1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.269 2005/12/01 14:21:25 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 The code for finding all the local interface addresses on a FreeBSD
10 system running IPv6 was broken. This may well have applied to all BSD
11 systems, as well as to others that have similar system calls. The broken
12 code found IPv4 interfaces correctly, but gave incorrect values for the
13 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
14 that it would not match correctly against @[] and not recognize the IPv6
21 PH/01 Two changes to the default runtime configuration:
23 (1) Move the checks for relay_from_hosts and authenticated clients from
24 after to before the (commented out) DNS black list checks.
26 (2) Add control=submission to the relay_from_hosts and authenticated
27 clients checks, on the grounds that messages accepted by these
28 statements are most likely to be submissions.
30 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
32 (1) Generate an error if the third argument for the ${prvs expansion is
35 (2) Treat a missing third argument of ${prvscheck as if it were an empty
38 (3) Reset the variables that are obtained from the first argument of
39 ${prvscheck and used in the second argument before leaving the code,
40 because their memory is reclaimed, so using them afterwards may do
43 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
44 one (it's much easier than Tom thought :-).
46 (5) Because of (4), we can now allow for the use of $prvscheck_result
47 inside the third argument.
49 PH/03 For some reason, the default setting of PATH when running a command from
50 a pipe transport was just "/usr/bin". I have changed it to
53 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
54 anything to be listed in the output from -bV.
56 PH/05 When a filter generated an autoreply, the entire To: header line was
57 quoted in the delivery log line, like this:
59 => >A.N.Other <ano@some.domain> <original@ddress> ...
61 This has been changed so that it extracts the operative address. There
62 may be more than one such address. If so, they are comma-separated, like
65 => >ano@some.domain,ona@other.domain <original@ddress> ...
67 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
68 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
69 not being looked up in the rDNS to get a host name, Exim was showing the
70 IP address twice in Received: lines, even though the IP addresses were
71 identical. For example:
73 Received: from [1.2.3.4] (helo=[1.2.3.4])
75 However, if the real host name was known, it was omitting the HELO data
76 if it matched the actual IP address. This has been tidied up so that it
77 doesn't show the same IP address twice.
79 PH/07 When both +timestamp and +memory debugging was on, the value given by
80 $tod_xxx expansions could be wrong, because the tod_stamp() function was
81 called by the debug printing, thereby overwriting the timestamp buffer.
82 Debugging no longer uses the tod_stamp() function when +timestamp is set.
84 PH/08 When the original message was included in an autoreply transport, it
85 always said "this is a copy of the message, including all the headers",
86 even if body_only or headers_only was set. It now gives an appropriate
89 PH/09 Applied a patch from the Sieve maintainer which:
92 o adds the (disabled) notify extension core
93 o adds some debug output for the result of if/elsif tests
94 o points to the current vacation draft in the documentation
95 and documents the missing references header update
99 o fixes a bug in processing the envelope test (when testing
100 multiple envelope elements, the last element determinted the
103 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
104 Electronic Mail") by including:
106 Auto-submitted: auto-generated
108 in the messages that it generates (bounce messages and others, such as
109 warnings). In the case of bounce messages for non-SMTP mesages, there was
110 also a typo: it was using "Auto_submitted" (underscore instead of
111 hyphen). Since every message generated by Exim is necessarily in response
112 to another message, thes have all been changed to:
114 Auto-Submitted: auto-replied
116 in accordance with these statements in the RFC:
118 The auto-replied keyword:
120 - SHOULD be used on messages sent in direct response to another
121 message by an automatic process,
123 - MUST NOT be used on manually-generated messages,
125 - MAY be used on Delivery Status Notifications (DSNs) and Message
126 Disposition Notifications (MDNs),
128 - MUST NOT be used on messages generated by automatic or periodic
129 processes, except for messages which are automatic responses to
132 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
133 to the default Received: header definition.
135 PH/12 Added log selector acl_warn_skipped (default on).
137 PH/13 After a successful wildlsearch lookup, discard the values of numeric
138 variables because (a) they are in the wrong storage pool and (b) even if
139 they were copied, it wouldn't work properly because of the caching.
141 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
142 checking when decoding. Apparently there are clients that generate
143 overlong encoded strings. Why am I not surprised?
145 PH/15 If the first argument of "${if match_address" was not empty, but did not
146 contain an "@" character, Exim crashed. Now it writes a panic log message
147 and treats the condition as false.
149 PH/16 In autoreply, treat an empty string for "once" the same as unset.
151 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
152 extension "envelope-auth". The code is finished and in agreement with
153 other implementations, but there is no documentation so far and in fact,
154 nobody wrote the draft yet. This extension is currently #undef'ed, thus
155 not changing the active code.
157 Print executed "if" and "elsif" statements when debugging is used. This
158 helps a great deal to understand what a filter does.
160 Document more things not specified clearly in RFC3028. I had all this
161 sorted out, when out of a sudden new issues came to my mind. Oops."
163 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
166 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
167 canonical form (as documented). However, after a host name lookup from
168 the IP address, check_host() was doing a simple string comparison with
169 addresses acquired from the DNS when checking that the found name did
170 have the original IP as one of its addresses. Since any found IPv6
171 addresses are likely to be in abbreviated form, the comparison could
172 fail. Luckily, there already exists a function for doing the comparison
173 by converting both addresses to binary, so now that is used instead of
176 PH/20 There was another similar case to PH/19, when a complete host name was
177 given in a host list; looking up its IP address could give an abbreviated
178 form, whereas the current host's name might or might not be abbreviated.
179 The same fix has been applied.
185 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
186 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
189 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
190 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
192 PH/03 Typo: missing ".o" in src/pcre/Makefile.
194 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
195 header line, restrict the check to what is listed in RFCs 2369 and 2929.
196 Also, for "Auto-Submitted", treat anything other than "no" as
197 non-personal, in accordance with RFC 3834. (Previously it treated
198 anything starting "auto-" as non-personal.)
200 TF/01 The control=submission/name=... option had a problem with syntax
201 errors if the name included a slash character. The /name= option
202 now slurps the rest of the string, so it can include any characters
203 but it must come last in the list of options (after /sender_retain
206 PH/05 Some modifications to the interface to the fake nameserver for the new
209 PH/06 Add ${if def:sender_address {(envelope-from <$sender_address>)\n\t}}
210 into the default Received: header string.
217 TK/01 Added the "success_on_redirect" address verification option. See
218 NewStuff for rationale and an example.
220 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
222 PH/02 Patch to exigrep to allow it to work on syslog lines.
224 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
225 fread() to skip over the body file's header line, because in Cygwin the
226 header line is locked and is inaccessible.
228 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
229 co-exist for some time) to make it clear that it is the Exim ID that is
230 referenced, not the Message-ID: header line.
232 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
233 string_format() function, because snprintf() does not exist on all
236 PH/06 The use of forbid_filter_existstest now also locks out the use of the
237 ${stat: expansion item.
239 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
240 protocol synchronization error", to keep the pedants happy.
242 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
243 well as for IRIX systems, when gcc is being used. See the host.c source
246 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
248 PH/10 Named domain lists were not working if used in a queue_smtp_domains
251 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
252 transport and to the smtp transport in LMTP mode.
254 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
256 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
257 run a filter in a subprocess. This could lead to confusion in subsequent
258 lookups in the parent process. There should also be a search_tidyup() at
259 the end of the subprocess.
261 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
262 only if the host matched helo_try_verify_hosts, which caused the
263 verification to occur when the EHLO/HELO command was issued. The ACL just
264 tested the remembered result. Now, if a previous verification attempt has
265 not happened, "verify = helo" does it there and then.
267 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
269 TK/03 Fix log output including CR from clamd.
271 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
272 exist provoked a memory error which could cause a segfault.
274 PH/15 Installed PCRE 6.2
276 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
278 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
279 of the problem. Specifically, suggested +O2 rather than +O1 for the
282 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
284 PH/20 If a delivery was routed to a non-standard port by means of an SRV
285 record, the port was not correctly logged when the outgoing_port log
286 selector was set (it logged the transort's default port).
288 PH/21 Added support for host-specific ports to manualroute, queryprogram,
289 fallback_hosts, and "hosts" in the smtp transport.
291 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
292 host errors such as "Connection refused".
294 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
295 authentication with radiusclient 0.4.9:
297 - Error returned from rc_read_config was caught wrongly
298 - Username/password not passed on to radius server due to wrong length.
300 The presumption is that some radiusclient API changes for 4.51/PH/17
301 were not taken care of correctly. The code is still untested by me (my
302 Linux distribution still has 0.3.2 of radiusclient), but it was
303 contributed by a Radius user.
305 PH/24 When doing a callout, the value of $domain wasn't set correctly when
306 expanding the "port" option of the smtp transport.
308 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
309 while reading a MIME header. Thanks to Tom Hughes for a patch.
311 PH/24 Include config.h inside local_scan.h so that configuration settings are
314 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
315 that in an ACL for RCPT (for example), you can examine exactly what was
318 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
319 commands, but it was not correctly comparing the address with the actual
320 client host address. Thus, it would show the EHLO address in Received:
321 header lines when this was not necessary.
323 PH/27 Added the % operator to ${eval:}.
325 PH/28 Exim tries to create and chdir to its spool directory when it starts;
326 it should be ignoring failures (because with -C, for example, it has lost
327 privilege). It wasn't ignoring creation failures other than "already
330 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
333 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
334 because an input line was too long, either on its own, or by virtue of
335 too many continuations, the temporary file was not being removed, and the
336 return code was incorrect.
338 PH/31 Missing "BOOL" in function definition in filtertest.c.
340 PH/32 Applied Sieve patches from the maintainer.
342 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
344 PH/33 Added "verify = not_blind".
346 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
347 Local/Makefile (with some defaults set). These are used in built scripts
348 such as exicyclog, but they have never been used in the exim_install
349 script (though there are many overriding facilities there). I have
350 arranged that the exim_install script now takes note of these two
353 PH/35 Installed configuration files for Dragonfly.
355 PH/36 When a locally submitted message by a trusted user did not contain a
356 From: header, and the sender address was obtained from -f or from an SMTP
357 MAIL command, and the trusted user did not use -F to supply a sender
358 name, $originator_name was incorrectly used when constructing a From:
359 header. Furthermore, $originator_name was used for submission mode
360 messages from external hosts without From: headers in a similar way,
361 which is clearly wrong.
363 PH/37 Added control=suppress_local_fixups.
365 PH/38 When log_selector = +received_sender was set, and the addition of the
366 sender made the log line's construction buffer exactly full, or one byte
367 less than full, an overflow happened when the terminating "\n" was
370 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
371 when the result of a list match is failure because a DNS lookup failed.
373 PH/40 RM_COMMAND is now used in the building process.
375 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
376 the "build-* directories that it finds.
378 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
379 address in a domain literal was a prefix of an interface address.
381 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
382 when verifying a sender address, unless rewrite_headers is false.
384 PH/44 Wrote a long comment about why errors_to addresses are verified as
385 recipients, not senders.
387 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
388 the ratelimit ACL was added.
390 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
392 PH/46 Added extra information about PostgreSQL errors to the error string.
394 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
395 avoiding the need to install special zones in a real server. This is
396 backwards compatible; if it can't find the fake resolver, it drops back.
397 Thus, both old and new test suites can be run.
399 TF/02 Added util/ratelimit.pl
401 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
402 clock is set back in time.
404 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
405 Candler <B.Candler@pobox.com>.
407 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
408 OK for addresses that are the result of redirections.
410 PH/48 A number of further additions for the benefit of the new test suite,
411 including a fake gethostbyname() that interfaces to the fake DNS resolver
414 TF/06 The fix for widen_domains has also been applied to qualify_single and
415 search_parents which are the other dnslookup options that can cause
418 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
421 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
423 TF/07 Exim produced the error message "an SRV record indicated no SMTP
424 service" if it encountered an MX record with an empty target hostname.
425 The message is now "an MX or SRV record indicated no SMTP service".
427 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
428 if the DNS of the sending site is misconfigured. This is quite a
429 common situation. This change restores the behaviour of treating a
430 helo verification defer as a failure.
432 PH/51 If self=fail was set on a router, the bounce message did not include the
433 actual error message.
439 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
441 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
442 command itself ended in error, the underlying message about the transport
443 filter timeout was being overwritten with the pipe command error. Now the
444 underlying error message should be appended to the second error message.
446 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
448 PH/02 Reduce the amount of output that "make" produces by default. Full output
449 can still be requested.
451 PH/03 The warning log line about a condition test deferring for a "warn" verb
452 was being output only once per connection, rather than after each
453 occurrence (because it was using the same function as for successful
454 "warn" verbs). This seems wrong, so I have changed it.
456 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
457 it should not have, which might have caused a crash in the right
458 circumstances, but probably never did.
460 PH/04 Installed a modified version of Tony Finch's patch to make submission
461 mode fix the return path as well as the Sender: header line, and to
462 add a /name= option so that you can make the user's friendly name appear
465 TF/03 Added the control = fakedefer ACL modifier.
467 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
468 Mark Lowes for thorough testing.
470 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
472 TK/03 Merged latest SRS patch from Miles Wilton.
474 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
475 with the definition in sysexits.h (which is #included earlier).
476 Fortunately, Exim does not actually use EX_OK. The code used to try to
477 preserve the sysexits.h value, by assumimg that macro definitions were
478 scanned for macro replacements. I have been disabused of this notion,
479 so now the code just undefines EX_OK before #including unistd.h.
481 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
482 in the smtp transport. When a block could not be written in a single
483 write() function, the timeout was being re-applied to each part-write.
484 This seems wrong - if the receiver was accepting one byte at a time it
485 would take for ever. The timeout is now adjusted when this happens. It
486 doesn't have to be particularly precise.
488 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
489 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
491 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
492 without a domain if the check to <postmaster@domain> fails.
494 SC/01 Eximstats: added -xls and the ability to specify output files
495 (patch written by Frank Heydlauf).
497 SC/02 Eximstats: use FileHandles for outputing results.
499 SC/03 Eximstats: allow any combination of xls, txt, and html output.
501 SC/04 Eximstats: fixed display of large numbers with -nvr option
503 SC/05 Eximstats: fixed merging of reports with empty tables.
505 SC/06 Eximstats: added the -include_original_destination flag
507 SC/07 Eximstats: removed tabs and trailing whitespace.
509 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
511 TK/06 MBOX spool code: Add real "From " MBOX separator line
512 so the .eml file is really in mbox format (even though
513 most programs do not really care). Patch from Alex Miller.
515 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
516 The latter is generated from $received_to and is only set if the
517 message has one envelope recipient. SA can use these headers,
518 obviously out-of-the-box. Patch from Alex Miller.
520 PH/08 The ${def test on a variable was returning false if the variable's
521 value was "0", contrary to what the specification has always said!
522 The result should be true unless the variable is empty.
524 PH/09 The syntax error of a character other than { following "${if
525 def:variable_name" (after optional whitespace) was not being diagnosed.
526 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
527 accidental colon was present, for example, could give incorrect results.
529 PH/10 Tidied the code in a number of places where the st_size field of a stat()
530 result is used (not including appendfile, where other changes are about
533 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
534 This involved changing a lot of size variables from int to off_t. It
535 should work with maildirs and everything.
537 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
538 spamd dying while we are connected to it.
540 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
543 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
544 function for BDB 4.3.
546 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
548 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
549 the sources for PCRE have been re-arranged and more files are now
552 PH/15 The code I had for printing potentially long long variables in PH/11
553 above was not the best (it lost precision). The length of off_t variables
554 is now inspected at build time, and an appropriate printing format (%ld
555 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
556 to be "long long int" or "long int". This is needed for the internal
557 formatting function string_vformat().
559 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
560 the configuration file to be ":syslog", then the script "guesses" where
561 the logs files are, rather than using the compiled in default. In our
562 case the guess is not the same as the compiled default, so the script
563 suddenly stopped working when I started to use syslog. The patch checks
564 to see if log_file_path is "". If so, it attempts to read it from exim
565 with no configuration file to get the compiled in version, before it
566 falls back to the previous guessing code."
568 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
569 implementing BATV in an Exim configuration. See NewStuff for the gory
572 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
573 Makefile that are specific to HP-UX.
575 PH/18 If the "use_postmaster" option was set for a recipient callout together
576 with the "random" option, the postmaster address was used as the MAIL
577 FROM address for the random test, but not for the subsequent recipient
578 test. It is now used for both.
580 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
581 patch removes a few documentation additions to RFC 3028, because the
582 latest draft now contains them. It adds the new en;ascii-case comparator
583 and a new error check for 8bit text in MIME parts. Comparator and
584 require names are now matched exactly. I enabled the subaddress
585 extension, but it is not well tested yet (read: it works for me)."
587 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
588 rework some of the code of TK/09 above to avoid the hardwired use of
589 "%lld" and "long long". Replaced the call to snprintf() with a call to
592 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
593 records point to non-existent hosts", "retry timeout exceeded", and
594 "retry time not reached for any host after a long failure period".
596 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
597 experimental DomainKeys support:
599 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
600 (2) On an error such as an illegally used "control", the wrong name for
601 the control was given.
603 These problems did NOT occur unless DomainKeys support was compiled.
605 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
607 PH/24 Added ${if match_ip condition.
609 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
610 databases so that it will be absolutely obvious if a crash occurs in the
611 DB library. This is a regular occurrence (often caused by mis-matched
614 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
615 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
616 picked up on a user's system that detects such things. There doesn't seem
617 to be a gcc warning option for this - only an attribute that has to be
618 put on the function's prototype. It seems that in Fedora Core 4 they have
619 set this on a number of new functions. No doubt there will be more in due
622 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
623 specify a transport. However, if an address that was verified by such a
624 router was the subject of a callout, Exim crashed because it tried to
625 read the rcpt_include_affixes from the non-existent transport. Now it
626 just assumes that the setting of that option is false. This bug was
627 introduced by 4.51/PH/31.
629 PH/28 Changed -d+all to exclude +memory, because that information is very
630 rarely of interest, but it makes the output a lot bigger. People tend to
631 do -d+all out of habit.
633 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
634 code in os-type was giving problems when libc.so lives in lib64, like on
637 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
638 aren't the modern standard, and it seems that some systems' include files
639 don't always have them. Exim was already checking for some of the newer
640 ones like T_AAAA, and defining it itself. I've added checks for all the
641 record types that Exim uses.
643 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
644 not automatically generating a new one, as it is supposed to. This
645 prevented TLS from working. If the file did exist, but contained invalid
646 data, a new version was generated, as expected. It was only the case of a
647 non-existent file that was broken.
649 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
650 with a change in libdomainkeys > 0.64.
652 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
653 from DNS. If the selector record carries the flag, it now has
654 precedence over the domain-wide flag.
656 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
658 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
659 the use of a port name that isn't defined in /etc/services) occurred, the
660 message was deferred as in a normal delivery, and thus remained on the
661 spool, instead of being failed because of the mua_wrapper setting. This
662 is now fixed, and I tidied up some of the mua_wrapper messages at the
665 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
666 the messages in a hash of arrays rather than using individual hashes.
667 This is a bit cleaner and results in dramatic memory savings, albeit
668 at a slight CPU cost.
670 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
671 as requested by Marc Sherman.
673 SC/10 Eximstats: added histograms for user specified patterns as requested
676 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
678 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
679 fopen() in the content-scanning modules that did not already have it.
685 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
686 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
688 TK/02 Fix ACL "control" statement not being available in MIME ACL.
690 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
692 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
693 to test Sieve filters that use "vacation".
695 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
696 that changes the way the GnuTLS parameters are stored in the cache file.
697 The new format can be generated externally. For backward compatibility,
698 if the data in the cache doesn't make sense, Exim assumes it has read an
699 old-format file, and it generates new data and writes a new file. This
700 means that you can't go back to an older release without removing the
703 PH/03 A redirect router that has both "unseen" and "one_time" set does not
704 work if there are any delivery delays because "one_time" forces the
705 parent to be marked "delivered", so its unseen clone is never tried
706 again. For this reason, Exim now forbids the simultaneous setting of
709 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
710 redirected to themselves ("homonym" addresses). Read the long ChangeLog
711 entry if you want to know the details. The fix, however, neglected to
712 consider the case when local delivery batching is involved. The test for
713 "previously delivered" was not happening when checking to see if an
714 address could be batched with a previous (undelivered) one; under
715 certain circumstances this could lead to multiple deliveries to the same
718 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
719 in its include files, and this causes problems building Exim.
721 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
722 header_syntax) but Exim was just ignoring anything given after a slash.
723 In particular, this caused confusion with an attempt to use "verify =
724 reverse_host_lookup/defer_ok". An error is now given when options are
725 supplied for verify items that do not have them. (Maybe reverse_host_
726 lookup should have a defer_ok option, but that's a different point.)
728 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
729 defined by RFC 821) to 2048, because there were problems with some AUTH
730 commands, and RFC 1869 says the size should be increased for extended
731 SMTP commands that take arguments.
733 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
736 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
737 "unknown" error; now it says that the functionality isn't in the binary.
739 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
740 an address' error message when a string expansion fails (syntax or
741 whatever). Otherwise the password may appear in the log. Following change
742 PH/42 below, there is no longer a chance of it appearing in a bounce
745 PH/11 Installed exipick version 20050225.0 from John Jetmore.
747 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
748 of its addresses was ever tried. (Bugzilla bug #2.)
750 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
751 the result incorrectly in the debug output. (It correctly added a newline
752 to what was transported.)
754 TF/01 Added $received_time.
756 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
757 commented out examples of how to interface to a virus scanner and to
758 SpamAssassin. Also added commented examples of av_scanner and
759 spamd_address settings.
761 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
762 and controls are allowed in which ACLs. There were a couple of minor
763 errors. Some of the entries in the conditions table (which is a table of
764 where they are NOT allowed) were getting very unwieldy; rewrote them as a
765 negation of where the condition IS allowed.
767 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
769 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
770 header file does not have a version number, so I've had to invent a new
771 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
772 API. The code is untested by me (my Linux distribution still has 0.3.2 of
773 radiusclient), but it was contributed by a Radius user.
775 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
776 files or directories, for OpenSSL.
778 PH/19 When an Exim process that is running as root has to create an Exim log
779 file, it does so in a subprocess that runs as exim:exim so as to get the
780 ownership right at creation (otherwise, other Exim processes might see
781 the file with the wrong ownership). There was no test for failure of this
782 fork() call, which would lead to the process getting stuck as it waited
783 for a non-existent subprocess. Forks do occasionally fail when resources
784 run out. I reviewed all the other calls to fork(); they all seem to check
787 PH/20 When checking for unexpected SMTP input at connect time (before writing
788 the banner), Exim was not dealing correctly with a non-positive return
789 from the read() function. If the client had disconnected by this time,
790 the result was a log entry for a synchronization error with an empty
791 string after "input=" when read() returned zero. If read() returned -1
792 (an event I could not check), uninitialized data bytes were printed.
793 There were reports of junk text (parts of files, etc) appearing after
796 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
798 PH/22 Added support for macro redefinition, and (re)definition in between
799 driver and ACL definitions.
801 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
802 forgetting to use the resulting value; it was using the unexpanded value.
804 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
805 hadn't been configured. The fix is from Juergen Kreileder, who
806 understands it better than I do:
808 "Here's what I see happening with three configured cyrus_sasl
809 authenticators configured (plain, login, cram-md5):
811 On startup auth_cyrus_sasl_init() gets called for each of these.
812 This means three calls to sasl_listmech() without a specified mech_list.
813 => SASL tests which mechs of all available mechs actually work
814 => three warnings about OTP not working
815 => the returned list contains: plain, login, cram-md5, digest-md5, ...
817 With the patch, sasl_listmech() also gets called three times. But now
818 SASL's mech_list option is set to the server_mech specified in the the
819 authenticator. Or in other words, the answer from sasl_listmech()
820 gets limited to just the mech you're testing for (which is different
822 => the return list contains just 'plain' or 'login', 'cram-md5' or
823 nothing depending on the value of ob->server_mech.
825 I've just tested the patch: Authentication still works fine,
826 unavailable mechs specified in the exim configuration are still
827 caught, and the auth.log warnings about OTP are gone."
829 PH/25 When debugging is enabled, the contents of the command line are added
830 to the debugging output, even when log_selector=+arguments is not
833 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
834 answer is "GNU", and only if the return is "GNU/something" is the answer
837 PH/27 $acl_verify_message is now set immediately after the failure of a
838 verification in an ACL, and so is available in subsequent modifiers. In
839 particular, the message can be preserved by coding like this:
841 warn !verify = sender
842 set acl_m0 = $acl_verify_message
844 Previously, $acl_verify_message was set only while expanding "message"
845 and "log_message" when a very denied access.
847 PH/28 Modified OS/os.c-Linux with
849 -#ifndef OS_LOAD_AVERAGE
850 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
852 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
853 nomenclature these days.)
855 PH/29 Installed patch from the Sieve maintainer that adds the options
856 sieve_useraddress and sieve_subaddress to the redirect router.
858 PH/30 In these circumstances:
859 . Two addresses routed to the same list of hosts;
860 . First host does not offer TLS;
861 . First host accepts first address;
862 . First host gives temporary error to second address;
863 . Second host offers TLS and a TLS session is established;
864 . Second host accepts second address.
865 Exim incorrectly logged both deliveries with the TLS parameters (cipher
866 and peerdn, if requested) that were in fact used only for the second
869 PH/31 When doing a callout as part of verifying an address, Exim was not paying
870 attention to any local part prefix or suffix that was matched by the
871 router that accepted the address. It now behaves in the same way as it
872 does for delivery: the affixes are removed from the local part unless
873 rcpt_include_affixes is set on the transport.
875 PH/32 Add the sender address, as F=<...>, to the log line when logging a
876 timeout during the DATA phase of an incoming message.
878 PH/33 Sieve envelope tests were broken for match types other than :is. I have
879 applied a patch sanctioned by the Sieve maintainer.
881 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
882 the uid or gid is negative. A case of a negative gid caused this to be
883 noticed. The fix allows for either to be negative.
885 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
886 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
889 PH/36 The change PH/12 above was broken. Fixed it.
891 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
892 the grounds that routing the same address twice would always produce the
893 same answer. This might have been true once, but it is certainly no
894 longer true now. Routing a child address may depend on the previous
895 routing that produced that child. Some complicated redirection strategies
896 went wrong when messages had multiple recipients, and made Exim's
897 behaviour dependent on the order in which the addresses were given.
899 I have moved the duplicate checking until after the routing is complete.
900 Exim scans the addresses that are assigned to local and remote
901 transports, and removes any duplicates. This means that more work will be
902 done, as duplicates will always all be routed, but duplicates are
903 presumably rare, so I don't expect this is of any significance.
905 For deliveries to pipes, files, and autoreplies, the duplicate checking
906 still happens during the routing process, since they are not going to be
909 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
910 It corrects a timeout issue with spamd. This is Ian's comment: "The
911 background is that sometimes spamd either never reads data from a
912 connection it has accepted, or it never writes response data. The exiscan
913 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
914 blindly assumes that writes won't block so it may never time out."
916 PH/39 Allow G after quota size as well as K and M.
918 PH/40 The value set for $authenticated_id in an authenticator may not contain
919 binary zeroes or newlines because the value is written to log lines and
920 to spool files. There was no check on this. Now the value is run through
921 the string_printing() function so that such characters are converted to
922 printable escape sequences.
924 PH/41 $message_linecount is a new variable that contains the total number of
925 lines in the message. Compare $body_linecount, which is the count for the
928 PH/42 Exim no longer gives details of delivery errors for specific addresses in
929 bounce and delay warning messages, except in certain special cases, which
932 (a) An SMTP error message from a remote host;
933 (b) A message specified in a :fail: redirection;
934 (c) A message specified in a "fail" command in a system filter;
935 (d) A message specified in a FAIL return from the queryprogram router;
936 (e) A message specified by the cannot_route_message router option.
938 In these cases only, Exim does include the error details in bounce and
939 warning messages. There are also a few cases where bland messages such
940 as "unrouteable address" or "local delivery error" are given.
942 PH/43 $value is now also set for the "else" part of a ${run expansion.
944 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
945 being worked on, but at least Exim now implements the latest version to
948 PH/45 In a pipe transport, although a timeout while waiting for the pipe
949 process to complete was treated as a delivery failure, a timeout while
950 writing the message to the pipe was logged, but erroneously treated as a
951 successful delivery. Such timeouts include transport filter timeouts. For
952 consistency with the overall process timeout, these timeouts are now
953 treated as errors, giving rise to delivery failures by default. However,
954 there is now a new Boolean option for the pipe transport called
955 timeout_defer, which, if set TRUE, converts the failures into defers for
956 both kinds of timeout. A transport filter timeout is now identified in
959 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
960 systems where "make" and "gmake" are different, calling "gmake" at top
961 level broke things. I've arranged for the value of $(MAKE) to be passed
962 from the Makefile to this script so that it can call the same version of
966 A note about Exim versions 4.44 and 4.50
967 ----------------------------------------
969 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
970 changes of various kinds. As a consequence, a big documentation update was
971 needed. This delayed the release for rather longer than seemed good, especially
972 in the light of a couple of (minor) security issues. Therefore, the changes
973 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
974 release. So 4.44 and 4.50 are in effect two different branches that both start
977 I have left the 4.50 change log unchanged; it contains all the changes since
978 4.43. The change log for 4.44 is below; many of its items are identical to
979 those for 4.50. This seems to be the most sensible way to preserve the
980 historical information.
986 1. Minor wording change to the doc/README.SIEVE file.
988 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
989 computation of the current number of files was incorrect.
991 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
992 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
995 4. Give more explanation in the error message when the command for a transport
996 filter fails to execute.
998 5. There are several places where Exim runs a non-Exim command in a
999 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1000 was being done only for the command run by the queryprogram router. It is
1001 now done for all such subprocesses. The other cases are: ${run, transport
1002 filters, and the commands run by the lmtp and pipe transports.
1004 6. Added CONFIGURE_GROUP build-time option.
1006 7. Some older OS have a limit of 256 on the maximum number of file
1007 descriptors. Exim was using setrlimit() to set 1000 as a large value
1008 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1009 systems. I've change it so that if it can't get 1000, it tries for 256.
1011 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1012 was an oversight, and furthermore, ever since the addition of extra
1013 controls (e.g. 4.43/32), the checks on when to allow different forms of
1014 "control" were broken. There should now be diagnostics for all cases when a
1015 control that does not make sense is encountered.
1017 9. Added the /retain_sender option to "control=submission".
1019 10. $recipients is now available in the predata ACL (oversight).
1021 11. Tidy the search cache before the fork to do a delivery from a message
1022 received from the command line. Otherwise the child will trigger a lookup
1023 failure and thereby defer the delivery if it tries to use (for example) a
1024 cached ldap connection that the parent has called unbind on.
1026 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1027 of $address_data from the recipient verification was clobbered by the
1028 sender verification.
1030 13. The value of address_data from a sender verification is now available in
1031 $sender_address_data in subsequent conditions in the ACL statement.
1033 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1035 15. Added a new option "connect=<time>" to callout options, to set a different
1038 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1039 was its contents. (It was OK if the option was not defined at all.)
1041 17. A "Completed" log line is now written for messages that are removed from
1042 the spool by the -Mrm option.
1044 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1045 information about exactly what failed.
1047 19. Added -dd to debug only the daemon process.
1049 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1050 handles timeouts, both on the server side and network timeouts. Renamed the
1051 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1053 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1054 It is now set to "smtps".
1056 22. $host_address is now set to the target address during the checking of
1057 ignore_target_hosts.
1059 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1060 being passed; this would have caused $sender_host_name to have been used if
1061 matching the list had actually called for a host name (not very likely,
1062 since this list is usually IP addresses). A host name is now passed as
1065 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1066 code that specifies a non-restarting handler (typically sigaction() in
1067 modern systems) in an attempt to fix a rare and obscure crash bug.
1069 25. Narrowed the window for a race in the daemon that could cause it to ignore
1070 SIGCHLD signals. This is not a major problem, because they are used only to
1071 wake it up if nothing else does.
1073 26. A malformed maildirsize file could cause Exim to calculate negative values
1074 for the mailbox size or file count. Odd effects could occur as a result.
1075 The maildirsize information is now recalculated if the size or filecount
1078 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1079 support for a long time. Removed HAVE_SYS_VFS_H.
1081 28. Installed the latest version of exipick from John Jetmore.
1083 29. In an address list, if the pattern was not a regular expression, an empty
1084 subject address (from a bounce message) matched only if the pattern was an
1085 empty string. Non-empty patterns were not even tested. This was the wrong
1086 because it is perfectly reasonable to use an empty address as part of a
1087 database query. An empty address is now tested by patterns that are
1088 lookups. However, all the other forms of pattern expect the subject to
1089 contain a local part and a domain, and therefore, for them, an empty
1090 address still always fails if the pattern is not itself empty.
1092 30. Exim went into a mad DNS loop when attempting to do a callout where the
1093 host was specified on an smtp transport, and looking it up yielded more
1094 than one IP address.
1096 31. Re-factored the code for checking spool and log partition space into a
1097 function that finds that data and another that does the check. The former
1098 is then used to implement four new variables: $spool_space, $log_space,
1099 $spool_inodes, and $log_inodes.
1101 32. The RFC2047 encoding function was originally intended for short strings
1102 such as real names; it was not keeping to the 75-character limit for
1103 encoded words that the RFC imposes. It now respects the limit, and
1104 generates multiple encoded words if necessary. To be on the safe side, I
1105 have increased the buffer size for the ${rfc2047: expansion operator from
1108 33. It is now permitted to omit both strings after an "if" condition; if the
1109 condition is true, the result is "true". As before, when the second string
1110 is omitted, a false condition yields an empty string. This makes it less
1111 cumbersome to write custom ACL and router conditions.
1113 34. Failure to deliver a bounce message always caused it to be frozen, even if
1114 there was an errors_to setting on the router. The errors_to setting is now
1117 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1118 canonical form (fully expanded) before being placed in
1119 $sender_host_address.
1121 36. The table in the code that translates DNS record types into text (T_A to
1122 "A" for instance) was missing entries for NS and CNAME. It is just possible
1123 that this could have caused confusion if both these types were looked up
1124 for the same domain, because the text type is used as part of Exim's
1125 per-process caching. But the chance of anyone hitting this buglet seems
1128 37. The dnsdb lookup has been extended in a number of ways.
1130 (1) There is a new type, "zns", which walks up the domain tree until it
1131 finds some nameserver records. It should be used with care.
1133 (2) There is a new type, "mxh", which is like "mx" except that it returns
1134 just the host names, not the priorities.
1136 (3) It is now possible to give a list of domains (or IP addresses) to be
1137 looked up. The behaviour when one of the lookups defers can be
1138 controlled by a keyword.
1140 (4) It is now possible to specify the separator character for use when
1141 multiple records are returned.
1143 38. The dnslists ACL condition has been extended: it is now possible to supply
1144 a list of IP addresses and/or domains to be looked up in a particular DNS
1147 39. Added log_selector=+queue_time_overall.
1149 40. When running the queue in the test harness, wait just a tad after forking a
1150 delivery process, to get repeatability of debugging output.
1152 41. Include certificate and key file names in error message when GnuTLS fails
1153 to set them up, because the GnuTLS error message doesn't include the name
1154 of the failing file when there is a problem reading it.
1156 42. Allow both -bf and -bF in the same test run.
1158 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1160 44. The "Exiscan patch" is now merged into the mainline Exim source.
1162 45. Sometimes the final signoff response after QUIT could fail to get
1163 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1164 < 0 before doing a fflush(). This bug looks as though it goes back to the
1165 introduction of TLS in release 3.20, but "sometimes" must have been rare
1166 because the tests only now provoked it.
1168 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1169 (this can affect the format of dates).
1171 47. exim_tidydb, when checking for the continued existence of a message for
1172 which it has found a message-specific retry record, was not finding
1173 messages that were in split spool directories. Consequently, it was
1174 deleting retry records that should have stayed in existence.
1176 48. Steve fixed some bugs in eximstats.
1178 49. The SPA authentication driver was not abandoning authentication and moving
1179 on to the next authenticator when an expansion was forced to fail,
1180 contradicting the general specification for all authenticators. Instead it
1181 was generating a temporary error. It now behaves as specified.
1183 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1184 (the order specifies the preference for clients). The order is now AES256,
1185 AES128, 3DES, ARCFOUR128.
1187 51. Small patch to Sieve code - explicitly set From: when generating an
1190 52. Exim crashed if a remote delivery caused a very long error message to be
1191 recorded - for instance if somebody sent an entire SpamAssassin report back
1192 as a large number of 550 error lines. This bug was coincidentally fixed by
1193 increasing the size of one of Exim's internal buffers (big_buffer) that
1194 happened as part of the Exiscan merge. However, to be on the safe side, I
1195 have made the code more robust (and fixed the comments that describe what
1198 53. Now that there can be additional text after "Completed" in log lines (if
1199 the queue_time_overall log selector is set), a one-byte patch to exigrep
1200 was needed to allow it to recognize "Completed" as not the last thing in
1203 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1204 patch that reportedly fixes this has been added. I am not expert enough to
1205 create a test for it. This is what the patch creator wrote:
1207 "I found a little strange behaviour of ldap code when working with
1208 Windows 2003 AD Domain, where users was placed in more than one
1209 Organization Units. When I tried to give exim partial DN, the exit code
1210 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1211 But simultaneously result of request was absolutely normal ldap result,
1212 so I produce this patch..."
1214 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1215 REFERENCE, so I have modified the code to exclude the patch when that macro
1218 55. Some experimental protocols are using DNS PTR records for new purposes. The
1219 keys for these records are domain names, not reversed IP addresses. The
1220 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1221 leaves it alone. Component reversal etc. now happens only for IP addresses.
1224 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1226 57. Double the size of the debug message buffer (to 2048) so that more of very
1227 long debug lines gets shown.
1229 58. The exicyclog utility now does better if the number of log files to keep
1230 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1232 59. Two changes related to the smtp_active_hostname option:
1234 (1) $smtp_active_hostname is now available as a variable.
1235 (2) The default for smtp_banner uses $smtp_active_hostname instead
1236 of $primary_hostname.
1238 60. The host_aton() function is supposed to be passed a string that is known
1239 to be a valid IP address. However, in the case of IPv6 addresses, it was
1240 not checking this. This is a hostage to fortune. Exim now panics and dies
1241 if the condition is not met. A case was found where this could be provoked
1242 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1243 components; fortuitously, this particular loophole had already been fixed
1244 by change 4.50/55 above.
1246 If there are any other similar loopholes, the new check in host_aton()
1247 itself should stop them being exploited. The report I received stated that
1248 data on the command line could provoke the exploit when Exim was running as
1249 exim, but did not say which command line option was involved. All I could
1250 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1251 running as the user.
1254 61. There was a buffer overflow vulnerability in the SPA authentication code
1255 (which came originally from the Samba project). I have added a test to the
1256 spa_base64_to_bits() function which I hope fixes it.
1259 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1260 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1262 63. The daemon start-up calls getloadavg() while still root for those OS that
1263 need the first call to be done as root, but it missed one case: when
1264 deliver_queue_load_max is set with deliver_drop_privilege. This is
1265 necessary for the benefit of the queue runner, because there is no re-exec
1266 when deliver_drop_privilege is set.
1268 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1269 This has been fixed.
1271 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1272 was in use, was not putting the data itself into the right store pool;
1273 consequently, it could be overwritten for a subsequent message in the same
1274 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1277 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1279 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1280 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1281 the function were treating the return as a boolean value, which happened to
1282 work because 0=false and not-0=true, but is not correct code.
1284 68. The host_aton() function was not handling scoped IPv6 addresses (those
1285 with, for example, "%eth0" on the end) correctly.
1287 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1288 negated items (that is, ~something) in unsigned ints. Some compilers
1289 apparently mutter when there is no cast.
1291 70. If an address verification called from an ACL failed, and did not produce a
1292 user-specific message (i.e. there was only a "system" message), nothing was
1293 put in $acl_verify_message. In this situation, it now puts the system
1296 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1297 session; change 4.31/43 added the unwanted input to the log line - except
1298 that it did not do this in the start of session case. It now does.
1300 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1301 This is wrong and can cause the other end to generate a synchronization
1302 error if it is another Exim or anything else that does the synchronization
1303 check. A QUIT command is no longer sent after a timeout.
1305 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1306 during host lookups.
1308 74. The defer_ok option of callout verification was not working if it was used
1309 when verifying addresses in header lines, that is, for this case:
1311 verify = header_sender/callout=defer_ok
1313 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1314 those file descriptors could be used for SMTP connections. If anything
1315 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1316 could be sent to the SMTP client, causing chaos. The daemon now opens
1317 stdin, stdout, and stderr to /dev/null when it puts itself into the
1320 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1321 log by default. The user can override this with suitable Perl magic.
1323 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1324 the log message when discard triggers, was not working for the DATA ACL or
1325 for the non-SMTP ACL.
1327 78. Error message wording change in sieve.c.
1329 79. If smtp_accept_max_per_host was set, the number of connections could be
1330 restricted to fewer than expected, because the daemon was trying to set up
1331 a new connection before checking whether the processes handling previous
1332 connections had finished. The check for completed processes is now done
1333 earlier. On busy systems, this bug wouldn't be noticed because something
1334 else would have woken the daemon, and it would have reaped the completed
1337 80. If a message was submitted locally by a user whose login name contained one
1338 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1339 It caused a spool format error. I have fixed the spool reading code. A
1340 related problem was that the "from" clause in the Received: line became
1341 illegal because of the space(s). It is now covered by ${quote_local_part.
1343 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1346 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1348 83. Patch from Sieve maintainer to fix unterminated string problem in
1349 "vacation" handling.
1351 84. Some minor changes to the Linux configuration files to help with other
1352 OS variants using glibc.
1354 85. One more patch for Sieve to update vacation handling to latest spec.
1357 ----------------------------------------------------
1358 See the note above about the 4.44 and 4.50 releases.
1359 ----------------------------------------------------
1365 1. Change 4.43/35 introduced a bug that caused file counts to be
1366 incorrectly computed when quota_filecount was set in an appendfile
1369 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1370 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1373 3. Give more explanation in the error message when the command for a transport
1374 filter fails to execute.
1376 4. There are several places where Exim runs a non-Exim command in a
1377 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1378 was being done only for the command run by the queryprogram router. It is
1379 now done for all such subprocesses. The other cases are: ${run, transport
1380 filters, and the commands run by the lmtp and pipe transports.
1382 5. Some older OS have a limit of 256 on the maximum number of file
1383 descriptors. Exim was using setrlimit() to set 1000 as a large value
1384 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1385 systems. I've change it so that if it can't get 1000, it tries for 256.
1387 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1388 was an oversight, and furthermore, ever since the addition of extra
1389 controls (e.g. 4.43/32), the checks on when to allow different forms of
1390 "control" were broken. There should now be diagnostics for all cases when a
1391 control that does not make sense is encountered.
1393 7. $recipients is now available in the predata ACL (oversight).
1395 8. Tidy the search cache before the fork to do a delivery from a message
1396 received from the command line. Otherwise the child will trigger a lookup
1397 failure and thereby defer the delivery if it tries to use (for example) a
1398 cached ldap connection that the parent has called unbind on.
1400 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1401 of $address_data from the recipient verification was clobbered by the
1402 sender verification.
1404 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1405 was its contents. (It was OK if the option was not defined at all.)
1407 11. A "Completed" log line is now written for messages that are removed from
1408 the spool by the -Mrm option.
1410 12. $host_address is now set to the target address during the checking of
1411 ignore_target_hosts.
1413 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1414 being passed; this would have caused $sender_host_name to have been used if
1415 matching the list had actually called for a host name (not very likely,
1416 since this list is usually IP addresses). A host name is now passed as
1419 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1420 code that specifies a non-restarting handler (typically sigaction() in
1421 modern systems) in an attempt to fix a rare and obscure crash bug.
1423 15. Narrowed the window for a race in the daemon that could cause it to ignore
1424 SIGCHLD signals. This is not a major problem, because they are used only to
1425 wake it up if nothing else does.
1427 16. A malformed maildirsize file could cause Exim to calculate negative values
1428 for the mailbox size or file count. Odd effects could occur as a result.
1429 The maildirsize information is now recalculated if the size or filecount
1432 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1433 support for a long time. Removed HAVE_SYS_VFS_H.
1435 18. Updated exipick to current release from John Jetmore.
1437 19. Allow an empty sender to be matched against a lookup in an address list.
1438 Previously the only cases considered were a regular expression, or an
1441 20. Exim went into a mad DNS lookup loop when doing a callout where the
1442 host was specified on the transport, if the DNS lookup yielded more than
1445 21. The RFC2047 encoding function was originally intended for short strings
1446 such as real names; it was not keeping to the 75-character limit for
1447 encoded words that the RFC imposes. It now respects the limit, and
1448 generates multiple encoded words if necessary. To be on the safe side, I
1449 have increased the buffer size for the ${rfc2047: expansion operator from
1452 22. Failure to deliver a bounce message always caused it to be frozen, even if
1453 there was an errors_to setting on the router. The errors_to setting is now
1456 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1457 canonical form (fully expanded) before being placed in
1458 $sender_host_address.
1460 24. Updated eximstats to version 1.33
1462 25. Include certificate and key file names in error message when GnuTLS fails
1463 to set them up, because the GnuTLS error message doesn't include the name
1464 of the failing file when there is a problem reading it.
1466 26. Expand error message when OpenSSL has problems setting up cert/key files.
1469 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1470 (this can affect the format of dates).
1472 28. exim_tidydb, when checking for the continued existence of a message for
1473 which it has found a message-specific retry record, was not finding
1474 messages that were in split spool directories. Consequently, it was
1475 deleting retry records that should have stayed in existence.
1477 29. eximstats updated to version 1.35
1478 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1479 1.35 - bugfix such that pie charts by volume are generated correctly
1481 30. The SPA authentication driver was not abandoning authentication and moving
1482 on to the next authenticator when an expansion was forced to fail,
1483 contradicting the general specification for all authenticators. Instead it
1484 was generating a temporary error. It now behaves as specified.
1486 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1487 (the order specifies the preference for clients). The order is now AES256,
1488 AES128, 3DES, ARCFOUR128.
1490 31. Small patch to Sieve code - explicitly set From: when generating an
1493 32. Exim crashed if a remote delivery caused a very long error message to be
1494 recorded - for instance if somebody sent an entire SpamAssassin report back
1495 as a large number of 550 error lines. This bug was coincidentally fixed by
1496 increasing the size of one of Exim's internal buffers (big_buffer) that
1497 happened as part of the Exiscan merge. However, to be on the safe side, I
1498 have made the code more robust (and fixed the comments that describe what
1501 33. Some experimental protocols are using DNS PTR records for new purposes. The
1502 keys for these records are domain names, not reversed IP addresses. The
1503 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1504 leaves it alone. Component reversal etc. now happens only for IP addresses.
1507 34. The host_aton() function is supposed to be passed a string that is known
1508 to be a valid IP address. However, in the case of IPv6 addresses, it was
1509 not checking this. This is a hostage to fortune. Exim now panics and dies
1510 if the condition is not met. A case was found where this could be provoked
1511 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1512 components; fortuitously, this particular loophole had already been fixed
1513 by change 4.50/55 or 4.44/33 above.
1515 If there are any other similar loopholes, the new check in host_aton()
1516 itself should stop them being exploited. The report I received stated that
1517 data on the command line could provoke the exploit when Exim was running as
1518 exim, but did not say which command line option was involved. All I could
1519 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1520 running as the user.
1523 35. There was a buffer overflow vulnerability in the SPA authentication code
1524 (which came originally from the Samba project). I have added a test to the
1525 spa_base64_to_bits() function which I hope fixes it.
1528 36. The daemon start-up calls getloadavg() while still root for those OS that
1529 need the first call to be done as root, but it missed one case: when
1530 deliver_queue_load_max is set with deliver_drop_privilege. This is
1531 necessary for the benefit of the queue runner, because there is no re-exec
1532 when deliver_drop_privilege is set.
1534 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1535 was in use, was not putting the data itself into the right store pool;
1536 consequently, it could be overwritten for a subsequent message in the same
1537 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1540 38. Sometimes the final signoff response after QUIT could fail to get
1541 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1542 < 0 before doing a fflush(). This bug looks as though it goes back to the
1543 introduction of TLS in release 3.20, but "sometimes" must have been rare
1544 because the tests only now provoked it.
1550 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1551 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1552 Now it returns an integer. A number of calls were still expecting a T/F
1553 return. Fortuitously, in all cases, the tests worked in OK situations,
1554 which is the norm. However, things would have gone wrong on any write
1555 failures on the smtp file descriptor. This function is used when sending
1556 messages over SMTP and also when doing verify callouts.
1558 2. When Exim is called to do synchronous delivery of a locally submitted
1559 message (the -odf or -odi options), it no longer closes stderr before doing
1562 3. Implemented the mua_wrapper option.
1564 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1566 5. Implemented the functions header_remove(), header_testname(),
1567 header_add_at_position(), and receive_remove_recipient(), and exported them
1570 6. If an ACL "warn" statement specified the addition of headers, Exim already
1571 inserted X-ACL-Warn: at the start if there was no header name. However, it
1572 was not making this test for the second and subsequent header lines if
1573 there were newlines in the string. This meant that an invalid header could
1574 be inserted if Exim was badly configured.
1576 7. Allow an ACL "warn" statement to add header lines at the start or after all
1577 the Received: headers, as well as at the end.
1579 8. Added the rcpt_4xx retry error code.
1581 9. Added postmaster_mailfrom=xxx to callout verification option.
1583 10. Added mailfrom=xxxx to the callout verification option, for verify=
1586 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1587 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1589 12. Inserted some casts to stop certain compilers warning when using pointer
1590 differences as field lengths or precisions in printf-type calls (mostly
1591 affecting debugging statements).
1593 13. Added optional readline() support for -be (dynamically loaded).
1595 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1596 same clock tick as a message's arrival, so that its received time was the
1597 same as the "first fail" time on the retry record, and that message
1598 remained on the queue past the ultimate address timeout, every queue runner
1599 would try a delivery (because it was past the ultimate address timeout) but
1600 after another failure, the ultimate address timeout, which should have then
1601 bounced the address, did not kick in. This was a "< instead of <=" error;
1602 in most cases the first failure would have been in the next clock tick
1603 after the received time, and all would be well.
1605 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1606 being recognized when the domain list was tested by the match_domain
1607 condition in an expansion string.
1609 16. Added the ${str2b64: operator.
1611 17. Exim was always calling setrlimit() to set a large limit for the number of
1612 processes, without checking whether the existing limit was already
1613 adequate. (It did check for the limit on file descriptors.) Furthermore,
1614 errors from getrlimit() and setrlimit() were being ignored. Now they are
1615 logged to the main and panic logs, but Exim does carry on, to try to do its
1616 job under whatever limits there are.
1618 18. Imported PCRE 5.0.
1620 19. Trivial typo in log message " temporarily refused connection" (the leading
1623 20. If the log selector return_path_on_delivery was set and an address was
1624 redirected to /dev/null, the delivery process crashed because it assumed
1625 that a return path would always be set for a "successful" delivery. In this
1626 case, the whole delivery is bypassed as an optimization, and therefore no
1629 21. Internal re-arrangement: the function for sending a challenge and reading
1630 a response while authentication was assuming a zero-terminated challenge
1631 string. It's now changed to take a pointer and a length, to allow for
1632 binary data in such strings.
1634 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1636 23. Exim was not respecting finduser_retries when seeking the login of the
1637 uid under which it was called; it was always trying 10 times. (The default
1638 setting of finduser_retries is zero.) Also, it was sleeping after the final
1639 failure, which is pointless.
1641 24. Implemented tls_on_connect_ports.
1643 25. Implemented acl_smtp_predata.
1645 26. If the domain in control=submission is set empty, Exim assumes that the
1646 authenticated id is a complete email address when it generates From: or
1647 Sender: header lines.
1649 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1650 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1651 chown and chgrp in /bin and hostname in /usr/bin.
1653 28. Exim was keeping the "process log" file open after each use, just as it
1654 does for the main log. This opens the possibility of it remaining open for
1655 long periods when the USR1 signal hits a daemon. Occasional processlog
1656 errors were reported, that could have been caused by this. Anyway, it seems
1657 much more sensible not to leave this file open at all, so that is what now
1660 29. The long-running daemon process does not normally write to the log once it
1661 has entered its main loop, and it closes the log before doing so. This is
1662 so that log files can straightforwardly be renamed and moved. However,
1663 there are a couple of unusual error situations where the daemon does write
1664 log entries, and I had neglected to close the log afterwards.
1666 30. The text of an SMTP error response that was received during a remote
1667 delivery was being truncated at 512 bytes. This is too short for some of
1668 the long messages that one sometimes sees. I've increased the limit to
1671 31. It is now possible to make retry rules that apply only when a message has a
1672 specific sender, in particular, an empty sender.
1674 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1675 it possible to be selective about when SMTP synchronization is enforced.
1677 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1679 32. Implemented hosts_connection_nolog.
1681 33. Added an ACL for QUIT.
1683 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1686 35. Added mailbox_size and mailbox_filecount to appendfile.
1688 36. Added control = no_multiline_responses to ACLs.
1690 37. There was a bug in the logic of the code that waits for the clock to tick
1691 in the case where the clock went backwards by a substantial amount such
1692 that the microsecond fraction of "now" was more than the microsecond
1693 fraction of "then" (but the whole seconds number was less).
1695 38. Added support for the libradius Radius client library this is found on
1696 FreeBSD (previously only the radiusclient library was supported).
1702 1. When certain lookups returned multiple values in the form name=value, the
1703 quoting of the values was not always being done properly. Specifically:
1704 (a) If the value started with a double quote, but contained no whitespace,
1706 (b) If the value contained whitespace other than a space character (i.e.
1707 tabs or newlines or carriage returns) it was not quoted.
1708 This fix has been applied to the mysql and pgsql lookups by writing a
1709 separate quoting function and calling it from the lookup code. The fix
1710 should probably also be applied to nisplus, ibase and oracle lookups, but
1711 since I cannot test any of those, I have not disturbed their existing code.
1713 2. A hit in the callout cache for a specific address caused a log line with no
1714 reason for rejecting RCPT. Now it says "Previous (cached) callout
1715 verification failure".
1717 3. There was an off-by-one bug in the queryprogram router. An over-long
1718 return line was truncated at 256 instead of 255 characters, thereby
1719 overflowing its buffer with the terminating zero. As well as fixing this, I
1720 have increased the buffer size to 1024 (and made a note to document this).
1722 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1723 when Exim is waiting for an SMTP response from a remote server, Exim
1724 restarts its select() call on the socket, thereby resetting its timeout.
1725 This is not a problem when such interrupts are rare. Somebody set up a cron
1726 job to run exiwhat every 2 minutes, which is less than the normal select()
1727 timeout (5 or 10 minutes). This meant that the select() timeout never
1728 kicked in because it was always reset. I have fixed this by comparing the
1729 time when an interrupt arrives with the time at the start of the first call
1730 to select(). If more time than the timeout has elapsed, the interrupt is
1731 treated as a timeout.
1733 5. Some internal re-factoring in preparation for the addition of Sieve
1734 extensions (by MH). In particular, the "personal" test is moved to a
1735 separate function, and given an option for scanning Cc: and Bcc: (which is
1736 not set for Exim filters).
1738 6. When Exim created an email address using the login of the caller as the
1739 local part (e.g. when creating a From: or Sender: header line), it was not
1740 quoting the local part when it contained special characters such as @.
1742 7. Installed new OpenBSD configuration files.
1744 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1745 try to make them clearer.
1747 9. Callout options, other than the timeout value, were being ignored when
1748 verifying sender addresses in header lines. For example, when using
1750 verify = header_sender/callout=no_cache
1752 the cache was (incorrectly) being used.
1754 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1755 only the Cygwin environment.
1757 11. When return_path_on_delivery was set as a log selector, if different remote
1758 addresses in the same message used different return paths and parallel
1759 remote delivery occurred, the wrong values would sometimes be logged.
1760 (Whenever a remote delivery process finished, the return path value from
1761 the most recently started remote delivery process was logged.)
1763 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1764 lines when AUTH and/or TLS are in use. This is the "received protocol"
1765 field. Exim used to use "asmtp" for authenticated SMTP, without any
1766 indication (in the protocol name) for TLS use. Now it follows the RFC and
1767 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1768 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1769 names appear in log lines as well as in Received: header lines.
1771 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1772 extensions, and comparison tests, and to fix some bugs.
1774 14. Changes to the "personal" filter test:
1776 (1) The test was buggy in that it was just doing the equivalent of
1777 "contains" tests on header lines. For example, if a user's address was
1778 anne@some.where, the "personal" test would incorrectly be true for
1780 To: susanne@some.where
1782 This test is now done by extracting each address from the header in turn,
1783 and checking the entire address. Other tests that are part of "personal"
1784 are now done using regular expressions (for example, to check local parts
1785 of addresses in From: header lines).
1787 (2) The list of non-personal local parts in From: addresses has been
1788 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1789 taken from the Sieve specification recommendations.
1791 (3) If the message contains any header line starting with "List-" it is
1792 treated as non-personal.
1794 (4) The test for "circular" in the Subject: header line has been removed
1795 because it now seems ill-conceived.
1797 15. Minor typos in src/EDITME comments corrected.
1799 16. Installed latest exipick from John Jetmore.
1801 17. If headers_add on a router specified a text string that was too long for
1802 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1803 of string_sprintf() is now avoided.
1805 18. $message_body_size was not set (it was always zero) when running the DATA
1806 ACL and the local_scan() function.
1808 19. For the "mail" command in an Exim filter, no default was being set for
1809 the once_repeat time, causing a random time value to be used if "once" was
1810 specified. (If the value happened to be <= 0, no repeat happened.) The
1811 default is now 0s, meaning "never repeat". The "vacation" command was OK
1812 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1813 (I found it when inspecting the code).
1815 20. There is now an overall timeout for performing a callout verification. It
1816 defaults to 4 times the callout timeout, which applies to individual SMTP
1817 commands during the callout. The overall timeout applies when there is more
1818 than one host that can be tried. The timeout is checked before trying the
1819 next host. This prevents very long delays if there are a large number of
1820 hosts and all are timing out (e.g. when the network connections are timing
1821 out). The value of the overall timeout can be changed by specifying an
1822 additional sub-option for "callout", called "maxwait". For example:
1824 verify = sender/callout=5s,maxwait=20s
1826 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1827 to the end before writing, but this should make it even safer).
1829 22. Exim was forgetting that it had advertised PIPELINING for the second and
1830 subsequent messages on an SMTP connection. It was also not resetting its
1831 memory on STARTTLS and an internal HELO.
1833 23. When Exim logs an SMTP synchronization error within a session, it now
1834 records whether PIPELINING has been advertised or not.
1836 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1837 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1838 rather than long int.
1840 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1842 26. Added the never_mail option to autoreply.
1848 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1849 crash if the getsockname() call failed; this can happen if a connection is
1850 closed very soon after it is established. The problem was simply in the
1851 order in which certain operations were done, causing Exim to try to write
1852 to the SMTP stream before it had set up the file descriptor. The bug has
1853 been fixed by making things happen in the correct order.
1859 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1860 before the connection was closed, thus losing the rejection response.
1862 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1863 some early Solaris releases, but causes trouble in current releases where
1864 socklen_t is defined.
1866 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1869 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1872 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1873 that releasing the top of it at the end releases what was used for sub-
1874 expansions (unless the block got too big). However, discard this block if
1875 the first thing is a variable or header, so that we can use its block when
1876 it is dynamic (useful for very large $message_headers, for example).
1878 6. Lookups now cache *every* query, not just the most recent. A new, separate
1879 store pool is used for this. It can be recovered when all lookup caches are
1880 flushed. Lookups now release memory at the end of their result strings.
1881 This has involved some general refactoring of the lookup sources.
1883 7. Some code has been added to the store_xxx() functions to reduce the amount
1884 of flapping under certain conditions.
1886 8. log_incoming_interface used to affect only the <= reception log lines. Now
1887 it causes the local interface and port to be added to several more SMTP log
1888 lines, for example "SMTP connection from", and rejection lines.
1890 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1892 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1894 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1895 could be overwritten at the end of the current message (or the start of a
1896 new message if it was set in a HELO ACL). The value is now preserved for
1897 the duration of the SMTP connection.
1899 12. If a transport had a headers_rewrite setting, and a matching header line
1900 contained an unqualified address, that address was qualified, even if it
1901 did not match any rewriting rules. The underlying bug was that the values
1902 of the flags that permit the existence of unqualified sender and recipient
1903 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1904 non-local messages, and by -bnq for local messages) were not being
1905 preserved with the message after it was received.
1907 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1908 "next input=" as part of the text comprising the host identity instead of
1909 the correct text. The code was using the same buffer for two different
1910 strings. However, depending on which order the printing function evaluated
1911 its arguments, the bug did not always show up. Under Linux, for example, my
1912 test suite worked just fine.
1914 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1915 doesn't work with some older versions of Perl. It has been changed to "my",
1916 which in any case is probably the better facility to use.
1918 15. A really picky compiler found some instances of statements for creating
1919 error messages that either had too many or two few arguments for the format
1922 16. The size of the buffer for calls to the DNS resolver has been increased
1923 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1924 for addresses that have a lot of PTR records. This alleviates a problem; it
1925 does not fully solve it.
1927 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1928 buffer now truncates the list and logs the incident, which is the same
1929 action as happens when Exim is looking up a host name and its aliases.
1930 Previously in this situation something unpredictable would happen;
1931 sometimes it was "internal error: store_reset failed".
1933 18. If a server dropped the connection unexpectedly when an Exim client was
1934 using GnuTLS and trying to read a response, the client delivery process
1935 crashed while trying to generate an error log message.
1937 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1938 string, for example:
1940 warn message = H1: something\nH2: something
1942 the text was added as a single header line from Exim's point of view
1943 though it ended up OK in the delivered message. However, searching for the
1944 second and subsequent header lines using $h_h2: did not work. This has been
1945 fixed. Similarly, if a system filter added multiple headers in this way,
1946 the routers could not see them.
1948 20. Expanded the error message when iplsearch is called with an invalid key to
1949 suggest using net-iplsearch in a host list.
1951 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1952 ACLs are no longer actually imposed (and a message to that effect is
1955 22. If a "gecos" field in a passwd entry contained escaped characters, in
1956 particular, if it contained a \" sequence, Exim got it wrong when building
1957 a From: or a Sender: header from that name. A second bug also caused
1958 incorrect handling when an unquoted " was present following a character
1959 that needed quoting.
1961 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1962 was not being matched caselessly.
1964 24. Arranged for all hyphens in the exim.8 source to be escaped with
1967 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1968 properly. Recipient callout cache records were still being keyed to include
1969 the sender, even when use_sender was set false. This led to far more
1970 callouts that were necessary. The sender is no longer included in the key
1971 when use_sender is false.
1973 26. Added "control = submission" modifier to ACLs.
1975 27. Added the ${base62d: operator to decode base 62 numbers.
1977 28. dnsdb lookups can now access SRV records.
1979 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1980 the configuration file.
1982 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1983 (-v) mode. This makes the output for a verbose queue run more intelligible.
1985 31. Added a use_postmaster feature to recipient callouts.
1987 32. Added the $body_zerocount variable, containing the number of binary zero
1988 bytes in the message body.
1990 33. The time of last modification of the "new" subdirectory is now used as the
1991 "mailbox time last read" when there is a quota error for a maildir
1994 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1996 35. Added +ignore_unknown as a special item in host lists.
1998 36. Code for decoding IPv6 addresses in host lists is now included, even if
1999 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2000 address was recognized as an IP address, but was then not correctly decoded
2001 into binary, causing unexpected and incorrect effects when compared with
2008 1. Very minor rewording of debugging text in manualroute to say "list of
2009 hosts" instead of "hostlist".
2011 2. If verify=header_syntax was set, and a header line with an unqualified
2012 address (no domain) and a large number of spaces between the end of the
2013 name and the colon was received, the reception process suffered a buffer
2014 overflow, and (when I tested it) crashed. This was caused by some obsolete
2015 code that should have been removed. The fix is to remove it!
2017 3. When running in the test harness, delay a bit after writing a bounce
2018 message to get a bit more predictability in the log output.
2020 4. Added a call to search_tidyup() just before forking a reception process. In
2021 theory, someone could use a lookup in the expansion of smtp_accept_max_
2022 per_host which, without the tidyup, could leave open a database connection.
2024 5. Added the variables $recipient_data and $sender_data which get set from a
2025 lookup success in an ACL "recipients" or "senders" condition, or a router
2026 "senders" option, similar to $domain_data and $local_part_data.
2028 6. Moved the writing of debug_print from before to after the "senders" test
2031 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2032 problems for message scanning, either using a data ACL, or using
2033 local_scan() because the Received: header was not generated till after they
2034 were called (in order to set the time as the time of reception completion).
2035 I have revised the way this works. The header is now generated after the
2036 body is received, but before the ACL or local_scan() are called. After they
2037 are run, the timestamp in the header is updated.
2043 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2044 before starting a queue runner without re-exec. This happened only when
2045 deliver_drop_privilege was set or when the Exim user was set to root. The
2046 effect of the bug was that timeouts during subsequent deliveries caused
2047 crashes instead of being properly handled. The handler is now left at its
2048 default (and expected) setting.
2050 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2051 message, again when deliver_drop_privilege is set or Exim is run as root.
2052 The bug described in (1) was not present in this case, but the tidying up
2053 of the other signals was missing. I have made the two cases consistent.
2055 3. The ignore_target_hosts setting on a manualroute router was being ignored
2056 for hosts that were looked up using the /MX notation.
2058 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2061 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2062 operated on the sender address. After changing the $sender_address to <>
2063 for the sender address verify, Exim was re-instated it as the original
2064 (before rewriting) address, but remembering that it had rewritten it, so it
2065 wasn't rewriting it again. This bug also had the effect of breaking the
2066 sender address verification caching when the sender address was rewritten.
2068 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2069 This has been changed so that if the ip literal address matches
2070 ignore_target_hosts, the router declines.
2072 7. Added expansion conditions match_domain, match_address, and match_local_
2073 part (NOT match_host).
2075 8. The placeholder for the Received: header didn't have a length field set.
2077 9. Added code to Exim itself and to exim_lock to test for a specific race
2078 condition that could lead to file corruption when using MBX delivery. The
2079 issue is with the lockfile that is created in /tmp. If this file is removed
2080 after a process has opened it but before that process has acquired a lock,
2081 there is the potential for a second process to recreate the file and also
2082 acquire a lock. This could lead to two Exim processes writing to the file
2083 at the same time. The added code performs the same test as UW imapd; it
2084 checks after acquiring the lock that its file descriptor still refers to
2085 the same named file.
2087 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2088 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2091 11. Added the smtp_active_hostname option. If used, this will typically be made
2092 to depend on the incoming interface address. Because $interface_address is
2093 not set up until the daemon has forked a reception process, error responses
2094 that can happen earlier (such as "too many connections") no longer contain
2097 12. If an expansion in a condition on a "warn" statement fails because a lookup
2098 defers, the "warn" statement is abandoned, and the next ACL statement is
2099 processed. Previously this caused the whole ACL to be aborted.
2101 13. Added the iplsearch lookup type.
2103 14. Added ident_timeout as a log selector.
2105 15. Added tls_certificate_verified as a log selector.
2107 16. Added a global option tls_require_ciphers (compare the smtp transport
2108 option of the same name). This controls incoming TLS connections.
2110 17. I finally figured out how to make tls_require_ciphers do a similar thing
2111 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2112 before starting the TLS session.
2114 18. Tabs are now shown as \t in -bP output.
2116 19. If the log selector return_path_on_delivery was set, Exim crashed when
2117 bouncing a message because it had too many Received: header lines.
2119 20. If two routers both had headers_remove settings, and the first one included
2120 a superfluous trailing colon, the final name in the first list and the
2121 first name in the second list were incorrectly joined into one item (with a
2122 colon in the middle).
2128 1. Added -C and -D options to the exinext utility, mainly to make it easier
2129 to include in the automated testing, but these could be helpful when
2130 multiple configurations are in use.
2132 2. The exinext utility was not formatting the output nicely when there was
2133 an alternate port involved in the retry record key, nor when there was a
2134 message id as well (for retries that were specific to a specific message
2135 and a specific host). It was also confused by IPv6 addresses, because of
2136 the additional colons they contain. I have fixed the IPv4 problem, and
2137 patched it up to do a reasonable job for IPv6.
2139 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2140 delivery, the log line now contains "pipelined" if PIPELINING was used.
2142 4. An SMTP transport process used to panic and die if the bind() call to set
2143 an explicit outgoing interface failed. This has been changed; it is now
2144 treated in the same way as a connect() failure.
2146 5. A reference to $sender_host_name in the part of a conditional expansion
2147 that was being skipped was still causing a DNS lookup. This no longer
2150 6. The def: expansion condition was not recognizing references to header lines
2151 that used bh_ and bheader_.
2153 7. Added the _cache feature to named lists.
2155 8. The code for checking quota_filecount in the appendfile transport was
2156 allowing one more file than it should have been.
2158 9. For compatibility with Sendmail, the command line option
2166 and sets the incoming protocol and host name (for trusted callers). The
2167 host name and its colon can be omitted when only the protocol is to be set.
2168 Note the Exim already has two private options, -pd and -ps, that refer to
2169 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2170 "s", but I don't think that's a major issue.
2172 10. A number of refactoring changes to the code, none of which should affect
2175 (a) The number of logging options was getting close to filling up the
2176 32-bit word that was used as a bit map. I have split them into two classes:
2177 those that are passed in the argument to log_write(), and those that are
2178 only ever tested independently outside of that function. These are now in
2179 separate 32-bit words, so there is plenty of room for expansion again.
2180 There is no change in the user interface or the logging behaviour.
2182 (b) When building, for example, log lines, the code previously used a
2183 macro that called string_cat() twice, in order to add two strings. This is
2184 not really sufficiently general. Furthermore, there was one instance where
2185 it was actually wrong because one of the argument was used twice, and in
2186 one call a function was used. (As it happened, calling the function twice
2187 did not affect the overall behaviour.) The macro has been replaced by a
2188 function that can join an arbitrary number of extra strings onto a growing
2191 (c) The code for expansion conditions now uses a table and a binary chop
2192 instead of a serial search (which was left over from when there were very
2193 few conditions). Also, it now recognizes conditions like "pam" even when
2194 the relevant support is not compiled in: a suitably worded error message is
2195 given if an attempt is made to use such a condition.
2197 11. Added ${time_interval:xxxxx}.
2199 12. A bug was causing one of the ddress fields not to be passed back correctly
2200 from remote delivery subprocesses. The field in question was not being
2201 subsequently used, so this caused to problems in practice.
2203 13. Added new log selectors queue_time and deliver_time.
2205 14. Might have fixed a bug in maildirsizefile handling that threw up
2206 "unexpected character" debug warnings, and recalculated the data
2207 unnecessarily. In any case, I expanded the warning message to give more
2210 15. Added the message "Restricted characters in address" to the statements in
2211 the default ACL that block characters like @ and % in local parts.
2213 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2214 Three changes have been made:
2216 (a) There was a serious bug; a negative response to MAIL caused the whole
2217 recipient domain to be cached as invalid, thereby blocking all messages
2218 to all local parts at the same domain, from all senders. This bug has
2219 been fixed. The domain is no longer cached after a negative response to
2220 MAIL if the sender used is not empty.
2222 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2225 (c) A new callout option, "use_sender" has been added for people who want
2226 the modified behaviour.
2232 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2235 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2236 indeed breaks things for older releases.
2238 3. Added additional logging to the case where there is a problem reading data
2239 from a filter that is running in a subprocess using a pipe, in order to
2240 try to track down a specific problem.
2242 4. Testing facility fudge: when running in the test harness and attempting
2243 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2244 getting "No route to host". Convert this to a timeout.
2246 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2249 6. Some OS don't have socklen_t but use size_t instead. This affects the
2250 fifth argument of getsockopt() amongst other things. This is now
2251 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2252 can be set for individual OS. I have set it for SunOS5, OSF1, and
2253 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2254 some earlier ones do not.
2256 7. Change 4.30/15 was not doing the test caselessly.
2258 8. The standard form for an IPv6 address literal was being rejected by address
2259 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2260 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2261 this, as well as the form without the "IPv6" on the front (but only when
2262 address literals are enabled, of course).
2264 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2266 10. Exim crashed if a message with an empty sender address specified by -f
2267 encountered a router with an errors_to setting. This could be provoked only
2268 by a command such as
2272 where an empty string was supplied; "<>" did not hit this bug.
2274 11. Installed PCRE release 4.5.
2276 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2277 remained set. It is now erased.
2279 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2280 times from message ids (which are base 36 rather than the normal 62).
2282 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2283 were being counted as actual protocol errors, and logged if the log
2284 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2285 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2286 and DATA following a set of rejected RCPTs do not count as protocol errors.
2287 In other words, Exim assumes they were pipelined, though this may not
2288 actually be the case. Of course, in all cases the client gets an
2289 appropriate error code.
2291 15. If a lookup fails in an ACL condition, a message about the failure may
2292 be available; it is used if testing the ACL cannot continue, because most
2293 such messages specify what the cause of the deferral is. However, some
2294 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2295 that caused an old message to be retained and used if a later statement
2296 caused a defer, replacing the real cause of the deferral.
2298 16. If an IP address had so many PTR records that the DNS lookup buffer
2299 was not large enough to hold them, Exim could crash while trying to process
2300 the truncated data. It now detects and logs this case.
2302 17. Further to 4.21/58, another change has been made: if (and only if) the
2303 first line of a message (the first header line) ends with CRLF, a bare LF
2304 in a subsequent header line has a space inserted after it, so as not to
2305 terminate the header.
2307 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2308 unnecessarily, used atoi() instead of strtol(), and didn't check the
2309 termination when getting file sizes from file names by regex.
2311 19. Completely re-implemented the support for maildirsize files, in the light
2312 of a number of problems with the previous contributed implementation
2313 (4.30/29). In particular:
2315 . If the quota is zero, the maildirsize file is maintained, but no quota is
2318 . If the maildir directory does not exist, it is created before any attempt
2319 to write a maildirsize file.
2321 . The quota value in the file is just a cache; if the quota is changed in
2322 the transport, the new value overrides.
2324 . A regular expression is available for excluding directories from the
2327 20. The autoreply transport checks the characters in options that define the
2328 message's headers; it allows continued headers, but it was checking with
2329 isspace() after an embedded newline instead of explicitly looking for a
2332 21. If all the "regular" hosts to which an address was routed had passed their
2333 expiry times, and had not reached their retry times, the address was
2334 bounced, even if fallback hosts were defined. Now Exim should go on to try
2337 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2338 equivalent code in the SMTP transport. Some hosts send humungous responses
2339 to HELO/EHLO, more than 1024 it seems.
2341 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2342 gives compiler warnings in some environments. I've now done it "properly",
2345 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2346 (because of problems with the built-in one) was declared to return uschar *
2347 instead of char *, causing compiler failure.
2349 25. Fixed a file descriptor leak when processing alias/forward files.
2351 26. Fixed a minor format string issue in dbfn.c.
2353 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2355 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2356 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2359 29. When a sender address is verified, it is cached, to save repeating the test
2360 when there is more than one recipient in a message. However, when the
2361 verification involves a callout, it is possible for different callout
2362 options to be set for different recipients. It is too complicated to keep
2363 track of this in the cache, so now Exim always runs a verification when a
2364 callout is required, relying on the callout cache for the optimization.
2365 The overhead is duplication of the address routing, but this should not be
2368 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2369 to be verified with callout=postmaster, and the main callout worked but the
2370 postmaster check failed, the verification correctly failed. However, if a
2371 subsequent RCPT command asked for sender verification *without* the
2372 postmaster check, incorrect caching caused this verification also to fail,
2375 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2376 it was not caching the DNS options (qualify_single, search_parents) that
2377 were used when the lookup failed. A subsequent lookup with different
2378 options therefore always gave the same answer, though there were cases
2379 where it should not have. (Example: a "domains = !$mx_any" option on a
2380 dnslookup router: the "domains" option is always processed without any
2381 widening, but the router might have qualify_single set.) Now Exim uses the
2382 cached value only when the same options are set.
2384 32. Added John Jetmore's "exipick" utility to the distribution.
2386 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2387 than a timeout (e.g. a certificate is required, and is not provided), an
2388 Exim server now closes the connection immediately. Previously it waited for
2389 the client to close - but if the client is SSL, it seems that they each
2390 wait for each other, leading to a delay before one of them times out.
2392 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2393 maintained 0.8.x compatibility because I don't think many are using it, and
2394 it is clearly obsolete.
2396 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2399 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2400 client certificate was expired. A simple patch fixes this, though I don't
2401 understand the full logic of why the verify callback is called multiple
2404 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2405 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2406 which causes problems with some clients (such as the Certicom SSL Plus
2407 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2408 disables the coutermeasure allowing Eudora to connect."
2410 38. Exim was not checking that a write() to a log file succeeded. This could
2411 lead to Bad Things if a log got too big, in particular if it hit a file
2412 size limit. Exim now panics and dies if it cannot write to a log file, just
2413 as it does if it cannot open a log file.
2415 39. Modified OS/Makefile-Linux so that it now contains
2417 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2419 The two -D definitions ensure that Exim is compiled with large file
2420 support, which makes it possible to handle log files that are bigger than
2423 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2424 instance) a domain was checked against a named list that involved a lookup,
2425 causing $domain_data to be set, then another domain was checked against the
2426 same list, then the first domain was re-checked, the value of $domain_data
2427 after the final check could be wrong. In particular, if the second check
2428 failed, it could be set empty. This bug probably also applied to
2431 41. The strip_trailing_dot option was not being applied to the address given
2432 with the -f command-line option.
2434 42. The code for reading a message's header from the spool was incrementing
2435 $received_count, but never initializing it. This meant that the value was
2436 incorrect (doubled) while delivering a message in the same process in which
2437 it was received. In the most common configuration of Exim, this never
2438 happens - a fresh exec is done - but it can happen when
2439 deliver_drop_privilege is set.
2441 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2442 it now includes up to 150 characters of the unexpected data in the log
2445 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2446 and building data strings. The size of both of these buffers was 10 000
2447 bytes - far larger than anybody would *ever* want, thought I. Needless to
2448 say, somebody hit the limit. I have increased the maximum line length to
2449 20 000 and the maximum data length of concatenated lines to 100 000. I have
2450 also fixed two bugs, because there was no checking on these buffers. Tsk,
2451 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2452 buffer is too small.
2454 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2455 lsearch lookups. Now it does.
2457 46. When parsing a route_list item in a manualroute router, a fixed-length
2458 buffer was used for the list of hosts. I made this 1024 bytes long,
2459 thinking that nobody would ever have a list of hosts that long. Wrong.
2460 Somebody had a whole pile of complicated expansion conditions, and the
2461 string was silently truncated, leading to an expansion error. It turns out
2462 that it is easier to change to an unlimited length (owing to other changes
2463 that have happened since this code was originally written) than to build
2464 structure for giving a limitation error. The length of the item that
2465 expands into the list of hosts is now unlimited.
2467 47. The lsearch lookup could not handle data where the length of text line was
2468 more than 4095 characters. Such lines were truncated, leading to shortened
2469 data being returned. It should now handle lines of any length.
2471 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2472 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2475 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2476 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2477 no longer happens in comment lines. A list of replacements is now placed
2478 at the head of all of the source files, except those whose only change is
2479 to replace PERL_COMMAND in the very first #! line.
2481 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2482 messages on the queue, with a bottom-up merge sort, using code contributed
2483 by Michael Haardt. This should make operations like -bp somewhat faster on
2484 large queues. It won't affect queue runners, except when queue_run_in_order
2487 51. Installed eximstats 1.31 in the distribution.
2489 52. Added support for SRV lookups to the dnslookup router.
2491 53. If an ACL referred to $message_body or $message_body_end, the value was not
2492 reset for any messages that followed in the same SMTP session.
2494 54. The store-handling optimization for building very long strings was not
2495 differentiating between the different store pools. I don't think this
2496 actually made any difference in practice, but I've tidied it.
2498 55. While running the routers to verify a sender address, $sender_address
2499 was still set to the sender address. This is wrong, because when routing to
2500 send a bounce to the sender, it would be empty. Therefore, I have changed
2501 it so that, while verifying a sender address, $sender_address is set to <>.
2502 (There is no change to what happens when verifying a recipient address.)
2504 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2505 target A or AAAA records (if not already returned) without resetting the
2506 qualify_single or search_parents options of the DNS resolver. These are
2507 inappropriate in this case because the targets of MX and SRV records must
2508 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2509 target that, when qualified, matched something in the local domain. These
2510 two options are now turned off when doing these lookups.
2512 57. It seems that at least some releases of Reiserfs (which does not have the
2513 concept of a fixed number of inodes) returns zero and not -1 for the
2514 number of available inodes. This interacted badly with check_spool_inodes,
2515 which assumed that -1 was the "no such thing" setting. What I have done is
2516 to check that the total number of inodes is greater than zero before doing
2517 the test of how many are available.
2519 58. When a "warn" ACL statement has a log_message modifier, the message is
2520 remembered, and not repeated. This is to avoid a lot of repetition when a
2521 message has many recipients that cause the same warning to be written.
2522 Howewer, Exim was preserving the list of already written lines for an
2523 entire SMTP session, which doesn't seem right. The memory is now reset if a
2524 new message is started.
2526 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2527 debugging output unless log_rewrite was also set.
2529 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2530 of (int)(handle) when we know that handle contains (void *)(-1).
2532 61. The Exim daemon panic-logs an error return when it closes the incoming
2533 connection. However "connection reset by peer" seems to be common, and
2534 isn't really an error worthy of noting specially, so that particular error
2537 62. When Exim is trying to find all the local interfaces, it used to panic and
2538 die if the ioctl to get the interface flags failed. However, it seems that
2539 on at least one OS (Solaris 9) it is possible to have an interface that is
2540 included in the list of interfaces, but for which you get a failure error
2541 for this call. This happens when the interface is not "plumbed" into a
2542 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2543 failure of the "get flags" call assumes that the interface is down.
2545 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2546 makes life easier for people who are doing arithmetic on fields extracted
2547 from dates, where you often get leading zeros that should not be
2548 interpreted as octal.
2550 64. Added qualify_domain to the redirect router, to override the global
2553 65. If a pathologically long header line contained very many addresses (the
2554 report of this problem mentioned 10 000) and each of them was rewritten,
2555 Exim could use up a very large amount of memory. (It kept on making new
2556 copies of the header line as it rewrote, and never released the old ones.)
2557 At the expense of a bit more processing, the header rewriting function has
2558 been changed so that it no longer eats memory in this way.
2560 66. The generation of the Received: header has been moved from the time that a
2561 message starts to be received, to the time that it finishes. The timestamp
2562 in the Received: header should now be very close to that of the <= log
2563 line. There are two side-effects of this change:
2565 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2566 logged header lines no longer include the local Received: line, because
2567 it has not yet been created. The same applies to a copy of the message
2568 that is returned to a non-SMTP sender when a message is rejected.
2570 (b) When a filter file is tested using -bf, no additional Received: header
2571 is added to the test message. After some thought, I decided that this
2574 This change does not affect the value of $received_for. It is still set
2575 after address rewriting, but before local_scan() is called.
2577 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2579 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2580 gave an unhelpful panic error message, and a defer error. I have managed to
2581 change this behaviour so that it now rejects any supplied certificate,
2582 which seems right, as the list of acceptable certificates is empty.
2584 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2585 gave an unhelpful defer error. I have not managed to make this reject any
2586 supplied certificates, but the error message it gives is "no certificate
2587 supplied", which is not helpful.
2589 70. exigrep's output now also includes lines that are not associated with any
2590 message, but which match the given pattern. Implemented by a patch from
2591 Martin Sluka, which also tidied up the Perl a bit.
2593 71. Recipient callout verification, like sender verification, was using <> in
2594 the MAIL FROM command. This isn't really the right thing, since the actual
2595 sender may affect whether the remote host accepts the recipient or not. I
2596 have changed it to use the actual sender in the callout; this means that
2597 the cache record is now keyed on a recipient/sender pair, not just the
2598 recipient address. There doesn't seem to be a real danger of callout loops,
2599 since a callout by the remote host to check the sender would use <>.
2600 [SEE ABOVE: changed after hitting problems.]
2602 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2603 temporary errors. However, in the case of such a code being given after
2604 the end of a data transmission (i.e. after ".") Exim was failing to write
2605 a retry record for the message. (Yes, there was some broken host that was
2606 actually sending 8xx at this point.)
2608 73. An unknown lookup type in a host list could cause Exim to panic-die when
2609 the list was checked. (An example that provoked this was putting <; in the
2610 middle of a list instead of at the start.) If this happened during a DATA
2611 ACL check, a -D file could be left lying around. This kind of configuration
2612 error no longer causes Exim to die; instead it causes a defer errror. The
2613 incident is still logged to the main and panic logs.
2615 74. Buglet left over from Exim 3 conversion. The message "too many messages
2616 in one connection" was written to the rejectlog but not the mainlog, except
2617 when address rewriting (yes!) was being logged.
2619 75. Added write_rejectlog option.
2621 76. When a system filter was run not as root (that is, when system_filter_user
2622 was set), the values of the $n variables were not being returned to the
2623 main process; thus, they were not subsequently available in the $sn
2626 77. Added +return_path_on_delivery log selector.
2628 78. A connection timeout was being treated differently from recipients deferred
2629 when testing hosts_max_try with a message that was older than the host's
2630 retry timeout. (The host should not be counted, thus allowing all hosts to
2631 be tried at least once before bouncing.) This may have been the cause of an
2632 occasionally reported bug whereby a message would remain on the queue
2633 longer than the retry timeout, but would be bounced if a delivery was
2634 forced. I say "may" because I never totally pinned down the problem;
2635 setting up timeout/retry tests is difficult. See also the next item.
2637 79. The ultimate address timeout was not being applied to errors that involved
2638 a combination of host plus message (for example, a timeout on a MAIL
2639 command). When an address resolved to a number of possible hosts, and they
2640 were not all tried for each delivery (e.g. because of hosts_max_try), a
2641 message could remain on the queue longer than the retry timeout.
2643 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2646 81. Fixed an obscure SMTP outgoing bug which required at least the following
2647 conditions: (a) there was another message waiting for the same server;
2648 (b) the server returned 5xx to all RCPT commands in the first message so
2649 that the message was not completed; (c) the server dropped the connection
2650 or gave a negative response to the RSET that Exim sends to abort the
2651 transaction. The observed case was a dropped connection after DATA that had
2652 been sent in pipelining mode. That is, the server had advertised PIPELINING
2653 but was not implementing it correctly. The effect of the bug was incorrect
2654 behaviour, such as trying another host, and this could lead to a crash.
2660 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2661 and daemon.c were passed as pointers to ints; they should have been
2662 pointers to socklen_t variables (which are typically unsigned ints).
2664 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2667 3. Fixed a really odd bug that affected only the testing scheme; patching a
2668 certain fixed string in the binary changed the value of another string that
2669 happened to be identical to the end of the original first string.
2671 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2672 name", it returns that address as the IP address. On some operating
2673 systems (e.g. Solaris), it also passes back the IP address string as the
2674 "host name". However, on others (e.g. Linux), it passes back an empty
2675 string. Exim wasn't checking for this, and was changing the host name to an
2676 empty string, assuming it had been canonicized.
2678 5. Although rare, it is permitted to have more than one PTR record for a given
2679 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2680 all the names associated with an address, because they do in Solaris.
2681 However, it seems that they do not in Linux for data that comes from the
2682 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2683 I found this out when I moved to a new Linux workstation and tried to run
2684 the Exim test suite.
2686 To get round this problem I have changed the code so that it now does its
2687 own call to the DNS to look up PTR records when searching for a host name.
2688 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2689 addresses that are only in /etc/hosts are still found.
2691 This behaviour is, however, controlled by an option called host_lookup_
2692 order, which defaults to "bydns:byaddr". If people want to use the other
2693 order, or indeed, just use one or the other means of lookup, they can
2694 specify it in this variable.
2696 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2697 some operating systems, this comes back from gethostbyaddr() as an empty
2698 string, and this is what Exim used to test for. However, it seems that in
2699 other systems, "." is yielded. Exim now tests for this case too.
2701 7. The values of check_spool_space and check_log_space are now held internally
2702 as a number of kilobytes instead of an absolute number of bytes. If a
2703 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2704 kilobyte. This means that much larger values can be stored.
2706 8. Exim monitor: an attempt to get the action menu when not actually pointing
2707 at a message produces an empty menu entitled "No message selected". This
2708 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2709 no entries in it ("Shell widget menu has zero width and/or height"). So I
2710 have added a single, blank menu entry in this case.
2712 9. Added ${quote_local_part.
2714 10. MIME decoding is now applied to the contents of Subject: header lines when
2717 11. Now that a reference to $sender_host_address automatically causes a reverse
2718 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2719 host lookup before query-style lookups in lists that might use this
2720 variable. This has therefore been abolished, and the "net-" prefix is no
2721 longer necessary for query-style lookups.
2723 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2724 have been a typo for LFLAGS, so it has been changed.
2726 13. The install script calls Exim with "-C /dev/null" in order to find the
2727 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2728 to be output. Howeve, since Exim outputs its version number before the
2729 error, it didn't break the script. It just looked ugly. I fixed this by
2730 always allowing "-C /dev/null" if the caller is root.
2732 14. Ignore overlarge ACL variable number when reading spool file - insurance
2733 against a later release with more variables having written the file.
2735 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2736 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2737 this, as well as the form without the "IPv6" on the front.
2739 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2740 OS/Makefile-Darwin file.
2742 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2743 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2745 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2746 variable. However, some OS use ints for inodes. I've added cast to long int
2747 to get rid of the compiler warning.
2749 19. I had forgotten to lock out "/../" in configuration file names when
2750 ALT_CONFIG_PREFIX was set.
2752 20. Routers used for verification do not need to specify transports. However,
2753 if such a router generated a host list, and callout was configured, Exim
2754 crashed, because it could not find a port number from the (non-existent)
2755 transport. It now assumes port 25 in this circumstance.
2757 21. Added the -t option to exigrep.
2759 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2760 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2761 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2762 them is set, LOOKUP_LSEARCH is forced.
2764 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2766 24. Added sender and host information to the "rejected by local_scan()" log
2767 line; previously there was no indication of these.
2769 25. Added .include_if_exists.
2771 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2772 Linux. It turns out that not all file systems support this. Apparently some
2773 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2774 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2775 not supported on the file descriptor, is now ignored when Exim is trying to
2776 sync a directory. This applies only to Linux.
2778 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2780 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2781 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2784 29. Added support for maildirsize files from supplied patch (modified a bit).
2786 30. The use of :fail: followed by an empty string could lead Exim to respond to
2787 sender verification failures with (e.g.):
2789 550 Verification failed for <xxx>
2790 550 Sender verify failed
2792 where the first response line was missing the '-' that indicates it is not
2793 the final line of the response.
2795 31. The loop for finding the name of the user that called Exim had a hardwired
2796 limit of 10; it now uses the value of finduser_retries, which is used for
2797 all other user lookups.
2799 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2802 33. Exim was neglecting to zero errno before one call of strtol() when
2803 expanding a string and expecting an integer value. On some systems this
2804 resulted in spurious "integer overflow" errors. Also, it was casting the
2805 result into an int without checking.
2807 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2808 did not work. The code looks as if it has *never* worked, though it appears
2809 to have been documented since at least releast 1.62. I have made it work.
2811 35. The "timeout_DNS" error in retry rules, also documented since at least
2812 1.62, also never worked. As it isn't clear exactly what this means, and
2813 clearly it isn't a major issue, I have abolished the feature by treating it
2814 as "timeout", and writing a warning to the main and panic logs.
2816 36. The display of retry rules for -brt wasn't always showing the error code
2819 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2820 timeout_connect_A, timeout_connect_MX.
2822 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2823 to the empty sender.
2825 39. The daemon was not analysing the content of -oX till after it had closed
2826 stderr and disconnected from the controlling terminal. This meant that any
2827 syntax errors were only noted on the panic log, and the return code from
2828 the command was 0. By re-arranging the code a little, I've made the
2829 decoding happen first, so such errors now appear on stderr, and the return
2830 code is 1. However, the actual setting up of the sockets still happens in
2831 the disconnected process, so errors there are still only recorded on the
2834 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2835 connections (as happens on some IP stacks) was logged at start up time as
2836 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2837 it from "IPv6 and IPv4", which means that two separate sockets are being
2840 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2841 says whether AF_INET or AF_INET6 was passed as an argument.
2843 42. Exiwhat output was messed up when time zones were included in log
2846 43. Exiwhat now gives more information about the daemon's listening ports,
2847 and whether -tls-on-connect was used.
2849 44. The "port" option of the smtp transport is now expanded.
2851 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2852 silently ignored. Now an error message is written to the main and panic
2855 46. There's a new ACL modifier called "logwrite" which writes to a log file
2856 as soon as it is encountered.
2858 47. Added $local_user_uid and $local_user_gid at routing time.
2860 48. Exim crashed when trying to verify a sender address that was being
2863 49. Exim was recognizing only a space character after ".include". It now also
2864 recognizes a tab character.
2866 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2867 extracting the relevant information from the specification. The man page no
2868 longer contains scrambled data for the -d option, and I've added a section
2869 at the front about calling Exim under different names.
2871 51. Added "extra_headers" argument to the "mail" command in filter files.
2873 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2876 53. Installed eximstats 1.29.
2878 54. Added transport_filter_timeout as a generic transport option.
2880 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2881 Cc: header lines. This was required by RFC 822, but it not required by RFC
2884 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2885 incoming messages that did not have them. Now it does so only if the
2886 message originates locally, that is, if there is no associated remote host
2887 address. When Resent- header lines are present, this applies to the Resent-
2888 lines rather than the non-Resent- lines.
2890 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2891 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2893 58. Messages for configuration errors now include the name of the main
2894 configuration file - useful now that there may be more than one file in a
2895 list (.included file names were always shown).
2897 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2898 for those rare installations that do not start the daemon as root or run it
2899 setuid root. I've cut out the call to initgroups() if the daemon is not
2902 60. The Exim user and group can now be bound into the binary as text strings
2903 that are looked up at the start of Exim's processing.
2905 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2907 62. Added $mailstore_basename variable.
2909 63. Installed patch to sieve.c from Michael Haardt.
2911 64. When Exim failed to open the panic log after failing to open the main log,
2912 the original message it was trying to log was written to stderr and debug
2913 output, but if they were not available (the usual case in production), it
2914 was lost. Now it is written to syslog before the two lines that record the
2915 failures to open the logs.
2917 65. Users' Exim filters run in subprocesses under the user's uid. It is
2918 possible for a "deliver" command or an alias in a "personal" command to
2919 provoke an address rewrite. If logging of address rewriting is configured,
2920 this fails because the process is not running as root or exim. There may be
2921 a better way of dealing with this, but for the moment (because 4.30 needs
2922 to be released), I have disabled address rewrite logging when running a
2923 filter in a non-root, non-exim process.
2929 1. The buildconfig auxiliary program wasn't quoting the value set for
2930 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2931 not defined. This bug was masked in 4.22 by the effect that was fixed in
2934 2. Some messages that were rejected after a message id was allocated were
2935 shown as "incomplete" by exigrep. It no longer does this for messages that
2936 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2938 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2939 have allow_domain_literals set, the ID did not get logged in the <= line.
2940 Domain literals are now always recognized in Message-ID: header lines.
2942 4. The first argument for a ${extract expansion item is the key name or field
2943 number. Leading and trailing spaces in this item were not being ignored,
2944 causing some misleading effects.
2946 5. When deliver_drop_privilege was set, single queue runner processes started
2947 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2948 same command in the process it spins off) were not dropping privilege.
2950 6. When the daemon running as "exim" started a queue runner, it always
2951 re-executed Exim in the spun-off process. This is a waste of effort when
2952 deliver_drop_privilege is set. The new process now just calls the
2953 queue-runner function directly.
2959 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2962 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2963 ignored. Though the use of -oP was forcing the writing of a pid file, it
2964 was always written to the default place.
2966 3. If the message "no IP address found for host xxxx" is generated during
2967 incoming verification, it is now followed by identification of the incoming
2968 connection (so you can more easily find what provoked it).
2970 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2972 5. Added some features to "harden" Exim a bit more against certain attacks:
2974 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2975 be put in Local/Makefile. This is like the never_users runtime option,
2976 but it cannot be overridden. The default setting is "root".
2978 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2979 prefix string with which any file named in a -C command line option
2982 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2983 is retained for -C and -D only if the caller of Exim is root. Without
2984 it, the exim user may also use -C and -D and retain privilege.
2986 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2987 command line option is disabled.
2989 6. Macro names set by the -D option must start with an upper case letter, just
2990 like macro names defined in the configuration file.
2992 7. Added "dereference=" facility to LDAP.
2994 8. Two instances of the typo "uknown" in the source files are fixed.
2996 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2997 the Configure-Makefile script screwed up while processing it.
2999 10. Incorporated PCRE 4.4.
3001 11. The SMTP synchronization check was not operating right at the start of an
3002 SMTP session. For example, it could not catch a HELO sent before the client
3003 waited for the greeting. There is now a check for outstanding input at the
3004 point when the greeting is written. Because of the duplex, asynchronous
3005 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3006 way, but not yet received, when the check is performed.
3008 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3009 on TCP/IP sockets, because this apparently causes some broken clients to
3012 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3013 unchanged) from the Cygwin maintainer.
3015 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3016 was supported instead of testing for mbx. Effectively a typo.
3018 15. The spa authenticator server code was not checking that the input it
3019 received was valid base64.
3021 16. The debug output line for the "set" modifier in ACLs was not showing the
3022 name of the variable that was being set.
3024 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3026 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3027 lookup on its own. Something else was needed to trigger the lookup. For
3028 example, a match in host_lookup or the need for a host name in a host list.
3029 Now, if $sender_host_name is referenced and the host name has not yet been
3030 looked up, a lookup is performed. If the lookup fails, the variable remains
3031 empty, and $host_lookup_failed is set to "1".
3033 19. Added "eqi" as a case-independent comparison operator.
3035 20. The saslauthd authentication condition could segfault if neither service
3036 nor realm was specified.
3038 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3039 error message that was logged was misleading, and incoming SMTP
3040 connections were dropped. The message is now more accurate, and temporary
3041 errors are given to SMTP connections.
3043 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3044 (except RSET) with a 421 error, until QUIT is received. However, it was
3045 failing to send a response to QUIT.
3047 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3048 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3049 ACL is now run after the helo_try_verify_hosts code.
3051 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3052 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3053 case-independent, so other case variants are also recognized). Apparently
3054 some systems use these upper case variants.
3056 25. If more than two messages were waiting for the same host, and a transport
3057 filter was specified for the transport, Exim sent two messages over the
3058 same TCP/IP connection, and then failed with "socket operation on non-
3059 socket" when it tried to send the third.
3061 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3063 27. The extern definition of crypt16() in expand.c was not being excluded when
3064 the OS had its own crypt16() function.
3066 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3067 as a preferred synonym for return_size_limit, both as an option and as an
3070 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3072 30. Changed the default configuration ACL to relax the local part checking rule
3073 for addresses that are not in any local domains. For these addresses,
3074 slashes and pipe symbols are allowed within local parts, but the sequence
3075 /../ is explicitly forbidden.
3077 31. SPA server authentication was not clearing the challenge buffer before
3080 32. log_message in a "warn" ACL statement was writing to the reject log as
3081 well as to the main log, which contradicts the documentation and doesn't
3082 seem right (because no rejection is happening). So I have stopped it.
3084 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3085 However, I am unable to do any testing of this.
3087 34. Fixed an infelicity in the appendfile transport. When checking directories
3088 for a mailbox, to see if any needed to be created, it was accidentally
3089 using path names with one or more superfluous leading slashes; tracing
3090 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3092 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3093 contents are added to the log line that is written for every discarded
3094 recipient. (Previously a log_message setting was ignored.)
3096 36. The ${quote: operator now quotes the string if it is empty.
3098 37. The install script runs exim in order to find its version number. If for
3099 some reason other than non-existence or emptiness, which it checks, it
3100 could not run './exim', it was installing it with an empty version number,
3101 i.e. as "exim-". This error state is now caught, and the installation is
3104 38. An argument was missing from the function that creates an error message
3105 when Exim fails to connect to the socket for saslauthd authentication.
3106 This could cause Exim to crash, or give a corrupted message.
3108 39. Added isip, isip4, and isip6 to ${if conditions.
3110 40. The ACL variables $acl_xx are now saved with the message, and can be
3111 accessed later in routers, transports, and filters.
3113 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3114 strings in the file are not string-expanded.
3116 42. If a MAIL command specified a SIZE value that was too large to fit into an
3117 int variable, the check against message_size_limit failed. Such values are
3118 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3119 day this will have to be increased, but I don't think I want to be around
3120 when emails are that large.
3127 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3128 iconv() is not standard in FreeBSD.
3130 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3131 IPv6 enabled. The observed symptom was a segmentation fault on return from
3132 the function os_common_find_running_interfaces() in src/os.c.
3134 3. In the check_special_case() function in daemon.c I had used "errno" as an
3135 argument name, which causes warnings on some systems. This was basically a
3136 typo, since it was named "eno" in the comments!
3138 4. The code that waits for the clock to tick (at a resolution of some fraction
3139 of a second) so as to ensure message-id uniqueness was always waiting for
3140 at least one whole tick, when it could have waited for less. [This is
3141 almost certainly not relevant at current processor speeds, where it is
3142 unlikely to ever wait at all. But we try to future-proof.]
3144 5. The function that sleeps for a time interval that includes fractions of a
3145 second contained a race. It did not block SIGALRM between setting the
3146 timer, and suspending (a couple of lines later). If the interval was short
3147 and the sigsuspend() was delayed until after it had expired, the suspension
3148 never ended. On busy systems this could lead to processes getting stuck for
3151 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3152 process, before it forks any delivery processes. The open lookup caching
3153 mechanism meant that the open file or database connection was passed into
3154 the delivery process. The problem was that delivery processes always tidy
3155 up cached lookup data. This could cause a problem for the next delivery
3156 process started by the queue runner, because the external queue runner
3157 process does not know about the closure. So the next delivery process
3158 still has data in the lookup cache. In the case of a file lookup, there was
3159 no problem because closing a file descriptor in a subprocess doesn't affect
3160 the parent. However, if the lookup was caching a connection to a database,
3161 the connection was closed, and the second delivery process was likely to
3162 see errors such as "PGSQL: query failed: server closed the connection
3163 unexpectedly". The problem has been fixed by closing all cached lookups
3164 in a queue runner before running a delivery process.
3166 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3167 seem to have the "const" qualifier which it has on other OS. I've
3170 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3171 *of the same type* (client or server) with the same public name that an
3172 error should be diagnosed.
3174 9. When Exim looked up a host name for an IP address, but failed to find the
3175 original IP address when looking up the host name (a safety check), it
3176 output the message "<ip address> does not match any IP for NULL", which was
3177 confusing, to say the least. The bug was that the host name should have
3178 appeared instead of "NULL".
3180 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3181 user that is built into the binary, and the -C or -D options is used, root
3182 privilege is dropped before the configuration file is read. In addition,
3183 logging is switched to stderr instead of the normal log files. If the
3184 configuration then re-defines the Exim user, the unprivileged environment
3185 is probably not what is expected, so Exim logs a panic warning message (but
3188 However, if deliver_drop_privilege is set, the unprivileged state may well
3189 be exactly what is intended, so the warning has been cut out in that case,
3190 and Exim is allowed to try to write to its normal log files.
3196 1. smtp_return_error_details was not giving details for temporary sender
3197 or receiver verification errors.
3199 2. Diagnose a configuration error if two authenticators have the same public
3202 3. Exim used not to create the message log file for a message until the first
3203 delivery attempt. This could be confusing when incoming messages were held
3204 for policy or load reasons. The message log file is now created at the time
3205 the message is received, and an initial "Received" line is written to it.
3207 4. The automatically generated man page for command line options had a minor
3208 bug that caused no ill effects; however, a more serious problem was that
3209 the procedure for building the man page automatically didn't always
3210 operate. Consequently, release 4.20 contains an out-of-date version. This
3211 shouldn't happen again.
3213 5. When building Exim with embedded Perl support, the script that builds the
3214 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3215 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3217 6. The freeze_tell option was not being used for messages that were frozen on
3218 arrival, either by an ACL or by local_scan().
3220 7. Added the smtp_incomplete_transaction log selector.
3222 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3223 was accepting AUTH without a new EHLO.
3225 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3226 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3229 10. Logging of TCP/IP connections (when configured) now happens in the main
3230 daemon process instead of the child process, so that the TCP/IP connection
3231 count is more accurate (but it can never be perfect).
3233 11. The use of "drop" in a nested ACL was not being handled correctly in the
3234 outer ACL. Now, if condition failure induced by the nested "drop" causes
3235 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3236 or "require"), the connection is dropped.
3238 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3239 that yield "discard" can now be used with an "accept" or a "discard" verb,
3240 but an error is generated for any others (because I can't see a useful way
3241 to define what should happen).
3243 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3244 are now processed in the same way as lines in the Exim configuration file.
3245 In particular, continuation lines are supported.
3247 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3249 15. Added -ti meaning -t -i.
3251 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3252 domains, and warn by logging if others are found.
3254 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3255 returned from GIFCONF when Exim is trying to find the list of interfaces on
3256 a host. The code in os.c has been modified to copy each ifreq to an aligned
3257 structure in all cases.
3259 Also, in some cases, the returned ifreq's were being copied to a 'struct
3260 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3261 means the last couple of bytes of an IPv6 address could be chopped if the
3262 ifreq contained only a normal sockaddr (14 bytes storage).
3264 18. Named domain lists were not supported in the hosts_treat_as_local option.
3265 An entry such as +xxxx was not recognized, and was treated as a literal
3268 19. Ensure that header lines added by a DATA ACL are included in the reject log
3269 if the ACL subsequently rejects the message.
3271 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3272 MD5 (which is deprecated).
3274 21. When testing a filter file using -bf, Exim was writing a message when it
3275 took the sender from a "From " line in the message, but it was not doing so
3276 when it took $return_path from a Return-Path: header line. It now does.
3278 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3279 with a valid header line field name (a series of printing characters
3280 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3282 23. Changed "disc" in the source to "disk" to conform to the documentation and
3283 the book and for uniformity.
3285 24. Ignore Sendmail's -Ooption=value command line item.
3287 25. When execve() failed while trying to run a command in a pipe transport,
3288 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3289 could be confused with a return value of 69 from the command itself. This
3290 has been changed to 127, the value the shell returns if it is asked to run
3291 a non-existent command. The wording for the related log line suggests a
3292 non-existent command as the problem.
3294 26. If received_header_text expands to an empty string, do not add a Received:
3295 header line to the message. (Well, it adds a token one on the spool, but
3296 marks it "old" so that it doesn't get used or transmitted.)
3298 27. Installed eximstats 1.28 (addition of -nt option).
3300 28. There was no check for failure on the call to getsockname() in the daemon
3301 code. This can fail if there is a shortage of resources on the system, with
3302 ENOMEM, for example. A temporary error is now given on failure.
3304 29. Contrary to the C standard, it seems that in some environments, the
3305 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3306 program. Exim now does this explicitly; it affects the formatting of
3307 timestamps using strftime().
3309 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3310 complaints. I've now initialized all the variables, to avoid this.
3312 32. Header lines added by a system filter were not being "seen" during
3313 transport-time rewrites.
3315 33. The info_callback() function passed to OpenSSL is set up with type void
3316 (*)(SSL *, int, int), as described somewhere. However, when calling the
3317 function (actually a macro) that sets it up, the type void(*)() is
3318 expected. I've put in a cast to prevent warnings from picky compilers.
3320 34. If a DNS black list lookup found a CNAME record, but there were no A
3321 records associated with the domain it pointed at, Exim crashed.
3323 35. If a DNS black list lookup returned more than one A record, Exim ignored
3324 all but the first. It now scans all returned addresses if a particular IP
3325 value is being sought. In this situation, the contents of the
3326 $dnslist_value variable are a list of all the addresses, separated by a
3329 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3330 to accept a host name and all its aliases if the forward lookup for any of
3331 them yielded the IP address of the incoming connection. Now it accepts only
3332 those names whose forward lookup yields the correct IP address. Any other
3333 names are discarded. This closes a loophole whereby a rogue DNS
3334 administrator could create reverse DNS records to break through a
3335 wildcarded host restriction in an ACL.
3337 37. If a user filter or a system filter that ran in a subprocess used any of
3338 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3339 the wrong values were passed to the pipe command ($thisaddress had the
3340 value of $0, $0 had the value of $1, etc). This bug was introduced by
3341 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3343 38. Improved the line breaking for long SMTP error messages from ACLs.
3344 Previously, if there was no break point between 40 and 75 characters, Exim
3345 left the rest of the message alone. Two changes have been made: (a) I've
3346 reduced the minimum length to 35 characters; (b) if it can't find a break
3347 point between 35 and 75 characters, it looks ahead and uses the first one
3348 that it finds. This may give the occasional overlong line, but at least the
3349 remaining text gets split now.
3351 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3352 file descriptors might be low, and that setting 1000 would always raise it.
3353 It turns out that in some environments, the limit is already over 1000 and
3354 that lowering it causes trouble. So now Exim takes care not to decrease it.
3356 40. When delivering a message, the value of $return_path is set to $sender_
3357 address at the start of routing (routers may change the value). By an
3358 oversight, this default was not being set up when an address was tested by
3359 -bt or -bv, which affected the outcome if any router or filter referred to
3362 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3363 log only when "message" or "log_message" are set. However, if one of the
3364 conditions was an address verification, or a call to a nested ACL, the
3365 messages generated by the underlying test were being passed through. This
3366 no longer happens. The underlying message is available in $acl_verify_
3367 message for both "message" and "log_message" expansions, so it can be
3368 passed through if needed.
3370 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3371 new expansion $bh_ to give the encoded byte string without charset
3372 translation. Translation happens only if iconv() is available; HAVE_ICONV
3373 indicates this at build time. HEADERS_CHARSET gives the charset to
3374 translate to; headers_charset can change it in the configuration, and
3375 "headers charset" can change it in an individual filter file.
3377 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3378 that creates RFC 2047 encoded "words" labels them as that charset instead
3379 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3380 expansion operator; (ii) when Exim creates a From: line for a local
3381 message; (iii) when a header line is rewritten to include a "phrase" part.
3383 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3384 buggy, causing it to skip the first lines of messages whose message ID
3385 ended in 'D'. This would not have bitten before Exim release 4.14, because
3386 message IDs were unlikely to end in 'D' before then. The effect was to have
3387 incorrect size information for certain domains.
3389 45. #include "config.h" was missing at the start of the crypt16.c module. This
3390 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3393 46. If there was a timeout during a "random" callout check, Exim treated it as
3394 a failure of the random address, and carried on sending RSET and the real
3395 address. If the delay was just some slowness somewhere, the response to the
3396 original RCPT would be taken as a response to RSET and so on, causing
3397 mayhem of various kinds.
3399 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3400 when I implemented it. It didn't allow for the fact that some option values
3401 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3402 the right test for positive values.
3404 48. Domain names in DNS records are case-independent. Exim always looks them up
3405 in lower case. Some resolvers return domain names in exactly the case they
3406 appear in the zone file, that is, they may contain uppercase letters. Not
3407 all resolvers do this - some return always lower case. Exim was treating a
3408 change of case by a resolver as a change of domain, similar to a widening
3409 of a domain abbreviation. This triggered its re-routing code and so it was
3410 trying to route what was effectively the same domain again. This normally
3411 caused routing to fail (because the router wouldn't handle the domain
3412 twice). Now Exim checks for this case specially, and just changes the
3413 casing of the domain that it ultimately uses when it transmits the message
3416 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3419 50. If a filter generated a file delivery with a non-absolute name (possible if
3420 no home directory exists for the router), the forbid_file option was not
3423 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3424 the existing equality matching.
3426 52. Exim was using ints instead of ino_t variables in some places where it was
3427 dealing with inode numbers.
3429 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3430 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3431 called TMPDIR, and if it finds it is different, it changes its value.
3433 54. The smtp_printf() function is now made available to local_scan() so
3434 additional output lines can be written before returning. There is also an
3435 smtp_fflush() function to enable the detection of a dropped connection.
3436 The variables smtp_input and smtp_batched_input are exported to
3439 55. Changed the default runtime configuration: the message "Unknown user"
3440 has been removed from the ACL, and instead placed on the localuser router,
3441 using the cannot_route_message feature. This means that any verification
3442 failures that generate their own messages won't get overridden. Similarly,
3443 the "Unrouteable address" message that was in the ACL for unverifiable
3444 relay addresses has also been removed.
3446 56. Added hosts_avoid_esmtp to the smtp transport.
3448 57. The exicyclog script was not checking for the esoteric option
3449 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3450 will work only if exicyclog is run under the appropriate euid.
3452 58. Following a discussion on the list, the rules by which Exim recognises line
3453 endings on incoming messages have been changed. The -dropcr and drop_cr
3454 options are now no-ops, retained only for backwards compatibility. The
3455 following line terminators are recognized: LF CRLF CR. However, special
3456 processing applies to CR:
3458 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3459 nor a local message in the state where . is a terminator.
3461 (ii) If a bare CR is encountered in a header line, an extra space is added
3462 after the line terminator so as not to end the header. The reasoning
3463 behind this is that bare CRs in header lines are most likely either
3464 to be mistakes, or people trying to play silly games.
3466 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3467 was being incorrectly given as 18 bytes larger than it should have been.
3468 This is a VOB (very old bug).
3470 60. This may never have affected anything current, but just in case it has:
3471 When the local host is found other than at the start of a list of hosts,
3472 the local host, those with the same MX, and any that follow, are discarded.
3473 When the list in question was part of a longer list of hosts, the following
3474 hosts (not currently being processed) were also being discarded. This no
3475 longer happens. I'm not sure if this situation could ever has previously
3478 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3481 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3482 header. This is something that is recommended in a new Internet Draft, and
3483 is something that is documented as being done by Sendmail. There are two
3484 possible values. For messages generated by the autoreply transport, Exim
3487 Auto-Submitted: auto-replied
3489 whereas for all other generated messages (e.g. bounces) it adds
3491 Auto-Submitted: auto-generated
3493 63. The "personal" condition in filters now includes a test for the
3494 Auto-Submitted: header. If it contains the string "auto-" the message it
3495 not considered personal.
3497 64. Added rcpt_include_affixes as a generic transport option.
3499 65. Added queue_only_override (default true).
3501 66. Added the syslog_duplication option.
3503 67. If what should have been the first header line of a message consisted of
3504 a space followed by a colon, Exim was mis-interpreting it as a header line.
3505 It isn't of course - it is syntactically invalid and should therefore be
3506 treated as the start of the message body. The misbehaviour could have
3507 caused a number of strange effects, including loss of data in subsequent
3508 header lines, and spool format errors.
3510 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3511 client host had authenticated. This control can now be exercised by an ACL
3512 for more flexibility.
3514 69. By default, callouts do not happen when testing with -bh. There is now a
3515 variant, -bhc, which does actually run the callout code, including
3516 consulting and updating the callout cache.
3518 70. Added support for saslauthd authentication, courtesy of Alexander
3521 71. If statvfs() failed on the spool or log directories while checking their
3522 size for availability, Exim confusingly gave the error "space shortage".
3523 Furthermore, in debugging mode it crashed with a floating point exception.
3524 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3525 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3526 serious problem, Exim now writes to the main and panic logs when this
3527 happens, with details of the failure. It then refuses to accept the
3528 incoming message, giving the message "spool directory problem" or "log
3529 directory problem" with a 421 code for SMTP messages.
3531 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3532 0, 1, and 2 exist, because some OS complain for execs without them (see
3533 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3534 descriptors. However, the code omitted to check that the open succeeded,
3535 causing mysterious errors if for some reason the permissions on /dev/null
3536 got screwed. Now Exim writes a message to the main and panic logs, and
3537 bombs out if it can't open /dev/null.
3539 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3540 interact so that it is all more flexible. It is supposed to remain
3541 backwards compatible. Also added extra_local_interfaces.
3543 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3544 to bomb out with an assertion failure - to the client this appears as a
3545 connection drop. This problem occurs in the part of the code that was taken
3546 from the Samba project. Fortunately, the assertion is in a very simple
3547 function, so I have fixed this by reproducing the function inline in the
3548 one place where it is called, and arranging for authentication to fail
3549 instead of killing the process with assert().
3551 75. The SPA client code was not working when the server requested OEM rather
3552 than Unicode encoding.
3554 76. Added code to make require_files with a specific uid setting more usable in
3555 the case where statting the file as root fails - usually a non-root-mounted
3556 NFS file system. When this happens and the failure is EACCES, Exim now
3557 forks a subprocess and does the per-uid checking as the relevant uid.
3559 77. Added process_log_path.
3561 78. If log_file_path was not explicitly set, a setting of check_log_space or
3562 check_log_inodes was ignored.
3564 79. If a space check for the spool or log partitions fails, the incident is now
3565 logged. Of course, in the latter case the data may get lost...
3567 80. Added the %p formatting code to string_format() so that it can be used to
3568 print addresses in debug_print(). Adjusted all the address printing in the
3569 debugging in store.c to use %p rather than %d.
3571 81. There was a concern that a line of code in smtp_in.c could overflow a
3572 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3573 initially expressed, the concern was not well-founded, because trailing
3574 spaces are removed early. However, if the trailing spaces were followed by
3575 a NULL, they did not get removed, so the overflow was possible. Two fixes
3578 (a) I re-wrote the offending code in a cleaner fashion.
3579 (b) If an incoming SMTP command contains a NULL character, it is rejected
3582 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3583 runs initgroups(), so that if the Exim user is in any additional groups,
3584 they will be used during message reception.
3590 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob