 connection (so you can more easily find what provoked it).
2857 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2859 5. Added some features to "harden" Exim a bit more against certain attacks:
2861 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2862 be put in Local/Makefile. This is like the never_users runtime option,
2863 but it cannot be overridden. The default setting is "root".
2865 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2866 prefix string with which any file named in a -C command line option
2869 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2870 is retained for -C and -D only if the caller of Exim is root. Without
2871 it, the exim user may also use -C and -D and retain privilege.
2873 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2874 command line option is disabled.
2876 6. Macro names set by the -D option must start with an upper case letter, just
2877 like macro names defined in the configuration file.
2879 7. Added "dereference=" facility to LDAP.
2881 8. Two instances of the typo "uknown" in the source files are fixed.
2883 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2884 the Configure-Makefile script screwed up while processing it.
2886 10. Incorporated PCRE 4.4.
2888 11. The SMTP synchronization check was not operating right at the start of an
2889 SMTP session. For example, it could not catch a HELO sent before the client
2890 waited for the greeting. There is now a check for outstanding input at the
2891 point when the greeting is written. Because of the duplex, asynchronous
2892 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2893 way, but not yet received, when the check is performed.
2895 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2896 on TCP/IP sockets, because this apparently causes some broken clients to
2899 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2900 unchanged) from the Cygwin maintainer.
2902 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2903 was supported instead of testing for mbx. Effectively a typo.
2905 15. The spa authenticator server code was not checking that the input it
2906 received was valid base64.
2908 16. The debug output line for the "set" modifier in ACLs was not showing the
2909 name of the variable that was being set.
2911 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2913 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2914 lookup on its own. Something else was needed to trigger the lookup. For
2915 example, a match in host_lookup or the need for a host name in a host list.
2916 Now, if $sender_host_name is referenced and the host name has not yet been
2917 looked up, a lookup is performed. If the lookup fails, the variable remains
2918 empty, and $host_lookup_failed is set to "1".
2920 19. Added "eqi" as a case-independent comparison operator.
2922 20. The saslauthd authentication condition could segfault if neither service
2923 nor realm was specified.
2925 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2926 error message that was logged was misleading, and incoming SMTP
2927 connections were dropped. The message is now more accurate, and temporary
2928 errors are given to SMTP connections.
2930 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2931 (except RSET) with a 421 error, until QUIT is received. However, it was
2932 failing to send a response to QUIT.
2934 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2935 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2936 ACL is now run after the helo_try_verify_hosts code.
2938 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2939 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2940 case-independent, so other case variants are also recognized). Apparently
2941 some systems use these upper case variants.
2943 25. If more than two messages were waiting for the same host, and a transport
2944 filter was specified for the transport, Exim sent two messages over the
2945 same TCP/IP connection, and then failed with "socket operation on non-
2946 socket" when it tried to send the third.
2948 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2950 27. The extern definition of crypt16() in expand.c was not being excluded when
2951 the OS had its own crypt16() function.
2953 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2954 as a preferred synonym for return_size_limit, both as an option and as an
2957 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2959 30. Changed the default configuration ACL to relax the local part checking rule
2960 for addresses that are not in any local domains. For these addresses,
2961 slashes and pipe symbols are allowed within local parts, but the sequence
2962 /../ is explicitly forbidden.
2964 31. SPA server authentication was not clearing the challenge buffer before
2967 32. log_message in a "warn" ACL statement was writing to the reject log as
2968 well as to the main log, which contradicts the documentation and doesn't
2969 seem right (because no rejection is happening). So I have stopped it.
2971 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2972 However, I am unable to do any testing of this.
2974 34. Fixed an infelicity in the appendfile transport. When checking directories
2975 for a mailbox, to see if any needed to be created, it was accidentally
2976 using path names with one or more superfluous leading slashes; tracing
2977 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2979 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2980 contents are added to the log line that is written for every discarded
2981 recipient. (Previously a log_message setting was ignored.)
2983 36. The ${quote: operator now quotes the string if it is empty.
2985 37. The install script runs exim in order to find its version number. If for
2986 some reason other than non-existence or emptiness, which it checks, it
2987 could not run './exim', it was installing it with an empty version number,
2988 i.e. as "exim-". This error state is now caught, and the installation is
2991 38. An argument was missing from the function that creates an error message
2992 when Exim fails to connect to the socket for saslauthd authentication.
2993 This could cause Exim to crash, or give a corrupted message.
2995 39. Added isip, isip4, and isip6 to ${if conditions.
2997 40. The ACL variables $acl_xx are now saved with the message, and can be
2998 accessed later in routers, transports, and filters.
3000 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3001 strings in the file are not string-expanded.
3003 42. If a MAIL command specified a SIZE value that was too large to fit into an
3004 int variable, the check against message_size_limit failed. Such values are
3005 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3006 day this will have to be increased, but I don't think I want to be around
3007 when emails are that large.
3014 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3015 iconv() is not standard in FreeBSD.
3017 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3018 IPv6 enabled. The observed symptom was a segmentation fault on return from
3019 the function os_common_find_running_interfaces() in src/os.c.
3021 3. In the check_special_case() function in daemon.c I had used "errno" as an
3022 argument name, which causes warnings on some systems. This was basically a
3023 typo, since it was named "eno" in the comments!
3025 4. The code that waits for the clock to tick (at a resolution of some fraction
3026 of a second) so as to ensure message-id uniqueness was always waiting for
3027 at least one whole tick, when it could have waited for less. [This is
3028 almost certainly not relevant at current processor speeds, where it is
3029 unlikely to ever wait at all. But we try to future-proof.]
3031 5. The function that sleeps for a time interval that includes fractions of a
3032 second contained a race. It did not block SIGALRM between setting the
3033 timer, and suspending (a couple of lines later). If the interval was short
3034 and the sigsuspend() was delayed until after it had expired, the suspension
3035 never ended. On busy systems this could lead to processes getting stuck for
3038 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3039 process, before it forks any delivery processes. The open lookup caching
3040 mechanism meant that the open file or database connection was passed into
3041 the delivery process. The problem was that delivery processes always tidy
3042 up cached lookup data. This could cause a problem for the next delivery
3043 process started by the queue runner, because the external queue runner
3044 process does not know about the closure. So the next delivery process
3045 still has data in the lookup cache. In the case of a file lookup, there was
3046 no problem because closing a file descriptor in a subprocess doesn't affect
3047 the parent. However, if the lookup was caching a connection to a database,
3048 the connection was closed, and the second delivery process was likely to
3049 see errors such as "PGSQL: query failed: server closed the connection
3050 unexpectedly". The problem has been fixed by closing all cached lookups
3051 in a queue runner before running a delivery process.
3053 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3054 seem to have the "const" qualifier which it has on other OS. I've
3057 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3058 *of the same type* (client or server) with the same public name that an
3059 error should be diagnosed.
3061 9. When Exim looked up a host name for an IP address, but failed to find the
3062 original IP address when looking up the host name (a safety check), it
3063 output the message "<ip address> does not match any IP for NULL", which was
3064 confusing, to say the least. The bug was that the host name should have
3065 appeared instead of "NULL".
3067 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3068 user that is built into the binary, and the -C or -D options is used, root
3069 privilege is dropped before the configuration file is read. In addition,
3070 logging is switched to stderr instead of the normal log files. If the
3071 configuration then re-defines the Exim user, the unprivileged environment
3072 is probably not what is expected, so Exim logs a panic warning message (but
3075 However, if deliver_drop_privilege is set, the unprivileged state may well
3076 be exactly what is intended, so the warning has been cut out in that case,
3077 and Exim is allowed to try to write to its normal log files.
3083 1. smtp_return_error_details was not giving details for temporary sender
3084 or receiver verification errors.
3086 2. Diagnose a configuration error if two authenticators have the same public
3089 3. Exim used not to create the message log file for a message until the first
3090 delivery attempt. This could be confusing when incoming messages were held
3091 for policy or load reasons. The message log file is now created at the time
3092 the message is received, and an initial "Received" line is written to it.
3094 4. The automatically generated man page for command line options had a minor
3095 bug that caused no ill effects; however, a more serious problem was that
3096 the procedure for building the man page automatically didn't always
3097 operate. Consequently, release 4.20 contains an out-of-date version. This
3098 shouldn't happen again.
3100 5. When building Exim with embedded Perl support, the script that builds the
3101 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3102 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3104 6. The freeze_tell option was not being used for messages that were frozen on
3105 arrival, either by an ACL or by local_scan().
3107 7. Added the smtp_incomplete_transaction log selector.
3109 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3110 was accepting AUTH without a new EHLO.
3112 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3113 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3116 10. Logging of TCP/IP connections (when configured) now happens in the main
3117 daemon process instead of the child process, so that the TCP/IP connection
3118 count is more accurate (but it can never be perfect).
3120 11. The use of "drop" in a nested ACL was not being handled correctly in the
3121 outer ACL. Now, if condition failure induced by the nested "drop" causes
3122 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3123 or "require"), the connection is dropped.
3125 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3126 that yield "discard" can now be used with an "accept" or a "discard" verb,
3127 but an error is generated for any others (because I can't see a useful way
3128 to define what should happen).
3130 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3131 are now processed in the same way as lines in the Exim configuration file.
3132 In particular, continuation lines are supported.
3134 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3136 15. Added -ti meaning -t -i.
3138 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3139 domains, and warn by logging if others are found.
3141 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3142 returned from GIFCONF when Exim is trying to find the list of interfaces on
3143 a host. The code in os.c has been modified to copy each ifreq to an aligned
3144 structure in all cases.
3146 Also, in some cases, the returned ifreq's were being copied to a 'struct
3147 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3148 means the last couple of bytes of an IPv6 address could be chopped if the
3149 ifreq contained only a normal sockaddr (14 bytes storage).
3151 18. Named domain lists were not supported in the hosts_treat_as_local option.
3152 An entry such as +xxxx was not recognized, and was treated as a literal
3155 19. Ensure that header lines added by a DATA ACL are included in the reject log
3156 if the ACL subsequently rejects the message.
3158 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3159 MD5 (which is deprecated).
3161 21. When testing a filter file using -bf, Exim was writing a message when it
3162 took the sender from a "From " line in the message, but it was not doing so
3163 when it took $return_path from a Return-Path: header line. It now does.
3165 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3166 with a valid header line field name (a series of printing characters
3167 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3169 23. Changed "disc" in the source to "disk" to conform to the documentation and
3170 the book and for uniformity.
3172 24. Ignore Sendmail's -Ooption=value command line item.
3174 25. When execve() failed while trying to run a command in a pipe transport,
3175 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3176 could be confused with a return value of 69 from the command itself. This
3177 has been changed to 127, the value the shell returns if it is asked to run
3178 a non-existent command. The wording for the related log line suggests a
3179 non-existent command as the problem.
3181 26. If received_header_text expands to an empty string, do not add a Received:
3182 header line to the message. (Well, it adds a token one on the spool, but
3183 marks it "old" so that it doesn't get used or transmitted.)
3185 27. Installed eximstats 1.28 (addition of -nt option).
3187 28. There was no check for failure on the call to getsockname() in the daemon
3188 code. This can fail if there is a shortage of resources on the system, with
3189 ENOMEM, for example. A temporary error is now given on failure.
3191 29. Contrary to the C standard, it seems that in some environments, the
3192 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3193 program. Exim now does this explicitly; it affects the formatting of
3194 timestamps using strftime().
3196 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3197 complaints. I've now initialized all the variables, to avoid this.
3199 32. Header lines added by a system filter were not being "seen" during
3200 transport-time rewrites.
3202 33. The info_callback() function passed to OpenSSL is set up with type void
3203 (*)(SSL *, int, int), as described somewhere. However, when calling the
3204 function (actually a macro) that sets it up, the type void(*)() is
3205 expected. I've put in a cast to prevent warnings from picky compilers.
3207 34. If a DNS black list lookup found a CNAME record, but there were no A
3208 records associated with the domain it pointed at, Exim crashed.
3210 35. If a DNS black list lookup returned more than one A record, Exim ignored
3211 all but the first. It now scans all returned addresses if a particular IP
3212 value is being sought. In this situation, the contents of the
3213 $dnslist_value variable are a list of all the addresses, separated by a
3216 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3217 to accept a host name and all its aliases if the forward lookup for any of
3218 them yielded the IP address of the incoming connection. Now it accepts only
3219 those names whose forward lookup yields the correct IP address. Any other
3220 names are discarded. This closes a loophole whereby a rogue DNS
3221 administrator could create reverse DNS records to break through a
3222 wildcarded host restriction in an ACL.
3224 37. If a user filter or a system filter that ran in a subprocess used any of
3225 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3226 the wrong values were passed to the pipe command ($thisaddress had the
3227 value of $0, $0 had the value of $1, etc). This bug was introduced by
3228 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3230 38. Improved the line breaking for long SMTP error messages from ACLs.
3231 Previously, if there was no break point between 40 and 75 characters, Exim
3232 left the rest of the message alone. Two changes have been made: (a) I've
3233 reduced the minimum length to 35 characters; (b) if it can't find a break
3234 point between 35 and 75 characters, it looks ahead and uses the first one
3235 that it finds. This may give the occasional overlong line, but at least the
3236 remaining text gets split now.
3238 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3239 file descriptors might be low, and that setting 1000 would always raise it.
3240 It turns out that in some environments, the limit is already over 1000 and
3241 that lowering it causes trouble. So now Exim takes care not to decrease it.
3243 40. When delivering a message, the value of $return_path is set to $sender_
3244 address at the start of routing (routers may change the value). By an
3245 oversight, this default was not being set up when an address was tested by
3246 -bt or -bv, which affected the outcome if any router or filter referred to
3249 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3250 log only when "message" or "log_message" are set. However, if one of the
3251 conditions was an address verification, or a call to a nested ACL, the
3252 messages generated by the underlying test were being passed through. This
3253 no longer happens. The underlying message is available in $acl_verify_
3254 message for both "message" and "log_message" expansions, so it can be
3255 passed through if needed.
3257 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3258 new expansion $bh_ to give the encoded byte string without charset
3259 translation. Translation happens only if iconv() is available; HAVE_ICONV
3260 indicates this at build time. HEADERS_CHARSET gives the charset to
3261 translate to; headers_charset can change it in the configuration, and
3262 "headers charset" can change it in an individual filter file.
3264 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3265 that creates RFC 2047 encoded "words" labels them as that charset instead
3266 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3267 expansion operator; (ii) when Exim creates a From: line for a local
3268 message; (iii) when a header line is rewritten to include a "phrase" part.
3270 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3271 buggy, causing it to skip the first lines of messages whose message ID
3272 ended in 'D'. This would not have bitten before Exim release 4.14, because
3273 message IDs were unlikely to end in 'D' before then. The effect was to have
3274 incorrect size information for certain domains.
3276 45. #include "config.h" was missing at the start of the crypt16.c module. This
3277 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3280 46. If there was a timeout during a "random" callout check, Exim treated it as
3281 a failure of the random address, and carried on sending RSET and the real
3282 address. If the delay was just some slowness somewhere, the response to the
3283 original RCPT would be taken as a response to RSET and so on, causing
3284 mayhem of various kinds.
3286 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3287 when I implemented it. It didn't allow for the fact that some option values
3288 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3289 the right test for positive values.
3291 48. Domain names in DNS records are case-independent. Exim always looks them up
3292 in lower case. Some resolvers return domain names in exactly the case they
3293 appear in the zone file, that is, they may contain uppercase letters. Not
3294 all resolvers do this - some return always lower case. Exim was treating a
3295 change of case by a resolver as a change of domain, similar to a widening
3296 of a domain abbreviation. This triggered its re-routing code and so it was
3297 trying to route what was effectively the same domain again. This normally
3298 caused routing to fail (because the router wouldn't handle the domain
3299 twice). Now Exim checks for this case specially, and just changes the
3300 casing of the domain that it ultimately uses when it transmits the message
3303 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3306 50. If a filter generated a file delivery with a non-absolute name (possible if
3307 no home directory exists for the router), the forbid_file option was not
3310 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3311 the existing equality matching.
3313 52. Exim was using ints instead of ino_t variables in some places where it was
3314 dealing with inode numbers.
3316 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3317 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3318 called TMPDIR, and if it finds it is different, it changes its value.
3320 54. The smtp_printf() function is now made available to local_scan() so
3321 additional output lines can be written before returning. There is also an
3322 smtp_fflush() function to enable the detection of a dropped connection.
3323 The variables smtp_input and smtp_batched_input are exported to
3326 55. Changed the default runtime configuration: the message "Unknown user"
3327 has been removed from the ACL, and instead placed on the localuser router,
3328 using the cannot_route_message feature. This means that any verification
3329 failures that generate their own messages won't get overridden. Similarly,
3330 the "Unrouteable address" message that was in the ACL for unverifiable
3331 relay addresses has also been removed.
3333 56. Added hosts_avoid_esmtp to the smtp transport.
3335 57. The exicyclog script was not checking for the esoteric option
3336 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3337 will work only if exicyclog is run under the appropriate euid.
3339 58. Following a discussion on the list, the rules by which Exim recognises line
3340 endings on incoming messages have been changed. The -dropcr and drop_cr
3341 options are now no-ops, retained only for backwards compatibility. The
3342 following line terminators are recognized: LF CRLF CR. However, special
3343 processing applies to CR:
3345 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3346 nor a local message in the state where . is a terminator.
3348 (ii) If a bare CR is encountered in a header line, an extra space is added
3349 after the line terminator so as not to end the header. The reasoning
3350 behind this is that bare CRs in header lines are most likely either
3351 to be mistakes, or people trying to play silly games.
3353 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3354 was being incorrectly given as 18 bytes larger than it should have been.
3355 This is a VOB (very old bug).
3357 60. This may never have affected anything current, but just in case it has:
3358 When the local host is found other than at the start of a list of hosts,
3359 the local host, those with the same MX, and any that follow, are discarded.
3360 When the list in question was part of a longer list of hosts, the following
3361 hosts (not currently being processed) were also being discarded. This no
3362 longer happens. I'm not sure if this situation could ever has previously
3365 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3368 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3369 header. This is something that is recommended in a new Internet Draft, and
3370 is something that is documented as being done by Sendmail. There are two
3371 possible values. For messages generated by the autoreply transport, Exim
3374 Auto-Submitted: auto-replied
3376 whereas for all other generated messages (e.g. bounces) it adds
3378 Auto-Submitted: auto-generated
3380 63. The "personal" condition in filters now includes a test for the
3381 Auto-Submitted: header. If it contains the string "auto-" the message it
3382 not considered personal.
3384 64. Added rcpt_include_affixes as a generic transport option.
3386 65. Added queue_only_override (default true).
3388 66. Added the syslog_duplication option.
3390 67. If what should have been the first header line of a message consisted of
3391 a space followed by a colon, Exim was mis-interpreting it as a header line.
3392 It isn't of course - it is syntactically invalid and should therefore be
3393 treated as the start of the message body. The misbehaviour could have
3394 caused a number of strange effects, including loss of data in subsequent
3395 header lines, and spool format errors.
3397 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3398 client host had authenticated. This control can now be exercised by an ACL
3399 for more flexibility.
3401 69. By default, callouts do not happen when testing with -bh. There is now a
3402 variant, -bhc, which does actually run the callout code, including
3403 consulting and updating the callout cache.
3405 70. Added support for saslauthd authentication, courtesy of Alexander
3408 71. If statvfs() failed on the spool or log directories while checking their
3409 size for availability, Exim confusingly gave the error "space shortage".
3410 Furthermore, in debugging mode it crashed with a floating point exception.
3411 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3412 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3413 serious problem, Exim now writes to the main and panic logs when this
3414 happens, with details of the failure. It then refuses to accept the
3415 incoming message, giving the message "spool directory problem" or "log
3416 directory problem" with a 421 code for SMTP messages.
3418 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3419 0, 1, and 2 exist, because some OS complain for execs without them (see
3420 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3421 descriptors. However, the code omitted to check that the open succeeded,
3422 causing mysterious errors if for some reason the permissions on /dev/null
3423 got screwed. Now Exim writes a message to the main and panic logs, and
3424 bombs out if it can't open /dev/null.
3426 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3427 interact so that it is all more flexible. It is supposed to remain
3428 backwards compatible. Also added extra_local_interfaces.
3430 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3431 to bomb out with an assertion failure - to the client this appears as a
3432 connection drop. This problem occurs in the part of the code that was taken
3433 from the Samba project. Fortunately, the assertion is in a very simple
3434 function, so I have fixed this by reproducing the function inline in the
3435 one place where it is called, and arranging for authentication to fail
3436 instead of killing the process with assert().
3438 75. The SPA client code was not working when the server requested OEM rather
3439 than Unicode encoding.
3441 76. Added code to make require_files with a specific uid setting more usable in
3442 the case where statting the file as root fails - usually a non-root-mounted
3443 NFS file system. When this happens and the failure is EACCES, Exim now
3444 forks a subprocess and does the per-uid checking as the relevant uid.
3446 77. Added process_log_path.
3448 78. If log_file_path was not explicitly set, a setting of check_log_space or
3449 check_log_inodes was ignored.
3451 79. If a space check for the spool or log partitions fails, the incident is now
3452 logged. Of course, in the latter case the data may get lost...
3454 80. Added the %p formatting code to string_format() so that it can be used to
3455 print addresses in debug_print(). Adjusted all the address printing in the
3456 debugging in store.c to use %p rather than %d.
3458 81. There was a concern that a line of code in smtp_in.c could overflow a
3459 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3460 initially expressed, the concern was not well-founded, because trailing
3461 spaces are removed early. However, if the trailing spaces were followed by
3462 a NULL, they did not get removed, so the overflow was possible. Two fixes
3465 (a) I re-wrote the offending code in a cleaner fashion.
3466 (b) If an incoming SMTP command contains a NULL character, it is rejected
3469 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3470 runs initgroups(), so that if the Exim user is in any additional groups,
3471 they will be used during message reception.
3477 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob