1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.249 2005/10/12 11:00:34 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Two changes to the default runtime configuration:
11 (1) Move the checks for relay_from_hosts and authenticated clients from
12 after to before the (commented out) DNS black list checks.
14 (2) Add control=submission to the relay_from_hosts and authenticated
15 clients checks, on the grounds that messages accepted by these
16 statements are most likely to be submissions.
18 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
20 (1) Generate an error if the third argument for the ${prvs expansion is
23 (2) Treat a missing third argument of ${prvscheck as if it were an empty
26 (3) Reset the variables that are obtained from the first argument of
27 ${prvscheck and used in the second argument before leaving the code,
28 because their memory is reclaimed, so using them afterwards may do
31 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
32 one (it's much easier than Tom thought :-).
34 (5) Because of (4), we can now allow for the use of $prvscheck_result
35 inside the third argument.
41 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
42 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
45 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
46 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
48 PH/03 Typo: missing ".o" in src/pcre/Makefile.
50 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
51 header line, restrict the check to what is listed in RFCs 2369 and 2929.
52 Also, for "Auto-Submitted", treat anything other than "no" as
53 non-personal, in accordance with RFC 3834. (Previously it treated
54 anything starting "auto-" as non-personal.)
56 TF/01 The control=submission/name=... option had a problem with syntax
57 errors if the name included a slash character. The /name= option
58 now slurps the rest of the string, so it can include any characters
59 but it must come last in the list of options (after /sender_retain
62 PH/05 Some modifications to the interface to the fake nameserver for the new
69 TK/01 Added the "success_on_redirect" address verification option. See
70 NewStuff for rationale and an example.
72 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
74 PH/02 Patch to exigrep to allow it to work on syslog lines.
76 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
77 fread() to skip over the body file's header line, because in Cygwin the
78 header line is locked and is inaccessible.
80 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
81 co-exist for some time) to make it clear that it is the Exim ID that is
82 referenced, not the Message-ID: header line.
84 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
85 string_format() function, because snprintf() does not exist on all
88 PH/06 The use of forbid_filter_existstest now also locks out the use of the
89 ${stat: expansion item.
91 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
92 protocol synchronization error", to keep the pedants happy.
94 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
95 well as for IRIX systems, when gcc is being used. See the host.c source
98 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
100 PH/10 Named domain lists were not working if used in a queue_smtp_domains
103 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
104 transport and to the smtp transport in LMTP mode.
106 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
108 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
109 run a filter in a subprocess. This could lead to confusion in subsequent
110 lookups in the parent process. There should also be a search_tidyup() at
111 the end of the subprocess.
113 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
114 only if the host matched helo_try_verify_hosts, which caused the
115 verification to occur when the EHLO/HELO command was issued. The ACL just
116 tested the remembered result. Now, if a previous verification attempt has
117 not happened, "verify = helo" does it there and then.
119 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
121 TK/03 Fix log output including CR from clamd.
123 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
124 exist provoked a memory error which could cause a segfault.
126 PH/15 Installed PCRE 6.2
128 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
130 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
131 of the problem. Specifically, suggested +O2 rather than +O1 for the
134 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
136 PH/20 If a delivery was routed to a non-standard port by means of an SRV
137 record, the port was not correctly logged when the outgoing_port log
138 selector was set (it logged the transort's default port).
140 PH/21 Added support for host-specific ports to manualroute, queryprogram,
141 fallback_hosts, and "hosts" in the smtp transport.
143 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
144 host errors such as "Connection refused".
146 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
147 authentication with radiusclient 0.4.9:
149 - Error returned from rc_read_config was caught wrongly
150 - Username/password not passed on to radius server due to wrong length.
152 The presumption is that some radiusclient API changes for 4.51/PH/17
153 were not taken care of correctly. The code is still untested by me (my
154 Linux distribution still has 0.3.2 of radiusclient), but it was
155 contributed by a Radius user.
157 PH/24 When doing a callout, the value of $domain wasn't set correctly when
158 expanding the "port" option of the smtp transport.
160 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
161 while reading a MIME header. Thanks to Tom Hughes for a patch.
163 PH/24 Include config.h inside local_scan.h so that configuration settings are
166 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
167 that in an ACL for RCPT (for example), you can examine exactly what was
170 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
171 commands, but it was not correctly comparing the address with the actual
172 client host address. Thus, it would show the EHLO address in Received:
173 header lines when this was not necessary.
175 PH/27 Added the % operator to ${eval:}.
177 PH/28 Exim tries to create and chdir to its spool directory when it starts;
178 it should be ignoring failures (because with -C, for example, it has lost
179 privilege). It wasn't ignoring creation failures other than "already
182 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
185 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
186 because an input line was too long, either on its own, or by virtue of
187 too many continuations, the temporary file was not being removed, and the
188 return code was incorrect.
190 PH/31 Missing "BOOL" in function definition in filtertest.c.
192 PH/32 Applied Sieve patches from the maintainer.
194 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
196 PH/33 Added "verify = not_blind".
198 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
199 Local/Makefile (with some defaults set). These are used in built scripts
200 such as exicyclog, but they have never been used in the exim_install
201 script (though there are many overriding facilities there). I have
202 arranged that the exim_install script now takes note of these two
205 PH/35 Installed configuration files for Dragonfly.
207 PH/36 When a locally submitted message by a trusted user did not contain a
208 From: header, and the sender address was obtained from -f or from an SMTP
209 MAIL command, and the trusted user did not use -F to supply a sender
210 name, $originator_name was incorrectly used when constructing a From:
211 header. Furthermore, $originator_name was used for submission mode
212 messages from external hosts without From: headers in a similar way,
213 which is clearly wrong.
215 PH/37 Added control=suppress_local_fixups.
217 PH/38 When log_selector = +received_sender was set, and the addition of the
218 sender made the log line's construction buffer exactly full, or one byte
219 less than full, an overflow happened when the terminating "\n" was
222 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
223 when the result of a list match is failure because a DNS lookup failed.
225 PH/40 RM_COMMAND is now used in the building process.
227 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
228 the "build-* directories that it finds.
230 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
231 address in a domain literal was a prefix of an interface address.
233 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
234 when verifying a sender address, unless rewrite_headers is false.
236 PH/44 Wrote a long comment about why errors_to addresses are verified as
237 recipients, not senders.
239 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
240 the ratelimit ACL was added.
242 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
244 PH/46 Added extra information about PostgreSQL errors to the error string.
246 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
247 avoiding the need to install special zones in a real server. This is
248 backwards compatible; if it can't find the fake resolver, it drops back.
249 Thus, both old and new test suites can be run.
251 TF/02 Added util/ratelimit.pl
253 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
254 clock is set back in time.
256 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
257 Candler <B.Candler@pobox.com>.
259 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
260 OK for addresses that are the result of redirections.
262 PH/48 A number of further additions for the benefit of the new test suite,
263 including a fake gethostbyname() that interfaces to the fake DNS resolver
266 TF/06 The fix for widen_domains has also been applied to qualify_single and
267 search_parents which are the other dnslookup options that can cause
270 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
273 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
275 TF/07 Exim produced the error message "an SRV record indicated no SMTP
276 service" if it encountered an MX record with an empty target hostname.
277 The message is now "an MX or SRV record indicated no SMTP service".
279 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
280 if the DNS of the sending site is misconfigured. This is quite a
281 common situation. This change restores the behaviour of treating a
282 helo verification defer as a failure.
284 PH/51 If self=fail was set on a router, the bounce message did not include the
285 actual error message.
291 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
293 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
294 command itself ended in error, the underlying message about the transport
295 filter timeout was being overwritten with the pipe command error. Now the
296 underlying error message should be appended to the second error message.
298 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
300 PH/02 Reduce the amount of output that "make" produces by default. Full output
301 can still be requested.
303 PH/03 The warning log line about a condition test deferring for a "warn" verb
304 was being output only once per connection, rather than after each
305 occurrence (because it was using the same function as for successful
306 "warn" verbs). This seems wrong, so I have changed it.
308 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
309 it should not have, which might have caused a crash in the right
310 circumstances, but probably never did.
312 PH/04 Installed a modified version of Tony Finch's patch to make submission
313 mode fix the return path as well as the Sender: header line, and to
314 add a /name= option so that you can make the user's friendly name appear
317 TF/03 Added the control = fakedefer ACL modifier.
319 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
320 Mark Lowes for thorough testing.
322 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
324 TK/03 Merged latest SRS patch from Miles Wilton.
326 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
327 with the definition in sysexits.h (which is #included earlier).
328 Fortunately, Exim does not actually use EX_OK. The code used to try to
329 preserve the sysexits.h value, by assumimg that macro definitions were
330 scanned for macro replacements. I have been disabused of this notion,
331 so now the code just undefines EX_OK before #including unistd.h.
333 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
334 in the smtp transport. When a block could not be written in a single
335 write() function, the timeout was being re-applied to each part-write.
336 This seems wrong - if the receiver was accepting one byte at a time it
337 would take for ever. The timeout is now adjusted when this happens. It
338 doesn't have to be particularly precise.
340 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
341 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
343 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
344 without a domain if the check to <postmaster@domain> fails.
346 SC/01 Eximstats: added -xls and the ability to specify output files
347 (patch written by Frank Heydlauf).
349 SC/02 Eximstats: use FileHandles for outputing results.
351 SC/03 Eximstats: allow any combination of xls, txt, and html output.
353 SC/04 Eximstats: fixed display of large numbers with -nvr option
355 SC/05 Eximstats: fixed merging of reports with empty tables.
357 SC/06 Eximstats: added the -include_original_destination flag
359 SC/07 Eximstats: removed tabs and trailing whitespace.
361 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
363 TK/06 MBOX spool code: Add real "From " MBOX separator line
364 so the .eml file is really in mbox format (even though
365 most programs do not really care). Patch from Alex Miller.
367 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
368 The latter is generated from $received_to and is only set if the
369 message has one envelope recipient. SA can use these headers,
370 obviously out-of-the-box. Patch from Alex Miller.
372 PH/08 The ${def test on a variable was returning false if the variable's
373 value was "0", contrary to what the specification has always said!
374 The result should be true unless the variable is empty.
376 PH/09 The syntax error of a character other than { following "${if
377 def:variable_name" (after optional whitespace) was not being diagnosed.
378 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
379 accidental colon was present, for example, could give incorrect results.
381 PH/10 Tidied the code in a number of places where the st_size field of a stat()
382 result is used (not including appendfile, where other changes are about
385 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
386 This involved changing a lot of size variables from int to off_t. It
387 should work with maildirs and everything.
389 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
390 spamd dying while we are connected to it.
392 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
395 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
396 function for BDB 4.3.
398 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
400 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
401 the sources for PCRE have been re-arranged and more files are now
404 PH/15 The code I had for printing potentially long long variables in PH/11
405 above was not the best (it lost precision). The length of off_t variables
406 is now inspected at build time, and an appropriate printing format (%ld
407 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
408 to be "long long int" or "long int". This is needed for the internal
409 formatting function string_vformat().
411 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
412 the configuration file to be ":syslog", then the script "guesses" where
413 the logs files are, rather than using the compiled in default. In our
414 case the guess is not the same as the compiled default, so the script
415 suddenly stopped working when I started to use syslog. The patch checks
416 to see if log_file_path is "". If so, it attempts to read it from exim
417 with no configuration file to get the compiled in version, before it
418 falls back to the previous guessing code."
420 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
421 implementing BATV in an Exim configuration. See NewStuff for the gory
424 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
425 Makefile that are specific to HP-UX.
427 PH/18 If the "use_postmaster" option was set for a recipient callout together
428 with the "random" option, the postmaster address was used as the MAIL
429 FROM address for the random test, but not for the subsequent recipient
430 test. It is now used for both.
432 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
433 patch removes a few documentation additions to RFC 3028, because the
434 latest draft now contains them. It adds the new en;ascii-case comparator
435 and a new error check for 8bit text in MIME parts. Comparator and
436 require names are now matched exactly. I enabled the subaddress
437 extension, but it is not well tested yet (read: it works for me)."
439 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
440 rework some of the code of TK/09 above to avoid the hardwired use of
441 "%lld" and "long long". Replaced the call to snprintf() with a call to
444 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
445 records point to non-existent hosts", "retry timeout exceeded", and
446 "retry time not reached for any host after a long failure period".
448 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
449 experimental DomainKeys support:
451 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
452 (2) On an error such as an illegally used "control", the wrong name for
453 the control was given.
455 These problems did NOT occur unless DomainKeys support was compiled.
457 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
459 PH/24 Added ${if match_ip condition.
461 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
462 databases so that it will be absolutely obvious if a crash occurs in the
463 DB library. This is a regular occurrence (often caused by mis-matched
466 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
467 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
468 picked up on a user's system that detects such things. There doesn't seem
469 to be a gcc warning option for this - only an attribute that has to be
470 put on the function's prototype. It seems that in Fedora Core 4 they have
471 set this on a number of new functions. No doubt there will be more in due
474 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
475 specify a transport. However, if an address that was verified by such a
476 router was the subject of a callout, Exim crashed because it tried to
477 read the rcpt_include_affixes from the non-existent transport. Now it
478 just assumes that the setting of that option is false. This bug was
479 introduced by 4.51/PH/31.
481 PH/28 Changed -d+all to exclude +memory, because that information is very
482 rarely of interest, but it makes the output a lot bigger. People tend to
483 do -d+all out of habit.
485 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
486 code in os-type was giving problems when libc.so lives in lib64, like on
489 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
490 aren't the modern standard, and it seems that some systems' include files
491 don't always have them. Exim was already checking for some of the newer
492 ones like T_AAAA, and defining it itself. I've added checks for all the
493 record types that Exim uses.
495 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
496 not automatically generating a new one, as it is supposed to. This
497 prevented TLS from working. If the file did exist, but contained invalid
498 data, a new version was generated, as expected. It was only the case of a
499 non-existent file that was broken.
501 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
502 with a change in libdomainkeys > 0.64.
504 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
505 from DNS. If the selector record carries the flag, it now has
506 precedence over the domain-wide flag.
508 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
510 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
511 the use of a port name that isn't defined in /etc/services) occurred, the
512 message was deferred as in a normal delivery, and thus remained on the
513 spool, instead of being failed because of the mua_wrapper setting. This
514 is now fixed, and I tidied up some of the mua_wrapper messages at the
517 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
518 the messages in a hash of arrays rather than using individual hashes.
519 This is a bit cleaner and results in dramatic memory savings, albeit
520 at a slight CPU cost.
522 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
523 as requested by Marc Sherman.
525 SC/10 Eximstats: added histograms for user specified patterns as requested
528 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
530 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
531 fopen() in the content-scanning modules that did not already have it.
537 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
538 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
540 TK/02 Fix ACL "control" statement not being available in MIME ACL.
542 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
544 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
545 to test Sieve filters that use "vacation".
547 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
548 that changes the way the GnuTLS parameters are stored in the cache file.
549 The new format can be generated externally. For backward compatibility,
550 if the data in the cache doesn't make sense, Exim assumes it has read an
551 old-format file, and it generates new data and writes a new file. This
552 means that you can't go back to an older release without removing the
555 PH/03 A redirect router that has both "unseen" and "one_time" set does not
556 work if there are any delivery delays because "one_time" forces the
557 parent to be marked "delivered", so its unseen clone is never tried
558 again. For this reason, Exim now forbids the simultaneous setting of
561 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
562 redirected to themselves ("homonym" addresses). Read the long ChangeLog
563 entry if you want to know the details. The fix, however, neglected to
564 consider the case when local delivery batching is involved. The test for
565 "previously delivered" was not happening when checking to see if an
566 address could be batched with a previous (undelivered) one; under
567 certain circumstances this could lead to multiple deliveries to the same
570 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
571 in its include files, and this causes problems building Exim.
573 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
574 header_syntax) but Exim was just ignoring anything given after a slash.
575 In particular, this caused confusion with an attempt to use "verify =
576 reverse_host_lookup/defer_ok". An error is now given when options are
577 supplied for verify items that do not have them. (Maybe reverse_host_
578 lookup should have a defer_ok option, but that's a different point.)
580 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
581 defined by RFC 821) to 2048, because there were problems with some AUTH
582 commands, and RFC 1869 says the size should be increased for extended
583 SMTP commands that take arguments.
585 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
588 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
589 "unknown" error; now it says that the functionality isn't in the binary.
591 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
592 an address' error message when a string expansion fails (syntax or
593 whatever). Otherwise the password may appear in the log. Following change
594 PH/42 below, there is no longer a chance of it appearing in a bounce
597 PH/11 Installed exipick version 20050225.0 from John Jetmore.
599 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
600 of its addresses was ever tried. (Bugzilla bug #2.)
602 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
603 the result incorrectly in the debug output. (It correctly added a newline
604 to what was transported.)
606 TF/01 Added $received_time.
608 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
609 commented out examples of how to interface to a virus scanner and to
610 SpamAssassin. Also added commented examples of av_scanner and
611 spamd_address settings.
613 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
614 and controls are allowed in which ACLs. There were a couple of minor
615 errors. Some of the entries in the conditions table (which is a table of
616 where they are NOT allowed) were getting very unwieldy; rewrote them as a
617 negation of where the condition IS allowed.
619 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
621 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
622 header file does not have a version number, so I've had to invent a new
623 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
624 API. The code is untested by me (my Linux distribution still has 0.3.2 of
625 radiusclient), but it was contributed by a Radius user.
627 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
628 files or directories, for OpenSSL.
630 PH/19 When an Exim process that is running as root has to create an Exim log
631 file, it does so in a subprocess that runs as exim:exim so as to get the
632 ownership right at creation (otherwise, other Exim processes might see
633 the file with the wrong ownership). There was no test for failure of this
634 fork() call, which would lead to the process getting stuck as it waited
635 for a non-existent subprocess. Forks do occasionally fail when resources
636 run out. I reviewed all the other calls to fork(); they all seem to check
639 PH/20 When checking for unexpected SMTP input at connect time (before writing
640 the banner), Exim was not dealing correctly with a non-positive return
641 from the read() function. If the client had disconnected by this time,
642 the result was a log entry for a synchronization error with an empty
643 string after "input=" when read() returned zero. If read() returned -1
644 (an event I could not check), uninitialized data bytes were printed.
645 There were reports of junk text (parts of files, etc) appearing after
648 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
650 PH/22 Added support for macro redefinition, and (re)definition in between
651 driver and ACL definitions.
653 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
654 forgetting to use the resulting value; it was using the unexpanded value.
656 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
657 hadn't been configured. The fix is from Juergen Kreileder, who
658 understands it better than I do:
660 "Here's what I see happening with three configured cyrus_sasl
661 authenticators configured (plain, login, cram-md5):
663 On startup auth_cyrus_sasl_init() gets called for each of these.
664 This means three calls to sasl_listmech() without a specified mech_list.
665 => SASL tests which mechs of all available mechs actually work
666 => three warnings about OTP not working
667 => the returned list contains: plain, login, cram-md5, digest-md5, ...
669 With the patch, sasl_listmech() also gets called three times. But now
670 SASL's mech_list option is set to the server_mech specified in the the
671 authenticator. Or in other words, the answer from sasl_listmech()
672 gets limited to just the mech you're testing for (which is different
674 => the return list contains just 'plain' or 'login', 'cram-md5' or
675 nothing depending on the value of ob->server_mech.
677 I've just tested the patch: Authentication still works fine,
678 unavailable mechs specified in the exim configuration are still
679 caught, and the auth.log warnings about OTP are gone."
681 PH/25 When debugging is enabled, the contents of the command line are added
682 to the debugging output, even when log_selector=+arguments is not
685 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
686 answer is "GNU", and only if the return is "GNU/something" is the answer
689 PH/27 $acl_verify_message is now set immediately after the failure of a
690 verification in an ACL, and so is available in subsequent modifiers. In
691 particular, the message can be preserved by coding like this:
693 warn !verify = sender
694 set acl_m0 = $acl_verify_message
696 Previously, $acl_verify_message was set only while expanding "message"
697 and "log_message" when a very denied access.
699 PH/28 Modified OS/os.c-Linux with
701 -#ifndef OS_LOAD_AVERAGE
702 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
704 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
705 nomenclature these days.)
707 PH/29 Installed patch from the Sieve maintainer that adds the options
708 sieve_useraddress and sieve_subaddress to the redirect router.
710 PH/30 In these circumstances:
711 . Two addresses routed to the same list of hosts;
712 . First host does not offer TLS;
713 . First host accepts first address;
714 . First host gives temporary error to second address;
715 . Second host offers TLS and a TLS session is established;
716 . Second host accepts second address.
717 Exim incorrectly logged both deliveries with the TLS parameters (cipher
718 and peerdn, if requested) that were in fact used only for the second
721 PH/31 When doing a callout as part of verifying an address, Exim was not paying
722 attention to any local part prefix or suffix that was matched by the
723 router that accepted the address. It now behaves in the same way as it
724 does for delivery: the affixes are removed from the local part unless
725 rcpt_include_affixes is set on the transport.
727 PH/32 Add the sender address, as F=<...>, to the log line when logging a
728 timeout during the DATA phase of an incoming message.
730 PH/33 Sieve envelope tests were broken for match types other than :is. I have
731 applied a patch sanctioned by the Sieve maintainer.
733 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
734 the uid or gid is negative. A case of a negative gid caused this to be
735 noticed. The fix allows for either to be negative.
737 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
738 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
741 PH/36 The change PH/12 above was broken. Fixed it.
743 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
744 the grounds that routing the same address twice would always produce the
745 same answer. This might have been true once, but it is certainly no
746 longer true now. Routing a child address may depend on the previous
747 routing that produced that child. Some complicated redirection strategies
748 went wrong when messages had multiple recipients, and made Exim's
749 behaviour dependent on the order in which the addresses were given.
751 I have moved the duplicate checking until after the routing is complete.
752 Exim scans the addresses that are assigned to local and remote
753 transports, and removes any duplicates. This means that more work will be
754 done, as duplicates will always all be routed, but duplicates are
755 presumably rare, so I don't expect this is of any significance.
757 For deliveries to pipes, files, and autoreplies, the duplicate checking
758 still happens during the routing process, since they are not going to be
761 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
762 It corrects a timeout issue with spamd. This is Ian's comment: "The
763 background is that sometimes spamd either never reads data from a
764 connection it has accepted, or it never writes response data. The exiscan
765 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
766 blindly assumes that writes won't block so it may never time out."
768 PH/39 Allow G after quota size as well as K and M.
770 PH/40 The value set for $authenticated_id in an authenticator may not contain
771 binary zeroes or newlines because the value is written to log lines and
772 to spool files. There was no check on this. Now the value is run through
773 the string_printing() function so that such characters are converted to
774 printable escape sequences.
776 PH/41 $message_linecount is a new variable that contains the total number of
777 lines in the message. Compare $body_linecount, which is the count for the
780 PH/42 Exim no longer gives details of delivery errors for specific addresses in
781 bounce and delay warning messages, except in certain special cases, which
784 (a) An SMTP error message from a remote host;
785 (b) A message specified in a :fail: redirection;
786 (c) A message specified in a "fail" command in a system filter;
787 (d) A message specified in a FAIL return from the queryprogram router;
788 (e) A message specified by the cannot_route_message router option.
790 In these cases only, Exim does include the error details in bounce and
791 warning messages. There are also a few cases where bland messages such
792 as "unrouteable address" or "local delivery error" are given.
794 PH/43 $value is now also set for the "else" part of a ${run expansion.
796 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
797 being worked on, but at least Exim now implements the latest version to
800 PH/45 In a pipe transport, although a timeout while waiting for the pipe
801 process to complete was treated as a delivery failure, a timeout while
802 writing the message to the pipe was logged, but erroneously treated as a
803 successful delivery. Such timeouts include transport filter timeouts. For
804 consistency with the overall process timeout, these timeouts are now
805 treated as errors, giving rise to delivery failures by default. However,
806 there is now a new Boolean option for the pipe transport called
807 timeout_defer, which, if set TRUE, converts the failures into defers for
808 both kinds of timeout. A transport filter timeout is now identified in
811 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
812 systems where "make" and "gmake" are different, calling "gmake" at top
813 level broke things. I've arranged for the value of $(MAKE) to be passed
814 from the Makefile to this script so that it can call the same version of
818 A note about Exim versions 4.44 and 4.50
819 ----------------------------------------
821 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
822 changes of various kinds. As a consequence, a big documentation update was
823 needed. This delayed the release for rather longer than seemed good, especially
824 in the light of a couple of (minor) security issues. Therefore, the changes
825 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
826 release. So 4.44 and 4.50 are in effect two different branches that both start
829 I have left the 4.50 change log unchanged; it contains all the changes since
830 4.43. The change log for 4.44 is below; many of its items are identical to
831 those for 4.50. This seems to be the most sensible way to preserve the
832 historical information.
838 1. Minor wording change to the doc/README.SIEVE file.
840 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
841 computation of the current number of files was incorrect.
843 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
844 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
847 4. Give more explanation in the error message when the command for a transport
848 filter fails to execute.
850 5. There are several places where Exim runs a non-Exim command in a
851 subprocess. The SIGUSR1 signal should be disabled for these processes. This
852 was being done only for the command run by the queryprogram router. It is
853 now done for all such subprocesses. The other cases are: ${run, transport
854 filters, and the commands run by the lmtp and pipe transports.
856 6. Added CONFIGURE_GROUP build-time option.
858 7. Some older OS have a limit of 256 on the maximum number of file
859 descriptors. Exim was using setrlimit() to set 1000 as a large value
860 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
861 systems. I've change it so that if it can't get 1000, it tries for 256.
863 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
864 was an oversight, and furthermore, ever since the addition of extra
865 controls (e.g. 4.43/32), the checks on when to allow different forms of
866 "control" were broken. There should now be diagnostics for all cases when a
867 control that does not make sense is encountered.
869 9. Added the /retain_sender option to "control=submission".
871 10. $recipients is now available in the predata ACL (oversight).
873 11. Tidy the search cache before the fork to do a delivery from a message
874 received from the command line. Otherwise the child will trigger a lookup
875 failure and thereby defer the delivery if it tries to use (for example) a
876 cached ldap connection that the parent has called unbind on.
878 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
879 of $address_data from the recipient verification was clobbered by the
882 13. The value of address_data from a sender verification is now available in
883 $sender_address_data in subsequent conditions in the ACL statement.
885 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
887 15. Added a new option "connect=<time>" to callout options, to set a different
890 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
891 was its contents. (It was OK if the option was not defined at all.)
893 17. A "Completed" log line is now written for messages that are removed from
894 the spool by the -Mrm option.
896 18. New variables $sender_verify_failure and $recipient_verify_failure contain
897 information about exactly what failed.
899 19. Added -dd to debug only the daemon process.
901 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
902 handles timeouts, both on the server side and network timeouts. Renamed the
903 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
905 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
906 It is now set to "smtps".
908 22. $host_address is now set to the target address during the checking of
911 23. When checking ignore_target_hosts for an ipliteral router, no host name was
912 being passed; this would have caused $sender_host_name to have been used if
913 matching the list had actually called for a host name (not very likely,
914 since this list is usually IP addresses). A host name is now passed as
917 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
918 code that specifies a non-restarting handler (typically sigaction() in
919 modern systems) in an attempt to fix a rare and obscure crash bug.
921 25. Narrowed the window for a race in the daemon that could cause it to ignore
922 SIGCHLD signals. This is not a major problem, because they are used only to
923 wake it up if nothing else does.
925 26. A malformed maildirsize file could cause Exim to calculate negative values
926 for the mailbox size or file count. Odd effects could occur as a result.
927 The maildirsize information is now recalculated if the size or filecount
930 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
931 support for a long time. Removed HAVE_SYS_VFS_H.
933 28. Installed the latest version of exipick from John Jetmore.
935 29. In an address list, if the pattern was not a regular expression, an empty
936 subject address (from a bounce message) matched only if the pattern was an
937 empty string. Non-empty patterns were not even tested. This was the wrong
938 because it is perfectly reasonable to use an empty address as part of a
939 database query. An empty address is now tested by patterns that are
940 lookups. However, all the other forms of pattern expect the subject to
941 contain a local part and a domain, and therefore, for them, an empty
942 address still always fails if the pattern is not itself empty.
944 30. Exim went into a mad DNS loop when attempting to do a callout where the
945 host was specified on an smtp transport, and looking it up yielded more
948 31. Re-factored the code for checking spool and log partition space into a
949 function that finds that data and another that does the check. The former
950 is then used to implement four new variables: $spool_space, $log_space,
951 $spool_inodes, and $log_inodes.
953 32. The RFC2047 encoding function was originally intended for short strings
954 such as real names; it was not keeping to the 75-character limit for
955 encoded words that the RFC imposes. It now respects the limit, and
956 generates multiple encoded words if necessary. To be on the safe side, I
957 have increased the buffer size for the ${rfc2047: expansion operator from
960 33. It is now permitted to omit both strings after an "if" condition; if the
961 condition is true, the result is "true". As before, when the second string
962 is omitted, a false condition yields an empty string. This makes it less
963 cumbersome to write custom ACL and router conditions.
965 34. Failure to deliver a bounce message always caused it to be frozen, even if
966 there was an errors_to setting on the router. The errors_to setting is now
969 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
970 canonical form (fully expanded) before being placed in
971 $sender_host_address.
973 36. The table in the code that translates DNS record types into text (T_A to
974 "A" for instance) was missing entries for NS and CNAME. It is just possible
975 that this could have caused confusion if both these types were looked up
976 for the same domain, because the text type is used as part of Exim's
977 per-process caching. But the chance of anyone hitting this buglet seems
980 37. The dnsdb lookup has been extended in a number of ways.
982 (1) There is a new type, "zns", which walks up the domain tree until it
983 finds some nameserver records. It should be used with care.
985 (2) There is a new type, "mxh", which is like "mx" except that it returns
986 just the host names, not the priorities.
988 (3) It is now possible to give a list of domains (or IP addresses) to be
989 looked up. The behaviour when one of the lookups defers can be
990 controlled by a keyword.
992 (4) It is now possible to specify the separator character for use when
993 multiple records are returned.
995 38. The dnslists ACL condition has been extended: it is now possible to supply
996 a list of IP addresses and/or domains to be looked up in a particular DNS
999 39. Added log_selector=+queue_time_overall.
1001 40. When running the queue in the test harness, wait just a tad after forking a
1002 delivery process, to get repeatability of debugging output.
1004 41. Include certificate and key file names in error message when GnuTLS fails
1005 to set them up, because the GnuTLS error message doesn't include the name
1006 of the failing file when there is a problem reading it.
1008 42. Allow both -bf and -bF in the same test run.
1010 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1012 44. The "Exiscan patch" is now merged into the mainline Exim source.
1014 45. Sometimes the final signoff response after QUIT could fail to get
1015 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1016 < 0 before doing a fflush(). This bug looks as though it goes back to the
1017 introduction of TLS in release 3.20, but "sometimes" must have been rare
1018 because the tests only now provoked it.
1020 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1021 (this can affect the format of dates).
1023 47. exim_tidydb, when checking for the continued existence of a message for
1024 which it has found a message-specific retry record, was not finding
1025 messages that were in split spool directories. Consequently, it was
1026 deleting retry records that should have stayed in existence.
1028 48. Steve fixed some bugs in eximstats.
1030 49. The SPA authentication driver was not abandoning authentication and moving
1031 on to the next authenticator when an expansion was forced to fail,
1032 contradicting the general specification for all authenticators. Instead it
1033 was generating a temporary error. It now behaves as specified.
1035 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1036 (the order specifies the preference for clients). The order is now AES256,
1037 AES128, 3DES, ARCFOUR128.
1039 51. Small patch to Sieve code - explicitly set From: when generating an
1042 52. Exim crashed if a remote delivery caused a very long error message to be
1043 recorded - for instance if somebody sent an entire SpamAssassin report back
1044 as a large number of 550 error lines. This bug was coincidentally fixed by
1045 increasing the size of one of Exim's internal buffers (big_buffer) that
1046 happened as part of the Exiscan merge. However, to be on the safe side, I
1047 have made the code more robust (and fixed the comments that describe what
1050 53. Now that there can be additional text after "Completed" in log lines (if
1051 the queue_time_overall log selector is set), a one-byte patch to exigrep
1052 was needed to allow it to recognize "Completed" as not the last thing in
1055 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1056 patch that reportedly fixes this has been added. I am not expert enough to
1057 create a test for it. This is what the patch creator wrote:
1059 "I found a little strange behaviour of ldap code when working with
1060 Windows 2003 AD Domain, where users was placed in more than one
1061 Organization Units. When I tried to give exim partial DN, the exit code
1062 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1063 But simultaneously result of request was absolutely normal ldap result,
1064 so I produce this patch..."
1066 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1067 REFERENCE, so I have modified the code to exclude the patch when that macro
1070 55. Some experimental protocols are using DNS PTR records for new purposes. The
1071 keys for these records are domain names, not reversed IP addresses. The
1072 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1073 leaves it alone. Component reversal etc. now happens only for IP addresses.
1076 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1078 57. Double the size of the debug message buffer (to 2048) so that more of very
1079 long debug lines gets shown.
1081 58. The exicyclog utility now does better if the number of log files to keep
1082 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1084 59. Two changes related to the smtp_active_hostname option:
1086 (1) $smtp_active_hostname is now available as a variable.
1087 (2) The default for smtp_banner uses $smtp_active_hostname instead
1088 of $primary_hostname.
1090 60. The host_aton() function is supposed to be passed a string that is known
1091 to be a valid IP address. However, in the case of IPv6 addresses, it was
1092 not checking this. This is a hostage to fortune. Exim now panics and dies
1093 if the condition is not met. A case was found where this could be provoked
1094 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1095 components; fortuitously, this particular loophole had already been fixed
1096 by change 4.50/55 above.
1098 If there are any other similar loopholes, the new check in host_aton()
1099 itself should stop them being exploited. The report I received stated that
1100 data on the command line could provoke the exploit when Exim was running as
1101 exim, but did not say which command line option was involved. All I could
1102 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1103 running as the user.
1106 61. There was a buffer overflow vulnerability in the SPA authentication code
1107 (which came originally from the Samba project). I have added a test to the
1108 spa_base64_to_bits() function which I hope fixes it.
1111 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1112 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1114 63. The daemon start-up calls getloadavg() while still root for those OS that
1115 need the first call to be done as root, but it missed one case: when
1116 deliver_queue_load_max is set with deliver_drop_privilege. This is
1117 necessary for the benefit of the queue runner, because there is no re-exec
1118 when deliver_drop_privilege is set.
1120 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1121 This has been fixed.
1123 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1124 was in use, was not putting the data itself into the right store pool;
1125 consequently, it could be overwritten for a subsequent message in the same
1126 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1129 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1131 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1132 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1133 the function were treating the return as a boolean value, which happened to
1134 work because 0=false and not-0=true, but is not correct code.
1136 68. The host_aton() function was not handling scoped IPv6 addresses (those
1137 with, for example, "%eth0" on the end) correctly.
1139 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1140 negated items (that is, ~something) in unsigned ints. Some compilers
1141 apparently mutter when there is no cast.
1143 70. If an address verification called from an ACL failed, and did not produce a
1144 user-specific message (i.e. there was only a "system" message), nothing was
1145 put in $acl_verify_message. In this situation, it now puts the system
1148 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1149 session; change 4.31/43 added the unwanted input to the log line - except
1150 that it did not do this in the start of session case. It now does.
1152 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1153 This is wrong and can cause the other end to generate a synchronization
1154 error if it is another Exim or anything else that does the synchronization
1155 check. A QUIT command is no longer sent after a timeout.
1157 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1158 during host lookups.
1160 74. The defer_ok option of callout verification was not working if it was used
1161 when verifying addresses in header lines, that is, for this case:
1163 verify = header_sender/callout=defer_ok
1165 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1166 those file descriptors could be used for SMTP connections. If anything
1167 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1168 could be sent to the SMTP client, causing chaos. The daemon now opens
1169 stdin, stdout, and stderr to /dev/null when it puts itself into the
1172 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1173 log by default. The user can override this with suitable Perl magic.
1175 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1176 the log message when discard triggers, was not working for the DATA ACL or
1177 for the non-SMTP ACL.
1179 78. Error message wording change in sieve.c.
1181 79. If smtp_accept_max_per_host was set, the number of connections could be
1182 restricted to fewer than expected, because the daemon was trying to set up
1183 a new connection before checking whether the processes handling previous
1184 connections had finished. The check for completed processes is now done
1185 earlier. On busy systems, this bug wouldn't be noticed because something
1186 else would have woken the daemon, and it would have reaped the completed
1189 80. If a message was submitted locally by a user whose login name contained one
1190 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1191 It caused a spool format error. I have fixed the spool reading code. A
1192 related problem was that the "from" clause in the Received: line became
1193 illegal because of the space(s). It is now covered by ${quote_local_part.
1195 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1198 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1200 83. Patch from Sieve maintainer to fix unterminated string problem in
1201 "vacation" handling.
1203 84. Some minor changes to the Linux configuration files to help with other
1204 OS variants using glibc.
1206 85. One more patch for Sieve to update vacation handling to latest spec.
1209 ----------------------------------------------------
1210 See the note above about the 4.44 and 4.50 releases.
1211 ----------------------------------------------------
1217 1. Change 4.43/35 introduced a bug that caused file counts to be
1218 incorrectly computed when quota_filecount was set in an appendfile
1221 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1222 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1225 3. Give more explanation in the error message when the command for a transport
1226 filter fails to execute.
1228 4. There are several places where Exim runs a non-Exim command in a
1229 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1230 was being done only for the command run by the queryprogram router. It is
1231 now done for all such subprocesses. The other cases are: ${run, transport
1232 filters, and the commands run by the lmtp and pipe transports.
1234 5. Some older OS have a limit of 256 on the maximum number of file
1235 descriptors. Exim was using setrlimit() to set 1000 as a large value
1236 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1237 systems. I've change it so that if it can't get 1000, it tries for 256.
1239 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1240 was an oversight, and furthermore, ever since the addition of extra
1241 controls (e.g. 4.43/32), the checks on when to allow different forms of
1242 "control" were broken. There should now be diagnostics for all cases when a
1243 control that does not make sense is encountered.
1245 7. $recipients is now available in the predata ACL (oversight).
1247 8. Tidy the search cache before the fork to do a delivery from a message
1248 received from the command line. Otherwise the child will trigger a lookup
1249 failure and thereby defer the delivery if it tries to use (for example) a
1250 cached ldap connection that the parent has called unbind on.
1252 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1253 of $address_data from the recipient verification was clobbered by the
1254 sender verification.
1256 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1257 was its contents. (It was OK if the option was not defined at all.)
1259 11. A "Completed" log line is now written for messages that are removed from
1260 the spool by the -Mrm option.
1262 12. $host_address is now set to the target address during the checking of
1263 ignore_target_hosts.
1265 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1266 being passed; this would have caused $sender_host_name to have been used if
1267 matching the list had actually called for a host name (not very likely,
1268 since this list is usually IP addresses). A host name is now passed as
1271 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1272 code that specifies a non-restarting handler (typically sigaction() in
1273 modern systems) in an attempt to fix a rare and obscure crash bug.
1275 15. Narrowed the window for a race in the daemon that could cause it to ignore
1276 SIGCHLD signals. This is not a major problem, because they are used only to
1277 wake it up if nothing else does.
1279 16. A malformed maildirsize file could cause Exim to calculate negative values
1280 for the mailbox size or file count. Odd effects could occur as a result.
1281 The maildirsize information is now recalculated if the size or filecount
1284 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1285 support for a long time. Removed HAVE_SYS_VFS_H.
1287 18. Updated exipick to current release from John Jetmore.
1289 19. Allow an empty sender to be matched against a lookup in an address list.
1290 Previously the only cases considered were a regular expression, or an
1293 20. Exim went into a mad DNS lookup loop when doing a callout where the
1294 host was specified on the transport, if the DNS lookup yielded more than
1297 21. The RFC2047 encoding function was originally intended for short strings
1298 such as real names; it was not keeping to the 75-character limit for
1299 encoded words that the RFC imposes. It now respects the limit, and
1300 generates multiple encoded words if necessary. To be on the safe side, I
1301 have increased the buffer size for the ${rfc2047: expansion operator from
1304 22. Failure to deliver a bounce message always caused it to be frozen, even if
1305 there was an errors_to setting on the router. The errors_to setting is now
1308 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1309 canonical form (fully expanded) before being placed in
1310 $sender_host_address.
1312 24. Updated eximstats to version 1.33
1314 25. Include certificate and key file names in error message when GnuTLS fails
1315 to set them up, because the GnuTLS error message doesn't include the name
1316 of the failing file when there is a problem reading it.
1318 26. Expand error message when OpenSSL has problems setting up cert/key files.
1321 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1322 (this can affect the format of dates).
1324 28. exim_tidydb, when checking for the continued existence of a message for
1325 which it has found a message-specific retry record, was not finding
1326 messages that were in split spool directories. Consequently, it was
1327 deleting retry records that should have stayed in existence.
1329 29. eximstats updated to version 1.35
1330 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1331 1.35 - bugfix such that pie charts by volume are generated correctly
1333 30. The SPA authentication driver was not abandoning authentication and moving
1334 on to the next authenticator when an expansion was forced to fail,
1335 contradicting the general specification for all authenticators. Instead it
1336 was generating a temporary error. It now behaves as specified.
1338 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1339 (the order specifies the preference for clients). The order is now AES256,
1340 AES128, 3DES, ARCFOUR128.
1342 31. Small patch to Sieve code - explicitly set From: when generating an
1345 32. Exim crashed if a remote delivery caused a very long error message to be
1346 recorded - for instance if somebody sent an entire SpamAssassin report back
1347 as a large number of 550 error lines. This bug was coincidentally fixed by
1348 increasing the size of one of Exim's internal buffers (big_buffer) that
1349 happened as part of the Exiscan merge. However, to be on the safe side, I
1350 have made the code more robust (and fixed the comments that describe what
1353 33. Some experimental protocols are using DNS PTR records for new purposes. The
1354 keys for these records are domain names, not reversed IP addresses. The
1355 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1356 leaves it alone. Component reversal etc. now happens only for IP addresses.
1359 34. The host_aton() function is supposed to be passed a string that is known
1360 to be a valid IP address. However, in the case of IPv6 addresses, it was
1361 not checking this. This is a hostage to fortune. Exim now panics and dies
1362 if the condition is not met. A case was found where this could be provoked
1363 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1364 components; fortuitously, this particular loophole had already been fixed
1365 by change 4.50/55 or 4.44/33 above.
1367 If there are any other similar loopholes, the new check in host_aton()
1368 itself should stop them being exploited. The report I received stated that
1369 data on the command line could provoke the exploit when Exim was running as
1370 exim, but did not say which command line option was involved. All I could
1371 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1372 running as the user.
1375 35. There was a buffer overflow vulnerability in the SPA authentication code
1376 (which came originally from the Samba project). I have added a test to the
1377 spa_base64_to_bits() function which I hope fixes it.
1380 36. The daemon start-up calls getloadavg() while still root for those OS that
1381 need the first call to be done as root, but it missed one case: when
1382 deliver_queue_load_max is set with deliver_drop_privilege. This is
1383 necessary for the benefit of the queue runner, because there is no re-exec
1384 when deliver_drop_privilege is set.
1386 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1387 was in use, was not putting the data itself into the right store pool;
1388 consequently, it could be overwritten for a subsequent message in the same
1389 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1392 38. Sometimes the final signoff response after QUIT could fail to get
1393 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1394 < 0 before doing a fflush(). This bug looks as though it goes back to the
1395 introduction of TLS in release 3.20, but "sometimes" must have been rare
1396 because the tests only now provoked it.
1402 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1403 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1404 Now it returns an integer. A number of calls were still expecting a T/F
1405 return. Fortuitously, in all cases, the tests worked in OK situations,
1406 which is the norm. However, things would have gone wrong on any write
1407 failures on the smtp file descriptor. This function is used when sending
1408 messages over SMTP and also when doing verify callouts.
1410 2. When Exim is called to do synchronous delivery of a locally submitted
1411 message (the -odf or -odi options), it no longer closes stderr before doing
1414 3. Implemented the mua_wrapper option.
1416 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1418 5. Implemented the functions header_remove(), header_testname(),
1419 header_add_at_position(), and receive_remove_recipient(), and exported them
1422 6. If an ACL "warn" statement specified the addition of headers, Exim already
1423 inserted X-ACL-Warn: at the start if there was no header name. However, it
1424 was not making this test for the second and subsequent header lines if
1425 there were newlines in the string. This meant that an invalid header could
1426 be inserted if Exim was badly configured.
1428 7. Allow an ACL "warn" statement to add header lines at the start or after all
1429 the Received: headers, as well as at the end.
1431 8. Added the rcpt_4xx retry error code.
1433 9. Added postmaster_mailfrom=xxx to callout verification option.
1435 10. Added mailfrom=xxxx to the callout verification option, for verify=
1438 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1439 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1441 12. Inserted some casts to stop certain compilers warning when using pointer
1442 differences as field lengths or precisions in printf-type calls (mostly
1443 affecting debugging statements).
1445 13. Added optional readline() support for -be (dynamically loaded).
1447 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1448 same clock tick as a message's arrival, so that its received time was the
1449 same as the "first fail" time on the retry record, and that message
1450 remained on the queue past the ultimate address timeout, every queue runner
1451 would try a delivery (because it was past the ultimate address timeout) but
1452 after another failure, the ultimate address timeout, which should have then
1453 bounced the address, did not kick in. This was a "< instead of <=" error;
1454 in most cases the first failure would have been in the next clock tick
1455 after the received time, and all would be well.
1457 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1458 being recognized when the domain list was tested by the match_domain
1459 condition in an expansion string.
1461 16. Added the ${str2b64: operator.
1463 17. Exim was always calling setrlimit() to set a large limit for the number of
1464 processes, without checking whether the existing limit was already
1465 adequate. (It did check for the limit on file descriptors.) Furthermore,
1466 errors from getrlimit() and setrlimit() were being ignored. Now they are
1467 logged to the main and panic logs, but Exim does carry on, to try to do its
1468 job under whatever limits there are.
1470 18. Imported PCRE 5.0.
1472 19. Trivial typo in log message " temporarily refused connection" (the leading
1475 20. If the log selector return_path_on_delivery was set and an address was
1476 redirected to /dev/null, the delivery process crashed because it assumed
1477 that a return path would always be set for a "successful" delivery. In this
1478 case, the whole delivery is bypassed as an optimization, and therefore no
1481 21. Internal re-arrangement: the function for sending a challenge and reading
1482 a response while authentication was assuming a zero-terminated challenge
1483 string. It's now changed to take a pointer and a length, to allow for
1484 binary data in such strings.
1486 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1488 23. Exim was not respecting finduser_retries when seeking the login of the
1489 uid under which it was called; it was always trying 10 times. (The default
1490 setting of finduser_retries is zero.) Also, it was sleeping after the final
1491 failure, which is pointless.
1493 24. Implemented tls_on_connect_ports.
1495 25. Implemented acl_smtp_predata.
1497 26. If the domain in control=submission is set empty, Exim assumes that the
1498 authenticated id is a complete email address when it generates From: or
1499 Sender: header lines.
1501 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1502 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1503 chown and chgrp in /bin and hostname in /usr/bin.
1505 28. Exim was keeping the "process log" file open after each use, just as it
1506 does for the main log. This opens the possibility of it remaining open for
1507 long periods when the USR1 signal hits a daemon. Occasional processlog
1508 errors were reported, that could have been caused by this. Anyway, it seems
1509 much more sensible not to leave this file open at all, so that is what now
1512 29. The long-running daemon process does not normally write to the log once it
1513 has entered its main loop, and it closes the log before doing so. This is
1514 so that log files can straightforwardly be renamed and moved. However,
1515 there are a couple of unusual error situations where the daemon does write
1516 log entries, and I had neglected to close the log afterwards.
1518 30. The text of an SMTP error response that was received during a remote
1519 delivery was being truncated at 512 bytes. This is too short for some of
1520 the long messages that one sometimes sees. I've increased the limit to
1523 31. It is now possible to make retry rules that apply only when a message has a
1524 specific sender, in particular, an empty sender.
1526 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1527 it possible to be selective about when SMTP synchronization is enforced.
1529 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1531 32. Implemented hosts_connection_nolog.
1533 33. Added an ACL for QUIT.
1535 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1538 35. Added mailbox_size and mailbox_filecount to appendfile.
1540 36. Added control = no_multiline_responses to ACLs.
1542 37. There was a bug in the logic of the code that waits for the clock to tick
1543 in the case where the clock went backwards by a substantial amount such
1544 that the microsecond fraction of "now" was more than the microsecond
1545 fraction of "then" (but the whole seconds number was less).
1547 38. Added support for the libradius Radius client library this is found on
1548 FreeBSD (previously only the radiusclient library was supported).
1554 1. When certain lookups returned multiple values in the form name=value, the
1555 quoting of the values was not always being done properly. Specifically:
1556 (a) If the value started with a double quote, but contained no whitespace,
1558 (b) If the value contained whitespace other than a space character (i.e.
1559 tabs or newlines or carriage returns) it was not quoted.
1560 This fix has been applied to the mysql and pgsql lookups by writing a
1561 separate quoting function and calling it from the lookup code. The fix
1562 should probably also be applied to nisplus, ibase and oracle lookups, but
1563 since I cannot test any of those, I have not disturbed their existing code.
1565 2. A hit in the callout cache for a specific address caused a log line with no
1566 reason for rejecting RCPT. Now it says "Previous (cached) callout
1567 verification failure".
1569 3. There was an off-by-one bug in the queryprogram router. An over-long
1570 return line was truncated at 256 instead of 255 characters, thereby
1571 overflowing its buffer with the terminating zero. As well as fixing this, I
1572 have increased the buffer size to 1024 (and made a note to document this).
1574 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1575 when Exim is waiting for an SMTP response from a remote server, Exim
1576 restarts its select() call on the socket, thereby resetting its timeout.
1577 This is not a problem when such interrupts are rare. Somebody set up a cron
1578 job to run exiwhat every 2 minutes, which is less than the normal select()
1579 timeout (5 or 10 minutes). This meant that the select() timeout never
1580 kicked in because it was always reset. I have fixed this by comparing the
1581 time when an interrupt arrives with the time at the start of the first call
1582 to select(). If more time than the timeout has elapsed, the interrupt is
1583 treated as a timeout.
1585 5. Some internal re-factoring in preparation for the addition of Sieve
1586 extensions (by MH). In particular, the "personal" test is moved to a
1587 separate function, and given an option for scanning Cc: and Bcc: (which is
1588 not set for Exim filters).
1590 6. When Exim created an email address using the login of the caller as the
1591 local part (e.g. when creating a From: or Sender: header line), it was not
1592 quoting the local part when it contained special characters such as @.
1594 7. Installed new OpenBSD configuration files.
1596 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1597 try to make them clearer.
1599 9. Callout options, other than the timeout value, were being ignored when
1600 verifying sender addresses in header lines. For example, when using
1602 verify = header_sender/callout=no_cache
1604 the cache was (incorrectly) being used.
1606 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1607 only the Cygwin environment.
1609 11. When return_path_on_delivery was set as a log selector, if different remote
1610 addresses in the same message used different return paths and parallel
1611 remote delivery occurred, the wrong values would sometimes be logged.
1612 (Whenever a remote delivery process finished, the return path value from
1613 the most recently started remote delivery process was logged.)
1615 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1616 lines when AUTH and/or TLS are in use. This is the "received protocol"
1617 field. Exim used to use "asmtp" for authenticated SMTP, without any
1618 indication (in the protocol name) for TLS use. Now it follows the RFC and
1619 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1620 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1621 names appear in log lines as well as in Received: header lines.
1623 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1624 extensions, and comparison tests, and to fix some bugs.
1626 14. Changes to the "personal" filter test:
1628 (1) The test was buggy in that it was just doing the equivalent of
1629 "contains" tests on header lines. For example, if a user's address was
1630 anne@some.where, the "personal" test would incorrectly be true for
1632 To: susanne@some.where
1634 This test is now done by extracting each address from the header in turn,
1635 and checking the entire address. Other tests that are part of "personal"
1636 are now done using regular expressions (for example, to check local parts
1637 of addresses in From: header lines).
1639 (2) The list of non-personal local parts in From: addresses has been
1640 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1641 taken from the Sieve specification recommendations.
1643 (3) If the message contains any header line starting with "List-" it is
1644 treated as non-personal.
1646 (4) The test for "circular" in the Subject: header line has been removed
1647 because it now seems ill-conceived.
1649 15. Minor typos in src/EDITME comments corrected.
1651 16. Installed latest exipick from John Jetmore.
1653 17. If headers_add on a router specified a text string that was too long for
1654 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1655 of string_sprintf() is now avoided.
1657 18. $message_body_size was not set (it was always zero) when running the DATA
1658 ACL and the local_scan() function.
1660 19. For the "mail" command in an Exim filter, no default was being set for
1661 the once_repeat time, causing a random time value to be used if "once" was
1662 specified. (If the value happened to be <= 0, no repeat happened.) The
1663 default is now 0s, meaning "never repeat". The "vacation" command was OK
1664 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1665 (I found it when inspecting the code).
1667 20. There is now an overall timeout for performing a callout verification. It
1668 defaults to 4 times the callout timeout, which applies to individual SMTP
1669 commands during the callout. The overall timeout applies when there is more
1670 than one host that can be tried. The timeout is checked before trying the
1671 next host. This prevents very long delays if there are a large number of
1672 hosts and all are timing out (e.g. when the network connections are timing
1673 out). The value of the overall timeout can be changed by specifying an
1674 additional sub-option for "callout", called "maxwait". For example:
1676 verify = sender/callout=5s,maxwait=20s
1678 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1679 to the end before writing, but this should make it even safer).
1681 22. Exim was forgetting that it had advertised PIPELINING for the second and
1682 subsequent messages on an SMTP connection. It was also not resetting its
1683 memory on STARTTLS and an internal HELO.
1685 23. When Exim logs an SMTP synchronization error within a session, it now
1686 records whether PIPELINING has been advertised or not.
1688 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1689 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1690 rather than long int.
1692 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1694 26. Added the never_mail option to autoreply.
1700 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1701 crash if the getsockname() call failed; this can happen if a connection is
1702 closed very soon after it is established. The problem was simply in the
1703 order in which certain operations were done, causing Exim to try to write
1704 to the SMTP stream before it had set up the file descriptor. The bug has
1705 been fixed by making things happen in the correct order.
1711 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1712 before the connection was closed, thus losing the rejection response.
1714 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1715 some early Solaris releases, but causes trouble in current releases where
1716 socklen_t is defined.
1718 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1721 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1724 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1725 that releasing the top of it at the end releases what was used for sub-
1726 expansions (unless the block got too big). However, discard this block if
1727 the first thing is a variable or header, so that we can use its block when
1728 it is dynamic (useful for very large $message_headers, for example).
1730 6. Lookups now cache *every* query, not just the most recent. A new, separate
1731 store pool is used for this. It can be recovered when all lookup caches are
1732 flushed. Lookups now release memory at the end of their result strings.
1733 This has involved some general refactoring of the lookup sources.
1735 7. Some code has been added to the store_xxx() functions to reduce the amount
1736 of flapping under certain conditions.
1738 8. log_incoming_interface used to affect only the <= reception log lines. Now
1739 it causes the local interface and port to be added to several more SMTP log
1740 lines, for example "SMTP connection from", and rejection lines.
1742 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1744 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1746 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1747 could be overwritten at the end of the current message (or the start of a
1748 new message if it was set in a HELO ACL). The value is now preserved for
1749 the duration of the SMTP connection.
1751 12. If a transport had a headers_rewrite setting, and a matching header line
1752 contained an unqualified address, that address was qualified, even if it
1753 did not match any rewriting rules. The underlying bug was that the values
1754 of the flags that permit the existence of unqualified sender and recipient
1755 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1756 non-local messages, and by -bnq for local messages) were not being
1757 preserved with the message after it was received.
1759 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1760 "next input=" as part of the text comprising the host identity instead of
1761 the correct text. The code was using the same buffer for two different
1762 strings. However, depending on which order the printing function evaluated
1763 its arguments, the bug did not always show up. Under Linux, for example, my
1764 test suite worked just fine.
1766 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1767 doesn't work with some older versions of Perl. It has been changed to "my",
1768 which in any case is probably the better facility to use.
1770 15. A really picky compiler found some instances of statements for creating
1771 error messages that either had too many or two few arguments for the format
1774 16. The size of the buffer for calls to the DNS resolver has been increased
1775 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1776 for addresses that have a lot of PTR records. This alleviates a problem; it
1777 does not fully solve it.
1779 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1780 buffer now truncates the list and logs the incident, which is the same
1781 action as happens when Exim is looking up a host name and its aliases.
1782 Previously in this situation something unpredictable would happen;
1783 sometimes it was "internal error: store_reset failed".
1785 18. If a server dropped the connection unexpectedly when an Exim client was
1786 using GnuTLS and trying to read a response, the client delivery process
1787 crashed while trying to generate an error log message.
1789 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1790 string, for example:
1792 warn message = H1: something\nH2: something
1794 the text was added as a single header line from Exim's point of view
1795 though it ended up OK in the delivered message. However, searching for the
1796 second and subsequent header lines using $h_h2: did not work. This has been
1797 fixed. Similarly, if a system filter added multiple headers in this way,
1798 the routers could not see them.
1800 20. Expanded the error message when iplsearch is called with an invalid key to
1801 suggest using net-iplsearch in a host list.
1803 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1804 ACLs are no longer actually imposed (and a message to that effect is
1807 22. If a "gecos" field in a passwd entry contained escaped characters, in
1808 particular, if it contained a \" sequence, Exim got it wrong when building
1809 a From: or a Sender: header from that name. A second bug also caused
1810 incorrect handling when an unquoted " was present following a character
1811 that needed quoting.
1813 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1814 was not being matched caselessly.
1816 24. Arranged for all hyphens in the exim.8 source to be escaped with
1819 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1820 properly. Recipient callout cache records were still being keyed to include
1821 the sender, even when use_sender was set false. This led to far more
1822 callouts that were necessary. The sender is no longer included in the key
1823 when use_sender is false.
1825 26. Added "control = submission" modifier to ACLs.
1827 27. Added the ${base62d: operator to decode base 62 numbers.
1829 28. dnsdb lookups can now access SRV records.
1831 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1832 the configuration file.
1834 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1835 (-v) mode. This makes the output for a verbose queue run more intelligible.
1837 31. Added a use_postmaster feature to recipient callouts.
1839 32. Added the $body_zerocount variable, containing the number of binary zero
1840 bytes in the message body.
1842 33. The time of last modification of the "new" subdirectory is now used as the
1843 "mailbox time last read" when there is a quota error for a maildir
1846 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1848 35. Added +ignore_unknown as a special item in host lists.
1850 36. Code for decoding IPv6 addresses in host lists is now included, even if
1851 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1852 address was recognized as an IP address, but was then not correctly decoded
1853 into binary, causing unexpected and incorrect effects when compared with
1860 1. Very minor rewording of debugging text in manualroute to say "list of
1861 hosts" instead of "hostlist".
1863 2. If verify=header_syntax was set, and a header line with an unqualified
1864 address (no domain) and a large number of spaces between the end of the
1865 name and the colon was received, the reception process suffered a buffer
1866 overflow, and (when I tested it) crashed. This was caused by some obsolete
1867 code that should have been removed. The fix is to remove it!
1869 3. When running in the test harness, delay a bit after writing a bounce
1870 message to get a bit more predictability in the log output.
1872 4. Added a call to search_tidyup() just before forking a reception process. In
1873 theory, someone could use a lookup in the expansion of smtp_accept_max_
1874 per_host which, without the tidyup, could leave open a database connection.
1876 5. Added the variables $recipient_data and $sender_data which get set from a
1877 lookup success in an ACL "recipients" or "senders" condition, or a router
1878 "senders" option, similar to $domain_data and $local_part_data.
1880 6. Moved the writing of debug_print from before to after the "senders" test
1883 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1884 problems for message scanning, either using a data ACL, or using
1885 local_scan() because the Received: header was not generated till after they
1886 were called (in order to set the time as the time of reception completion).
1887 I have revised the way this works. The header is now generated after the
1888 body is received, but before the ACL or local_scan() are called. After they
1889 are run, the timestamp in the header is updated.
1895 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1896 before starting a queue runner without re-exec. This happened only when
1897 deliver_drop_privilege was set or when the Exim user was set to root. The
1898 effect of the bug was that timeouts during subsequent deliveries caused
1899 crashes instead of being properly handled. The handler is now left at its
1900 default (and expected) setting.
1902 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1903 message, again when deliver_drop_privilege is set or Exim is run as root.
1904 The bug described in (1) was not present in this case, but the tidying up
1905 of the other signals was missing. I have made the two cases consistent.
1907 3. The ignore_target_hosts setting on a manualroute router was being ignored
1908 for hosts that were looked up using the /MX notation.
1910 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1913 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1914 operated on the sender address. After changing the $sender_address to <>
1915 for the sender address verify, Exim was re-instated it as the original
1916 (before rewriting) address, but remembering that it had rewritten it, so it
1917 wasn't rewriting it again. This bug also had the effect of breaking the
1918 sender address verification caching when the sender address was rewritten.
1920 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1921 This has been changed so that if the ip literal address matches
1922 ignore_target_hosts, the router declines.
1924 7. Added expansion conditions match_domain, match_address, and match_local_
1925 part (NOT match_host).
1927 8. The placeholder for the Received: header didn't have a length field set.
1929 9. Added code to Exim itself and to exim_lock to test for a specific race
1930 condition that could lead to file corruption when using MBX delivery. The
1931 issue is with the lockfile that is created in /tmp. If this file is removed
1932 after a process has opened it but before that process has acquired a lock,
1933 there is the potential for a second process to recreate the file and also
1934 acquire a lock. This could lead to two Exim processes writing to the file
1935 at the same time. The added code performs the same test as UW imapd; it
1936 checks after acquiring the lock that its file descriptor still refers to
1937 the same named file.
1939 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1940 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1943 11. Added the smtp_active_hostname option. If used, this will typically be made
1944 to depend on the incoming interface address. Because $interface_address is
1945 not set up until the daemon has forked a reception process, error responses
1946 that can happen earlier (such as "too many connections") no longer contain
1949 12. If an expansion in a condition on a "warn" statement fails because a lookup
1950 defers, the "warn" statement is abandoned, and the next ACL statement is
1951 processed. Previously this caused the whole ACL to be aborted.
1953 13. Added the iplsearch lookup type.
1955 14. Added ident_timeout as a log selector.
1957 15. Added tls_certificate_verified as a log selector.
1959 16. Added a global option tls_require_ciphers (compare the smtp transport
1960 option of the same name). This controls incoming TLS connections.
1962 17. I finally figured out how to make tls_require_ciphers do a similar thing
1963 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1964 before starting the TLS session.
1966 18. Tabs are now shown as \t in -bP output.
1968 19. If the log selector return_path_on_delivery was set, Exim crashed when
1969 bouncing a message because it had too many Received: header lines.
1971 20. If two routers both had headers_remove settings, and the first one included
1972 a superfluous trailing colon, the final name in the first list and the
1973 first name in the second list were incorrectly joined into one item (with a
1974 colon in the middle).
1980 1. Added -C and -D options to the exinext utility, mainly to make it easier
1981 to include in the automated testing, but these could be helpful when
1982 multiple configurations are in use.
1984 2. The exinext utility was not formatting the output nicely when there was
1985 an alternate port involved in the retry record key, nor when there was a
1986 message id as well (for retries that were specific to a specific message
1987 and a specific host). It was also confused by IPv6 addresses, because of
1988 the additional colons they contain. I have fixed the IPv4 problem, and
1989 patched it up to do a reasonable job for IPv6.
1991 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1992 delivery, the log line now contains "pipelined" if PIPELINING was used.
1994 4. An SMTP transport process used to panic and die if the bind() call to set
1995 an explicit outgoing interface failed. This has been changed; it is now
1996 treated in the same way as a connect() failure.
1998 5. A reference to $sender_host_name in the part of a conditional expansion
1999 that was being skipped was still causing a DNS lookup. This no longer
2002 6. The def: expansion condition was not recognizing references to header lines
2003 that used bh_ and bheader_.
2005 7. Added the _cache feature to named lists.
2007 8. The code for checking quota_filecount in the appendfile transport was
2008 allowing one more file than it should have been.
2010 9. For compatibility with Sendmail, the command line option
2018 and sets the incoming protocol and host name (for trusted callers). The
2019 host name and its colon can be omitted when only the protocol is to be set.
2020 Note the Exim already has two private options, -pd and -ps, that refer to
2021 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2022 "s", but I don't think that's a major issue.
2024 10. A number of refactoring changes to the code, none of which should affect
2027 (a) The number of logging options was getting close to filling up the
2028 32-bit word that was used as a bit map. I have split them into two classes:
2029 those that are passed in the argument to log_write(), and those that are
2030 only ever tested independently outside of that function. These are now in
2031 separate 32-bit words, so there is plenty of room for expansion again.
2032 There is no change in the user interface or the logging behaviour.
2034 (b) When building, for example, log lines, the code previously used a
2035 macro that called string_cat() twice, in order to add two strings. This is
2036 not really sufficiently general. Furthermore, there was one instance where
2037 it was actually wrong because one of the argument was used twice, and in
2038 one call a function was used. (As it happened, calling the function twice
2039 did not affect the overall behaviour.) The macro has been replaced by a
2040 function that can join an arbitrary number of extra strings onto a growing
2043 (c) The code for expansion conditions now uses a table and a binary chop
2044 instead of a serial search (which was left over from when there were very
2045 few conditions). Also, it now recognizes conditions like "pam" even when
2046 the relevant support is not compiled in: a suitably worded error message is
2047 given if an attempt is made to use such a condition.
2049 11. Added ${time_interval:xxxxx}.
2051 12. A bug was causing one of the ddress fields not to be passed back correctly
2052 from remote delivery subprocesses. The field in question was not being
2053 subsequently used, so this caused to problems in practice.
2055 13. Added new log selectors queue_time and deliver_time.
2057 14. Might have fixed a bug in maildirsizefile handling that threw up
2058 "unexpected character" debug warnings, and recalculated the data
2059 unnecessarily. In any case, I expanded the warning message to give more
2062 15. Added the message "Restricted characters in address" to the statements in
2063 the default ACL that block characters like @ and % in local parts.
2065 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2066 Three changes have been made:
2068 (a) There was a serious bug; a negative response to MAIL caused the whole
2069 recipient domain to be cached as invalid, thereby blocking all messages
2070 to all local parts at the same domain, from all senders. This bug has
2071 been fixed. The domain is no longer cached after a negative response to
2072 MAIL if the sender used is not empty.
2074 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2077 (c) A new callout option, "use_sender" has been added for people who want
2078 the modified behaviour.
2084 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2087 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2088 indeed breaks things for older releases.
2090 3. Added additional logging to the case where there is a problem reading data
2091 from a filter that is running in a subprocess using a pipe, in order to
2092 try to track down a specific problem.
2094 4. Testing facility fudge: when running in the test harness and attempting
2095 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2096 getting "No route to host". Convert this to a timeout.
2098 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2101 6. Some OS don't have socklen_t but use size_t instead. This affects the
2102 fifth argument of getsockopt() amongst other things. This is now
2103 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2104 can be set for individual OS. I have set it for SunOS5, OSF1, and
2105 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2106 some earlier ones do not.
2108 7. Change 4.30/15 was not doing the test caselessly.
2110 8. The standard form for an IPv6 address literal was being rejected by address
2111 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2112 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2113 this, as well as the form without the "IPv6" on the front (but only when
2114 address literals are enabled, of course).
2116 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2118 10. Exim crashed if a message with an empty sender address specified by -f
2119 encountered a router with an errors_to setting. This could be provoked only
2120 by a command such as
2124 where an empty string was supplied; "<>" did not hit this bug.
2126 11. Installed PCRE release 4.5.
2128 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2129 remained set. It is now erased.
2131 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2132 times from message ids (which are base 36 rather than the normal 62).
2134 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2135 were being counted as actual protocol errors, and logged if the log
2136 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2137 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2138 and DATA following a set of rejected RCPTs do not count as protocol errors.
2139 In other words, Exim assumes they were pipelined, though this may not
2140 actually be the case. Of course, in all cases the client gets an
2141 appropriate error code.
2143 15. If a lookup fails in an ACL condition, a message about the failure may
2144 be available; it is used if testing the ACL cannot continue, because most
2145 such messages specify what the cause of the deferral is. However, some
2146 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2147 that caused an old message to be retained and used if a later statement
2148 caused a defer, replacing the real cause of the deferral.
2150 16. If an IP address had so many PTR records that the DNS lookup buffer
2151 was not large enough to hold them, Exim could crash while trying to process
2152 the truncated data. It now detects and logs this case.
2154 17. Further to 4.21/58, another change has been made: if (and only if) the
2155 first line of a message (the first header line) ends with CRLF, a bare LF
2156 in a subsequent header line has a space inserted after it, so as not to
2157 terminate the header.
2159 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2160 unnecessarily, used atoi() instead of strtol(), and didn't check the
2161 termination when getting file sizes from file names by regex.
2163 19. Completely re-implemented the support for maildirsize files, in the light
2164 of a number of problems with the previous contributed implementation
2165 (4.30/29). In particular:
2167 . If the quota is zero, the maildirsize file is maintained, but no quota is
2170 . If the maildir directory does not exist, it is created before any attempt
2171 to write a maildirsize file.
2173 . The quota value in the file is just a cache; if the quota is changed in
2174 the transport, the new value overrides.
2176 . A regular expression is available for excluding directories from the
2179 20. The autoreply transport checks the characters in options that define the
2180 message's headers; it allows continued headers, but it was checking with
2181 isspace() after an embedded newline instead of explicitly looking for a
2184 21. If all the "regular" hosts to which an address was routed had passed their
2185 expiry times, and had not reached their retry times, the address was
2186 bounced, even if fallback hosts were defined. Now Exim should go on to try
2189 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2190 equivalent code in the SMTP transport. Some hosts send humungous responses
2191 to HELO/EHLO, more than 1024 it seems.
2193 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2194 gives compiler warnings in some environments. I've now done it "properly",
2197 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2198 (because of problems with the built-in one) was declared to return uschar *
2199 instead of char *, causing compiler failure.
2201 25. Fixed a file descriptor leak when processing alias/forward files.
2203 26. Fixed a minor format string issue in dbfn.c.
2205 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2207 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2208 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2211 29. When a sender address is verified, it is cached, to save repeating the test
2212 when there is more than one recipient in a message. However, when the
2213 verification involves a callout, it is possible for different callout
2214 options to be set for different recipients. It is too complicated to keep
2215 track of this in the cache, so now Exim always runs a verification when a
2216 callout is required, relying on the callout cache for the optimization.
2217 The overhead is duplication of the address routing, but this should not be
2220 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2221 to be verified with callout=postmaster, and the main callout worked but the
2222 postmaster check failed, the verification correctly failed. However, if a
2223 subsequent RCPT command asked for sender verification *without* the
2224 postmaster check, incorrect caching caused this verification also to fail,
2227 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2228 it was not caching the DNS options (qualify_single, search_parents) that
2229 were used when the lookup failed. A subsequent lookup with different
2230 options therefore always gave the same answer, though there were cases
2231 where it should not have. (Example: a "domains = !$mx_any" option on a
2232 dnslookup router: the "domains" option is always processed without any
2233 widening, but the router might have qualify_single set.) Now Exim uses the
2234 cached value only when the same options are set.
2236 32. Added John Jetmore's "exipick" utility to the distribution.
2238 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2239 than a timeout (e.g. a certificate is required, and is not provided), an
2240 Exim server now closes the connection immediately. Previously it waited for
2241 the client to close - but if the client is SSL, it seems that they each
2242 wait for each other, leading to a delay before one of them times out.
2244 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2245 maintained 0.8.x compatibility because I don't think many are using it, and
2246 it is clearly obsolete.
2248 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2251 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2252 client certificate was expired. A simple patch fixes this, though I don't
2253 understand the full logic of why the verify callback is called multiple
2256 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2257 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2258 which causes problems with some clients (such as the Certicom SSL Plus
2259 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2260 disables the coutermeasure allowing Eudora to connect."
2262 38. Exim was not checking that a write() to a log file succeeded. This could
2263 lead to Bad Things if a log got too big, in particular if it hit a file
2264 size limit. Exim now panics and dies if it cannot write to a log file, just
2265 as it does if it cannot open a log file.
2267 39. Modified OS/Makefile-Linux so that it now contains
2269 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2271 The two -D definitions ensure that Exim is compiled with large file
2272 support, which makes it possible to handle log files that are bigger than
2275 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2276 instance) a domain was checked against a named list that involved a lookup,
2277 causing $domain_data to be set, then another domain was checked against the
2278 same list, then the first domain was re-checked, the value of $domain_data
2279 after the final check could be wrong. In particular, if the second check
2280 failed, it could be set empty. This bug probably also applied to
2283 41. The strip_trailing_dot option was not being applied to the address given
2284 with the -f command-line option.
2286 42. The code for reading a message's header from the spool was incrementing
2287 $received_count, but never initializing it. This meant that the value was
2288 incorrect (doubled) while delivering a message in the same process in which
2289 it was received. In the most common configuration of Exim, this never
2290 happens - a fresh exec is done - but it can happen when
2291 deliver_drop_privilege is set.
2293 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2294 it now includes up to 150 characters of the unexpected data in the log
2297 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2298 and building data strings. The size of both of these buffers was 10 000
2299 bytes - far larger than anybody would *ever* want, thought I. Needless to
2300 say, somebody hit the limit. I have increased the maximum line length to
2301 20 000 and the maximum data length of concatenated lines to 100 000. I have
2302 also fixed two bugs, because there was no checking on these buffers. Tsk,
2303 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2304 buffer is too small.
2306 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2307 lsearch lookups. Now it does.
2309 46. When parsing a route_list item in a manualroute router, a fixed-length
2310 buffer was used for the list of hosts. I made this 1024 bytes long,
2311 thinking that nobody would ever have a list of hosts that long. Wrong.
2312 Somebody had a whole pile of complicated expansion conditions, and the
2313 string was silently truncated, leading to an expansion error. It turns out
2314 that it is easier to change to an unlimited length (owing to other changes
2315 that have happened since this code was originally written) than to build
2316 structure for giving a limitation error. The length of the item that
2317 expands into the list of hosts is now unlimited.
2319 47. The lsearch lookup could not handle data where the length of text line was
2320 more than 4095 characters. Such lines were truncated, leading to shortened
2321 data being returned. It should now handle lines of any length.
2323 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2324 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2327 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2328 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2329 no longer happens in comment lines. A list of replacements is now placed
2330 at the head of all of the source files, except those whose only change is
2331 to replace PERL_COMMAND in the very first #! line.
2333 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2334 messages on the queue, with a bottom-up merge sort, using code contributed
2335 by Michael Haardt. This should make operations like -bp somewhat faster on
2336 large queues. It won't affect queue runners, except when queue_run_in_order
2339 51. Installed eximstats 1.31 in the distribution.
2341 52. Added support for SRV lookups to the dnslookup router.
2343 53. If an ACL referred to $message_body or $message_body_end, the value was not
2344 reset for any messages that followed in the same SMTP session.
2346 54. The store-handling optimization for building very long strings was not
2347 differentiating between the different store pools. I don't think this
2348 actually made any difference in practice, but I've tidied it.
2350 55. While running the routers to verify a sender address, $sender_address
2351 was still set to the sender address. This is wrong, because when routing to
2352 send a bounce to the sender, it would be empty. Therefore, I have changed
2353 it so that, while verifying a sender address, $sender_address is set to <>.
2354 (There is no change to what happens when verifying a recipient address.)
2356 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2357 target A or AAAA records (if not already returned) without resetting the
2358 qualify_single or search_parents options of the DNS resolver. These are
2359 inappropriate in this case because the targets of MX and SRV records must
2360 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2361 target that, when qualified, matched something in the local domain. These
2362 two options are now turned off when doing these lookups.
2364 57. It seems that at least some releases of Reiserfs (which does not have the
2365 concept of a fixed number of inodes) returns zero and not -1 for the
2366 number of available inodes. This interacted badly with check_spool_inodes,
2367 which assumed that -1 was the "no such thing" setting. What I have done is
2368 to check that the total number of inodes is greater than zero before doing
2369 the test of how many are available.
2371 58. When a "warn" ACL statement has a log_message modifier, the message is
2372 remembered, and not repeated. This is to avoid a lot of repetition when a
2373 message has many recipients that cause the same warning to be written.
2374 Howewer, Exim was preserving the list of already written lines for an
2375 entire SMTP session, which doesn't seem right. The memory is now reset if a
2376 new message is started.
2378 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2379 debugging output unless log_rewrite was also set.
2381 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2382 of (int)(handle) when we know that handle contains (void *)(-1).
2384 61. The Exim daemon panic-logs an error return when it closes the incoming
2385 connection. However "connection reset by peer" seems to be common, and
2386 isn't really an error worthy of noting specially, so that particular error
2389 62. When Exim is trying to find all the local interfaces, it used to panic and
2390 die if the ioctl to get the interface flags failed. However, it seems that
2391 on at least one OS (Solaris 9) it is possible to have an interface that is
2392 included in the list of interfaces, but for which you get a failure error
2393 for this call. This happens when the interface is not "plumbed" into a
2394 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2395 failure of the "get flags" call assumes that the interface is down.
2397 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2398 makes life easier for people who are doing arithmetic on fields extracted
2399 from dates, where you often get leading zeros that should not be
2400 interpreted as octal.
2402 64. Added qualify_domain to the redirect router, to override the global
2405 65. If a pathologically long header line contained very many addresses (the
2406 report of this problem mentioned 10 000) and each of them was rewritten,
2407 Exim could use up a very large amount of memory. (It kept on making new
2408 copies of the header line as it rewrote, and never released the old ones.)
2409 At the expense of a bit more processing, the header rewriting function has
2410 been changed so that it no longer eats memory in this way.
2412 66. The generation of the Received: header has been moved from the time that a
2413 message starts to be received, to the time that it finishes. The timestamp
2414 in the Received: header should now be very close to that of the <= log
2415 line. There are two side-effects of this change:
2417 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2418 logged header lines no longer include the local Received: line, because
2419 it has not yet been created. The same applies to a copy of the message
2420 that is returned to a non-SMTP sender when a message is rejected.
2422 (b) When a filter file is tested using -bf, no additional Received: header
2423 is added to the test message. After some thought, I decided that this
2426 This change does not affect the value of $received_for. It is still set
2427 after address rewriting, but before local_scan() is called.
2429 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2431 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2432 gave an unhelpful panic error message, and a defer error. I have managed to
2433 change this behaviour so that it now rejects any supplied certificate,
2434 which seems right, as the list of acceptable certificates is empty.
2436 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2437 gave an unhelpful defer error. I have not managed to make this reject any
2438 supplied certificates, but the error message it gives is "no certificate
2439 supplied", which is not helpful.
2441 70. exigrep's output now also includes lines that are not associated with any
2442 message, but which match the given pattern. Implemented by a patch from
2443 Martin Sluka, which also tidied up the Perl a bit.
2445 71. Recipient callout verification, like sender verification, was using <> in
2446 the MAIL FROM command. This isn't really the right thing, since the actual
2447 sender may affect whether the remote host accepts the recipient or not. I
2448 have changed it to use the actual sender in the callout; this means that
2449 the cache record is now keyed on a recipient/sender pair, not just the
2450 recipient address. There doesn't seem to be a real danger of callout loops,
2451 since a callout by the remote host to check the sender would use <>.
2452 [SEE ABOVE: changed after hitting problems.]
2454 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2455 temporary errors. However, in the case of such a code being given after
2456 the end of a data transmission (i.e. after ".") Exim was failing to write
2457 a retry record for the message. (Yes, there was some broken host that was
2458 actually sending 8xx at this point.)
2460 73. An unknown lookup type in a host list could cause Exim to panic-die when
2461 the list was checked. (An example that provoked this was putting <; in the
2462 middle of a list instead of at the start.) If this happened during a DATA
2463 ACL check, a -D file could be left lying around. This kind of configuration
2464 error no longer causes Exim to die; instead it causes a defer errror. The
2465 incident is still logged to the main and panic logs.
2467 74. Buglet left over from Exim 3 conversion. The message "too many messages
2468 in one connection" was written to the rejectlog but not the mainlog, except
2469 when address rewriting (yes!) was being logged.
2471 75. Added write_rejectlog option.
2473 76. When a system filter was run not as root (that is, when system_filter_user
2474 was set), the values of the $n variables were not being returned to the
2475 main process; thus, they were not subsequently available in the $sn
2478 77. Added +return_path_on_delivery log selector.
2480 78. A connection timeout was being treated differently from recipients deferred
2481 when testing hosts_max_try with a message that was older than the host's
2482 retry timeout. (The host should not be counted, thus allowing all hosts to
2483 be tried at least once before bouncing.) This may have been the cause of an
2484 occasionally reported bug whereby a message would remain on the queue
2485 longer than the retry timeout, but would be bounced if a delivery was
2486 forced. I say "may" because I never totally pinned down the problem;
2487 setting up timeout/retry tests is difficult. See also the next item.
2489 79. The ultimate address timeout was not being applied to errors that involved
2490 a combination of host plus message (for example, a timeout on a MAIL
2491 command). When an address resolved to a number of possible hosts, and they
2492 were not all tried for each delivery (e.g. because of hosts_max_try), a
2493 message could remain on the queue longer than the retry timeout.
2495 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2498 81. Fixed an obscure SMTP outgoing bug which required at least the following
2499 conditions: (a) there was another message waiting for the same server;
2500 (b) the server returned 5xx to all RCPT commands in the first message so
2501 that the message was not completed; (c) the server dropped the connection
2502 or gave a negative response to the RSET that Exim sends to abort the
2503 transaction. The observed case was a dropped connection after DATA that had
2504 been sent in pipelining mode. That is, the server had advertised PIPELINING
2505 but was not implementing it correctly. The effect of the bug was incorrect
2506 behaviour, such as trying another host, and this could lead to a crash.
2512 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2513 and daemon.c were passed as pointers to ints; they should have been
2514 pointers to socklen_t variables (which are typically unsigned ints).
2516 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2519 3. Fixed a really odd bug that affected only the testing scheme; patching a
2520 certain fixed string in the binary changed the value of another string that
2521 happened to be identical to the end of the original first string.
2523 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2524 name", it returns that address as the IP address. On some operating
2525 systems (e.g. Solaris), it also passes back the IP address string as the
2526 "host name". However, on others (e.g. Linux), it passes back an empty
2527 string. Exim wasn't checking for this, and was changing the host name to an
2528 empty string, assuming it had been canonicized.
2530 5. Although rare, it is permitted to have more than one PTR record for a given
2531 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2532 all the names associated with an address, because they do in Solaris.
2533 However, it seems that they do not in Linux for data that comes from the
2534 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2535 I found this out when I moved to a new Linux workstation and tried to run
2536 the Exim test suite.
2538 To get round this problem I have changed the code so that it now does its
2539 own call to the DNS to look up PTR records when searching for a host name.
2540 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2541 addresses that are only in /etc/hosts are still found.
2543 This behaviour is, however, controlled by an option called host_lookup_
2544 order, which defaults to "bydns:byaddr". If people want to use the other
2545 order, or indeed, just use one or the other means of lookup, they can
2546 specify it in this variable.
2548 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2549 some operating systems, this comes back from gethostbyaddr() as an empty
2550 string, and this is what Exim used to test for. However, it seems that in
2551 other systems, "." is yielded. Exim now tests for this case too.
2553 7. The values of check_spool_space and check_log_space are now held internally
2554 as a number of kilobytes instead of an absolute number of bytes. If a
2555 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2556 kilobyte. This means that much larger values can be stored.
2558 8. Exim monitor: an attempt to get the action menu when not actually pointing
2559 at a message produces an empty menu entitled "No message selected". This
2560 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2561 no entries in it ("Shell widget menu has zero width and/or height"). So I
2562 have added a single, blank menu entry in this case.
2564 9. Added ${quote_local_part.
2566 10. MIME decoding is now applied to the contents of Subject: header lines when
2569 11. Now that a reference to $sender_host_address automatically causes a reverse
2570 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2571 host lookup before query-style lookups in lists that might use this
2572 variable. This has therefore been abolished, and the "net-" prefix is no
2573 longer necessary for query-style lookups.
2575 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2576 have been a typo for LFLAGS, so it has been changed.
2578 13. The install script calls Exim with "-C /dev/null" in order to find the
2579 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2580 to be output. Howeve, since Exim outputs its version number before the
2581 error, it didn't break the script. It just looked ugly. I fixed this by
2582 always allowing "-C /dev/null" if the caller is root.
2584 14. Ignore overlarge ACL variable number when reading spool file - insurance
2585 against a later release with more variables having written the file.
2587 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2588 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2589 this, as well as the form without the "IPv6" on the front.
2591 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2592 OS/Makefile-Darwin file.
2594 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2595 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2597 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2598 variable. However, some OS use ints for inodes. I've added cast to long int
2599 to get rid of the compiler warning.
2601 19. I had forgotten to lock out "/../" in configuration file names when
2602 ALT_CONFIG_PREFIX was set.
2604 20. Routers used for verification do not need to specify transports. However,
2605 if such a router generated a host list, and callout was configured, Exim
2606 crashed, because it could not find a port number from the (non-existent)
2607 transport. It now assumes port 25 in this circumstance.
2609 21. Added the -t option to exigrep.
2611 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2612 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2613 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2614 them is set, LOOKUP_LSEARCH is forced.
2616 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2618 24. Added sender and host information to the "rejected by local_scan()" log
2619 line; previously there was no indication of these.
2621 25. Added .include_if_exists.
2623 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2624 Linux. It turns out that not all file systems support this. Apparently some
2625 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2626 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2627 not supported on the file descriptor, is now ignored when Exim is trying to
2628 sync a directory. This applies only to Linux.
2630 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2632 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2633 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2636 29. Added support for maildirsize files from supplied patch (modified a bit).
2638 30. The use of :fail: followed by an empty string could lead Exim to respond to
2639 sender verification failures with (e.g.):
2641 550 Verification failed for <xxx>
2642 550 Sender verify failed
2644 where the first response line was missing the '-' that indicates it is not
2645 the final line of the response.
2647 31. The loop for finding the name of the user that called Exim had a hardwired
2648 limit of 10; it now uses the value of finduser_retries, which is used for
2649 all other user lookups.
2651 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2654 33. Exim was neglecting to zero errno before one call of strtol() when
2655 expanding a string and expecting an integer value. On some systems this
2656 resulted in spurious "integer overflow" errors. Also, it was casting the
2657 result into an int without checking.
2659 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2660 did not work. The code looks as if it has *never* worked, though it appears
2661 to have been documented since at least releast 1.62. I have made it work.
2663 35. The "timeout_DNS" error in retry rules, also documented since at least
2664 1.62, also never worked. As it isn't clear exactly what this means, and
2665 clearly it isn't a major issue, I have abolished the feature by treating it
2666 as "timeout", and writing a warning to the main and panic logs.
2668 36. The display of retry rules for -brt wasn't always showing the error code
2671 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2672 timeout_connect_A, timeout_connect_MX.
2674 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2675 to the empty sender.
2677 39. The daemon was not analysing the content of -oX till after it had closed
2678 stderr and disconnected from the controlling terminal. This meant that any
2679 syntax errors were only noted on the panic log, and the return code from
2680 the command was 0. By re-arranging the code a little, I've made the
2681 decoding happen first, so such errors now appear on stderr, and the return
2682 code is 1. However, the actual setting up of the sockets still happens in
2683 the disconnected process, so errors there are still only recorded on the
2686 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2687 connections (as happens on some IP stacks) was logged at start up time as
2688 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2689 it from "IPv6 and IPv4", which means that two separate sockets are being
2692 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2693 says whether AF_INET or AF_INET6 was passed as an argument.
2695 42. Exiwhat output was messed up when time zones were included in log
2698 43. Exiwhat now gives more information about the daemon's listening ports,
2699 and whether -tls-on-connect was used.
2701 44. The "port" option of the smtp transport is now expanded.
2703 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2704 silently ignored. Now an error message is written to the main and panic
2707 46. There's a new ACL modifier called "logwrite" which writes to a log file
2708 as soon as it is encountered.
2710 47. Added $local_user_uid and $local_user_gid at routing time.
2712 48. Exim crashed when trying to verify a sender address that was being
2715 49. Exim was recognizing only a space character after ".include". It now also
2716 recognizes a tab character.
2718 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2719 extracting the relevant information from the specification. The man page no
2720 longer contains scrambled data for the -d option, and I've added a section
2721 at the front about calling Exim under different names.
2723 51. Added "extra_headers" argument to the "mail" command in filter files.
2725 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2728 53. Installed eximstats 1.29.
2730 54. Added transport_filter_timeout as a generic transport option.
2732 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2733 Cc: header lines. This was required by RFC 822, but it not required by RFC
2736 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2737 incoming messages that did not have them. Now it does so only if the
2738 message originates locally, that is, if there is no associated remote host
2739 address. When Resent- header lines are present, this applies to the Resent-
2740 lines rather than the non-Resent- lines.
2742 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2743 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2745 58. Messages for configuration errors now include the name of the main
2746 configuration file - useful now that there may be more than one file in a
2747 list (.included file names were always shown).
2749 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2750 for those rare installations that do not start the daemon as root or run it
2751 setuid root. I've cut out the call to initgroups() if the daemon is not
2754 60. The Exim user and group can now be bound into the binary as text strings
2755 that are looked up at the start of Exim's processing.
2757 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2759 62. Added $mailstore_basename variable.
2761 63. Installed patch to sieve.c from Michael Haardt.
2763 64. When Exim failed to open the panic log after failing to open the main log,
2764 the original message it was trying to log was written to stderr and debug
2765 output, but if they were not available (the usual case in production), it
2766 was lost. Now it is written to syslog before the two lines that record the
2767 failures to open the logs.
2769 65. Users' Exim filters run in subprocesses under the user's uid. It is
2770 possible for a "deliver" command or an alias in a "personal" command to
2771 provoke an address rewrite. If logging of address rewriting is configured,
2772 this fails because the process is not running as root or exim. There may be
2773 a better way of dealing with this, but for the moment (because 4.30 needs
2774 to be released), I have disabled address rewrite logging when running a
2775 filter in a non-root, non-exim process.
2781 1. The buildconfig auxiliary program wasn't quoting the value set for
2782 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2783 not defined. This bug was masked in 4.22 by the effect that was fixed in
2786 2. Some messages that were rejected after a message id was allocated were
2787 shown as "incomplete" by exigrep. It no longer does this for messages that
2788 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2790 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2791 have allow_domain_literals set, the ID did not get logged in the <= line.
2792 Domain literals are now always recognized in Message-ID: header lines.
2794 4. The first argument for a ${extract expansion item is the key name or field
2795 number. Leading and trailing spaces in this item were not being ignored,
2796 causing some misleading effects.
2798 5. When deliver_drop_privilege was set, single queue runner processes started
2799 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2800 same command in the process it spins off) were not dropping privilege.
2802 6. When the daemon running as "exim" started a queue runner, it always
2803 re-executed Exim in the spun-off process. This is a waste of effort when
2804 deliver_drop_privilege is set. The new process now just calls the
2805 queue-runner function directly.
2811 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2814 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2815 ignored. Though the use of -oP was forcing the writing of a pid file, it
2816 was always written to the default place.
2818 3. If the message "no IP address found for host xxxx" is generated during
2819 incoming verification, it is now followed by identification of the incoming
2820 connection (so you can more easily find what provoked it).
2822 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2824 5. Added some features to "harden" Exim a bit more against certain attacks:
2826 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2827 be put in Local/Makefile. This is like the never_users runtime option,
2828 but it cannot be overridden. The default setting is "root".
2830 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2831 prefix string with which any file named in a -C command line option
2834 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2835 is retained for -C and -D only if the caller of Exim is root. Without
2836 it, the exim user may also use -C and -D and retain privilege.
2838 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2839 command line option is disabled.
2841 6. Macro names set by the -D option must start with an upper case letter, just
2842 like macro names defined in the configuration file.
2844 7. Added "dereference=" facility to LDAP.
2846 8. Two instances of the typo "uknown" in the source files are fixed.
2848 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2849 the Configure-Makefile script screwed up while processing it.
2851 10. Incorporated PCRE 4.4.
2853 11. The SMTP synchronization check was not operating right at the start of an
2854 SMTP session. For example, it could not catch a HELO sent before the client
2855 waited for the greeting. There is now a check for outstanding input at the
2856 point when the greeting is written. Because of the duplex, asynchronous
2857 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2858 way, but not yet received, when the check is performed.
2860 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2861 on TCP/IP sockets, because this apparently causes some broken clients to
2864 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2865 unchanged) from the Cygwin maintainer.
2867 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2868 was supported instead of testing for mbx. Effectively a typo.
2870 15. The spa authenticator server code was not checking that the input it
2871 received was valid base64.
2873 16. The debug output line for the "set" modifier in ACLs was not showing the
2874 name of the variable that was being set.
2876 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2878 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2879 lookup on its own. Something else was needed to trigger the lookup. For
2880 example, a match in host_lookup or the need for a host name in a host list.
2881 Now, if $sender_host_name is referenced and the host name has not yet been
2882 looked up, a lookup is performed. If the lookup fails, the variable remains
2883 empty, and $host_lookup_failed is set to "1".
2885 19. Added "eqi" as a case-independent comparison operator.
2887 20. The saslauthd authentication condition could segfault if neither service
2888 nor realm was specified.
2890 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2891 error message that was logged was misleading, and incoming SMTP
2892 connections were dropped. The message is now more accurate, and temporary
2893 errors are given to SMTP connections.
2895 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2896 (except RSET) with a 421 error, until QUIT is received. However, it was
2897 failing to send a response to QUIT.
2899 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2900 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2901 ACL is now run after the helo_try_verify_hosts code.
2903 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2904 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2905 case-independent, so other case variants are also recognized). Apparently
2906 some systems use these upper case variants.
2908 25. If more than two messages were waiting for the same host, and a transport
2909 filter was specified for the transport, Exim sent two messages over the
2910 same TCP/IP connection, and then failed with "socket operation on non-
2911 socket" when it tried to send the third.
2913 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2915 27. The extern definition of crypt16() in expand.c was not being excluded when
2916 the OS had its own crypt16() function.
2918 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2919 as a preferred synonym for return_size_limit, both as an option and as an
2922 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2924 30. Changed the default configuration ACL to relax the local part checking rule
2925 for addresses that are not in any local domains. For these addresses,
2926 slashes and pipe symbols are allowed within local parts, but the sequence
2927 /../ is explicitly forbidden.
2929 31. SPA server authentication was not clearing the challenge buffer before
2932 32. log_message in a "warn" ACL statement was writing to the reject log as
2933 well as to the main log, which contradicts the documentation and doesn't
2934 seem right (because no rejection is happening). So I have stopped it.
2936 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2937 However, I am unable to do any testing of this.
2939 34. Fixed an infelicity in the appendfile transport. When checking directories
2940 for a mailbox, to see if any needed to be created, it was accidentally
2941 using path names with one or more superfluous leading slashes; tracing
2942 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2944 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2945 contents are added to the log line that is written for every discarded
2946 recipient. (Previously a log_message setting was ignored.)
2948 36. The ${quote: operator now quotes the string if it is empty.
2950 37. The install script runs exim in order to find its version number. If for
2951 some reason other than non-existence or emptiness, which it checks, it
2952 could not run './exim', it was installing it with an empty version number,
2953 i.e. as "exim-". This error state is now caught, and the installation is
2956 38. An argument was missing from the function that creates an error message
2957 when Exim fails to connect to the socket for saslauthd authentication.
2958 This could cause Exim to crash, or give a corrupted message.
2960 39. Added isip, isip4, and isip6 to ${if conditions.
2962 40. The ACL variables $acl_xx are now saved with the message, and can be
2963 accessed later in routers, transports, and filters.
2965 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2966 strings in the file are not string-expanded.
2968 42. If a MAIL command specified a SIZE value that was too large to fit into an
2969 int variable, the check against message_size_limit failed. Such values are
2970 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2971 day this will have to be increased, but I don't think I want to be around
2972 when emails are that large.
2979 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2980 iconv() is not standard in FreeBSD.
2982 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2983 IPv6 enabled. The observed symptom was a segmentation fault on return from
2984 the function os_common_find_running_interfaces() in src/os.c.
2986 3. In the check_special_case() function in daemon.c I had used "errno" as an
2987 argument name, which causes warnings on some systems. This was basically a
2988 typo, since it was named "eno" in the comments!
2990 4. The code that waits for the clock to tick (at a resolution of some fraction
2991 of a second) so as to ensure message-id uniqueness was always waiting for
2992 at least one whole tick, when it could have waited for less. [This is
2993 almost certainly not relevant at current processor speeds, where it is
2994 unlikely to ever wait at all. But we try to future-proof.]
2996 5. The function that sleeps for a time interval that includes fractions of a
2997 second contained a race. It did not block SIGALRM between setting the
2998 timer, and suspending (a couple of lines later). If the interval was short
2999 and the sigsuspend() was delayed until after it had expired, the suspension
3000 never ended. On busy systems this could lead to processes getting stuck for
3003 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3004 process, before it forks any delivery processes. The open lookup caching
3005 mechanism meant that the open file or database connection was passed into
3006 the delivery process. The problem was that delivery processes always tidy
3007 up cached lookup data. This could cause a problem for the next delivery
3008 process started by the queue runner, because the external queue runner
3009 process does not know about the closure. So the next delivery process
3010 still has data in the lookup cache. In the case of a file lookup, there was
3011 no problem because closing a file descriptor in a subprocess doesn't affect
3012 the parent. However, if the lookup was caching a connection to a database,
3013 the connection was closed, and the second delivery process was likely to
3014 see errors such as "PGSQL: query failed: server closed the connection
3015 unexpectedly". The problem has been fixed by closing all cached lookups
3016 in a queue runner before running a delivery process.
3018 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3019 seem to have the "const" qualifier which it has on other OS. I've
3022 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3023 *of the same type* (client or server) with the same public name that an
3024 error should be diagnosed.
3026 9. When Exim looked up a host name for an IP address, but failed to find the
3027 original IP address when looking up the host name (a safety check), it
3028 output the message "<ip address> does not match any IP for NULL", which was
3029 confusing, to say the least. The bug was that the host name should have
3030 appeared instead of "NULL".
3032 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3033 user that is built into the binary, and the -C or -D options is used, root
3034 privilege is dropped before the configuration file is read. In addition,
3035 logging is switched to stderr instead of the normal log files. If the
3036 configuration then re-defines the Exim user, the unprivileged environment
3037 is probably not what is expected, so Exim logs a panic warning message (but
3040 However, if deliver_drop_privilege is set, the unprivileged state may well
3041 be exactly what is intended, so the warning has been cut out in that case,
3042 and Exim is allowed to try to write to its normal log files.
3048 1. smtp_return_error_details was not giving details for temporary sender
3049 or receiver verification errors.
3051 2. Diagnose a configuration error if two authenticators have the same public
3054 3. Exim used not to create the message log file for a message until the first
3055 delivery attempt. This could be confusing when incoming messages were held
3056 for policy or load reasons. The message log file is now created at the time
3057 the message is received, and an initial "Received" line is written to it.
3059 4. The automatically generated man page for command line options had a minor
3060 bug that caused no ill effects; however, a more serious problem was that
3061 the procedure for building the man page automatically didn't always
3062 operate. Consequently, release 4.20 contains an out-of-date version. This
3063 shouldn't happen again.
3065 5. When building Exim with embedded Perl support, the script that builds the
3066 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3067 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3069 6. The freeze_tell option was not being used for messages that were frozen on
3070 arrival, either by an ACL or by local_scan().
3072 7. Added the smtp_incomplete_transaction log selector.
3074 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3075 was accepting AUTH without a new EHLO.
3077 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3078 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3081 10. Logging of TCP/IP connections (when configured) now happens in the main
3082 daemon process instead of the child process, so that the TCP/IP connection
3083 count is more accurate (but it can never be perfect).
3085 11. The use of "drop" in a nested ACL was not being handled correctly in the
3086 outer ACL. Now, if condition failure induced by the nested "drop" causes
3087 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3088 or "require"), the connection is dropped.
3090 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3091 that yield "discard" can now be used with an "accept" or a "discard" verb,
3092 but an error is generated for any others (because I can't see a useful way
3093 to define what should happen).
3095 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3096 are now processed in the same way as lines in the Exim configuration file.
3097 In particular, continuation lines are supported.
3099 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3101 15. Added -ti meaning -t -i.
3103 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3104 domains, and warn by logging if others are found.
3106 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3107 returned from GIFCONF when Exim is trying to find the list of interfaces on
3108 a host. The code in os.c has been modified to copy each ifreq to an aligned
3109 structure in all cases.
3111 Also, in some cases, the returned ifreq's were being copied to a 'struct
3112 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3113 means the last couple of bytes of an IPv6 address could be chopped if the
3114 ifreq contained only a normal sockaddr (14 bytes storage).
3116 18. Named domain lists were not supported in the hosts_treat_as_local option.
3117 An entry such as +xxxx was not recognized, and was treated as a literal
3120 19. Ensure that header lines added by a DATA ACL are included in the reject log
3121 if the ACL subsequently rejects the message.
3123 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3124 MD5 (which is deprecated).
3126 21. When testing a filter file using -bf, Exim was writing a message when it
3127 took the sender from a "From " line in the message, but it was not doing so
3128 when it took $return_path from a Return-Path: header line. It now does.
3130 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3131 with a valid header line field name (a series of printing characters
3132 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3134 23. Changed "disc" in the source to "disk" to conform to the documentation and
3135 the book and for uniformity.
3137 24. Ignore Sendmail's -Ooption=value command line item.
3139 25. When execve() failed while trying to run a command in a pipe transport,
3140 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3141 could be confused with a return value of 69 from the command itself. This
3142 has been changed to 127, the value the shell returns if it is asked to run
3143 a non-existent command. The wording for the related log line suggests a
3144 non-existent command as the problem.
3146 26. If received_header_text expands to an empty string, do not add a Received:
3147 header line to the message. (Well, it adds a token one on the spool, but
3148 marks it "old" so that it doesn't get used or transmitted.)
3150 27. Installed eximstats 1.28 (addition of -nt option).
3152 28. There was no check for failure on the call to getsockname() in the daemon
3153 code. This can fail if there is a shortage of resources on the system, with
3154 ENOMEM, for example. A temporary error is now given on failure.
3156 29. Contrary to the C standard, it seems that in some environments, the
3157 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3158 program. Exim now does this explicitly; it affects the formatting of
3159 timestamps using strftime().
3161 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3162 complaints. I've now initialized all the variables, to avoid this.
3164 32. Header lines added by a system filter were not being "seen" during
3165 transport-time rewrites.
3167 33. The info_callback() function passed to OpenSSL is set up with type void
3168 (*)(SSL *, int, int), as described somewhere. However, when calling the
3169 function (actually a macro) that sets it up, the type void(*)() is
3170 expected. I've put in a cast to prevent warnings from picky compilers.
3172 34. If a DNS black list lookup found a CNAME record, but there were no A
3173 records associated with the domain it pointed at, Exim crashed.
3175 35. If a DNS black list lookup returned more than one A record, Exim ignored
3176 all but the first. It now scans all returned addresses if a particular IP
3177 value is being sought. In this situation, the contents of the
3178 $dnslist_value variable are a list of all the addresses, separated by a
3181 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3182 to accept a host name and all its aliases if the forward lookup for any of
3183 them yielded the IP address of the incoming connection. Now it accepts only
3184 those names whose forward lookup yields the correct IP address. Any other
3185 names are discarded. This closes a loophole whereby a rogue DNS
3186 administrator could create reverse DNS records to break through a
3187 wildcarded host restriction in an ACL.
3189 37. If a user filter or a system filter that ran in a subprocess used any of
3190 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3191 the wrong values were passed to the pipe command ($thisaddress had the
3192 value of $0, $0 had the value of $1, etc). This bug was introduced by
3193 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3195 38. Improved the line breaking for long SMTP error messages from ACLs.
3196 Previously, if there was no break point between 40 and 75 characters, Exim
3197 left the rest of the message alone. Two changes have been made: (a) I've
3198 reduced the minimum length to 35 characters; (b) if it can't find a break
3199 point between 35 and 75 characters, it looks ahead and uses the first one
3200 that it finds. This may give the occasional overlong line, but at least the
3201 remaining text gets split now.
3203 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3204 file descriptors might be low, and that setting 1000 would always raise it.
3205 It turns out that in some environments, the limit is already over 1000 and
3206 that lowering it causes trouble. So now Exim takes care not to decrease it.
3208 40. When delivering a message, the value of $return_path is set to $sender_
3209 address at the start of routing (routers may change the value). By an
3210 oversight, this default was not being set up when an address was tested by
3211 -bt or -bv, which affected the outcome if any router or filter referred to
3214 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3215 log only when "message" or "log_message" are set. However, if one of the
3216 conditions was an address verification, or a call to a nested ACL, the
3217 messages generated by the underlying test were being passed through. This
3218 no longer happens. The underlying message is available in $acl_verify_
3219 message for both "message" and "log_message" expansions, so it can be
3220 passed through if needed.
3222 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3223 new expansion $bh_ to give the encoded byte string without charset
3224 translation. Translation happens only if iconv() is available; HAVE_ICONV
3225 indicates this at build time. HEADERS_CHARSET gives the charset to
3226 translate to; headers_charset can change it in the configuration, and
3227 "headers charset" can change it in an individual filter file.
3229 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3230 that creates RFC 2047 encoded "words" labels them as that charset instead
3231 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3232 expansion operator; (ii) when Exim creates a From: line for a local
3233 message; (iii) when a header line is rewritten to include a "phrase" part.
3235 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3236 buggy, causing it to skip the first lines of messages whose message ID
3237 ended in 'D'. This would not have bitten before Exim release 4.14, because
3238 message IDs were unlikely to end in 'D' before then. The effect was to have
3239 incorrect size information for certain domains.
3241 45. #include "config.h" was missing at the start of the crypt16.c module. This
3242 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3245 46. If there was a timeout during a "random" callout check, Exim treated it as
3246 a failure of the random address, and carried on sending RSET and the real
3247 address. If the delay was just some slowness somewhere, the response to the
3248 original RCPT would be taken as a response to RSET and so on, causing
3249 mayhem of various kinds.
3251 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3252 when I implemented it. It didn't allow for the fact that some option values
3253 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3254 the right test for positive values.
3256 48. Domain names in DNS records are case-independent. Exim always looks them up
3257 in lower case. Some resolvers return domain names in exactly the case they
3258 appear in the zone file, that is, they may contain uppercase letters. Not
3259 all resolvers do this - some return always lower case. Exim was treating a
3260 change of case by a resolver as a change of domain, similar to a widening
3261 of a domain abbreviation. This triggered its re-routing code and so it was
3262 trying to route what was effectively the same domain again. This normally
3263 caused routing to fail (because the router wouldn't handle the domain
3264 twice). Now Exim checks for this case specially, and just changes the
3265 casing of the domain that it ultimately uses when it transmits the message
3268 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3271 50. If a filter generated a file delivery with a non-absolute name (possible if
3272 no home directory exists for the router), the forbid_file option was not
3275 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3276 the existing equality matching.
3278 52. Exim was using ints instead of ino_t variables in some places where it was
3279 dealing with inode numbers.
3281 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3282 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3283 called TMPDIR, and if it finds it is different, it changes its value.
3285 54. The smtp_printf() function is now made available to local_scan() so
3286 additional output lines can be written before returning. There is also an
3287 smtp_fflush() function to enable the detection of a dropped connection.
3288 The variables smtp_input and smtp_batched_input are exported to
3291 55. Changed the default runtime configuration: the message "Unknown user"
3292 has been removed from the ACL, and instead placed on the localuser router,
3293 using the cannot_route_message feature. This means that any verification
3294 failures that generate their own messages won't get overridden. Similarly,
3295 the "Unrouteable address" message that was in the ACL for unverifiable
3296 relay addresses has also been removed.
3298 56. Added hosts_avoid_esmtp to the smtp transport.
3300 57. The exicyclog script was not checking for the esoteric option
3301 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3302 will work only if exicyclog is run under the appropriate euid.
3304 58. Following a discussion on the list, the rules by which Exim recognises line
3305 endings on incoming messages have been changed. The -dropcr and drop_cr
3306 options are now no-ops, retained only for backwards compatibility. The
3307 following line terminators are recognized: LF CRLF CR. However, special
3308 processing applies to CR:
3310 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3311 nor a local message in the state where . is a terminator.
3313 (ii) If a bare CR is encountered in a header line, an extra space is added
3314 after the line terminator so as not to end the header. The reasoning
3315 behind this is that bare CRs in header lines are most likely either
3316 to be mistakes, or people trying to play silly games.
3318 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3319 was being incorrectly given as 18 bytes larger than it should have been.
3320 This is a VOB (very old bug).
3322 60. This may never have affected anything current, but just in case it has:
3323 When the local host is found other than at the start of a list of hosts,
3324 the local host, those with the same MX, and any that follow, are discarded.
3325 When the list in question was part of a longer list of hosts, the following
3326 hosts (not currently being processed) were also being discarded. This no
3327 longer happens. I'm not sure if this situation could ever has previously
3330 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3333 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3334 header. This is something that is recommended in a new Internet Draft, and
3335 is something that is documented as being done by Sendmail. There are two
3336 possible values. For messages generated by the autoreply transport, Exim
3339 Auto-Submitted: auto-replied
3341 whereas for all other generated messages (e.g. bounces) it adds
3343 Auto-Submitted: auto-generated
3345 63. The "personal" condition in filters now includes a test for the
3346 Auto-Submitted: header. If it contains the string "auto-" the message it
3347 not considered personal.
3349 64. Added rcpt_include_affixes as a generic transport option.
3351 65. Added queue_only_override (default true).
3353 66. Added the syslog_duplication option.
3355 67. If what should have been the first header line of a message consisted of
3356 a space followed by a colon, Exim was mis-interpreting it as a header line.
3357 It isn't of course - it is syntactically invalid and should therefore be
3358 treated as the start of the message body. The misbehaviour could have
3359 caused a number of strange effects, including loss of data in subsequent
3360 header lines, and spool format errors.
3362 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3363 client host had authenticated. This control can now be exercised by an ACL
3364 for more flexibility.
3366 69. By default, callouts do not happen when testing with -bh. There is now a
3367 variant, -bhc, which does actually run the callout code, including
3368 consulting and updating the callout cache.
3370 70. Added support for saslauthd authentication, courtesy of Alexander
3373 71. If statvfs() failed on the spool or log directories while checking their
3374 size for availability, Exim confusingly gave the error "space shortage".
3375 Furthermore, in debugging mode it crashed with a floating point exception.
3376 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3377 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3378 serious problem, Exim now writes to the main and panic logs when this
3379 happens, with details of the failure. It then refuses to accept the
3380 incoming message, giving the message "spool directory problem" or "log
3381 directory problem" with a 421 code for SMTP messages.
3383 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3384 0, 1, and 2 exist, because some OS complain for execs without them (see
3385 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3386 descriptors. However, the code omitted to check that the open succeeded,
3387 causing mysterious errors if for some reason the permissions on /dev/null
3388 got screwed. Now Exim writes a message to the main and panic logs, and
3389 bombs out if it can't open /dev/null.
3391 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3392 interact so that it is all more flexible. It is supposed to remain
3393 backwards compatible. Also added extra_local_interfaces.
3395 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3396 to bomb out with an assertion failure - to the client this appears as a
3397 connection drop. This problem occurs in the part of the code that was taken
3398 from the Samba project. Fortunately, the assertion is in a very simple
3399 function, so I have fixed this by reproducing the function inline in the
3400 one place where it is called, and arranging for authentication to fail
3401 instead of killing the process with assert().
3403 75. The SPA client code was not working when the server requested OEM rather
3404 than Unicode encoding.
3406 76. Added code to make require_files with a specific uid setting more usable in
3407 the case where statting the file as root fails - usually a non-root-mounted
3408 NFS file system. When this happens and the failure is EACCES, Exim now
3409 forks a subprocess and does the per-uid checking as the relevant uid.
3411 77. Added process_log_path.
3413 78. If log_file_path was not explicitly set, a setting of check_log_space or
3414 check_log_inodes was ignored.
3416 79. If a space check for the spool or log partitions fails, the incident is now
3417 logged. Of course, in the latter case the data may get lost...
3419 80. Added the %p formatting code to string_format() so that it can be used to
3420 print addresses in debug_print(). Adjusted all the address printing in the
3421 debugging in store.c to use %p rather than %d.
3423 81. There was a concern that a line of code in smtp_in.c could overflow a
3424 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3425 initially expressed, the concern was not well-founded, because trailing
3426 spaces are removed early. However, if the trailing spaces were followed by
3427 a NULL, they did not get removed, so the overflow was possible. Two fixes
3430 (a) I re-wrote the offending code in a cleaner fashion.
3431 (b) If an incoming SMTP command contains a NULL character, it is rejected
3434 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3435 runs initgroups(), so that if the Exim user is in any additional groups,
3436 they will be used during message reception.
3442 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob