1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.247 2005/10/11 13:50:48 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.
22 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
23 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
26 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
27 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
29 PH/03 Typo: missing ".o" in src/pcre/Makefile.
31 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
32 header line, restrict the check to what is listed in RFCs 2369 and 2929.
33 Also, for "Auto-Submitted", treat anything other than "no" as
34 non-personal, in accordance with RFC 3834. (Previously it treated
35 anything starting "auto-" as non-personal.)
37 TF/01 The control=submission/name=... option had a problem with syntax
38 errors if the name included a slash character. The /name= option
39 now slurps the rest of the string, so it can include any characters
40 but it must come last in the list of options (after /sender_retain
43 PH/05 Some modifications to the interface to the fake nameserver for the new
50 TK/01 Added the "success_on_redirect" address verification option. See
51 NewStuff for rationale and an example.
53 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
55 PH/02 Patch to exigrep to allow it to work on syslog lines.
57 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
58 fread() to skip over the body file's header line, because in Cygwin the
59 header line is locked and is inaccessible.
61 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
62 co-exist for some time) to make it clear that it is the Exim ID that is
63 referenced, not the Message-ID: header line.
65 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
66 string_format() function, because snprintf() does not exist on all
69 PH/06 The use of forbid_filter_existstest now also locks out the use of the
70 ${stat: expansion item.
72 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
73 protocol synchronization error", to keep the pedants happy.
75 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
76 well as for IRIX systems, when gcc is being used. See the host.c source
79 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
81 PH/10 Named domain lists were not working if used in a queue_smtp_domains
84 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
85 transport and to the smtp transport in LMTP mode.
87 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
89 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
90 run a filter in a subprocess. This could lead to confusion in subsequent
91 lookups in the parent process. There should also be a search_tidyup() at
92 the end of the subprocess.
94 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
95 only if the host matched helo_try_verify_hosts, which caused the
96 verification to occur when the EHLO/HELO command was issued. The ACL just
97 tested the remembered result. Now, if a previous verification attempt has
98 not happened, "verify = helo" does it there and then.
100 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
102 TK/03 Fix log output including CR from clamd.
104 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
105 exist provoked a memory error which could cause a segfault.
107 PH/15 Installed PCRE 6.2
109 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
111 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
112 of the problem. Specifically, suggested +O2 rather than +O1 for the
115 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
117 PH/20 If a delivery was routed to a non-standard port by means of an SRV
118 record, the port was not correctly logged when the outgoing_port log
119 selector was set (it logged the transort's default port).
121 PH/21 Added support for host-specific ports to manualroute, queryprogram,
122 fallback_hosts, and "hosts" in the smtp transport.
124 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
125 host errors such as "Connection refused".
127 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
128 authentication with radiusclient 0.4.9:
130 - Error returned from rc_read_config was caught wrongly
131 - Username/password not passed on to radius server due to wrong length.
133 The presumption is that some radiusclient API changes for 4.51/PH/17
134 were not taken care of correctly. The code is still untested by me (my
135 Linux distribution still has 0.3.2 of radiusclient), but it was
136 contributed by a Radius user.
138 PH/24 When doing a callout, the value of $domain wasn't set correctly when
139 expanding the "port" option of the smtp transport.
141 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
142 while reading a MIME header. Thanks to Tom Hughes for a patch.
144 PH/24 Include config.h inside local_scan.h so that configuration settings are
147 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
148 that in an ACL for RCPT (for example), you can examine exactly what was
151 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
152 commands, but it was not correctly comparing the address with the actual
153 client host address. Thus, it would show the EHLO address in Received:
154 header lines when this was not necessary.
156 PH/27 Added the % operator to ${eval:}.
158 PH/28 Exim tries to create and chdir to its spool directory when it starts;
159 it should be ignoring failures (because with -C, for example, it has lost
160 privilege). It wasn't ignoring creation failures other than "already
163 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
166 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
167 because an input line was too long, either on its own, or by virtue of
168 too many continuations, the temporary file was not being removed, and the
169 return code was incorrect.
171 PH/31 Missing "BOOL" in function definition in filtertest.c.
173 PH/32 Applied Sieve patches from the maintainer.
175 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
177 PH/33 Added "verify = not_blind".
179 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
180 Local/Makefile (with some defaults set). These are used in built scripts
181 such as exicyclog, but they have never been used in the exim_install
182 script (though there are many overriding facilities there). I have
183 arranged that the exim_install script now takes note of these two
186 PH/35 Installed configuration files for Dragonfly.
188 PH/36 When a locally submitted message by a trusted user did not contain a
189 From: header, and the sender address was obtained from -f or from an SMTP
190 MAIL command, and the trusted user did not use -F to supply a sender
191 name, $originator_name was incorrectly used when constructing a From:
192 header. Furthermore, $originator_name was used for submission mode
193 messages from external hosts without From: headers in a similar way,
194 which is clearly wrong.
196 PH/37 Added control=suppress_local_fixups.
198 PH/38 When log_selector = +received_sender was set, and the addition of the
199 sender made the log line's construction buffer exactly full, or one byte
200 less than full, an overflow happened when the terminating "\n" was
203 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
204 when the result of a list match is failure because a DNS lookup failed.
206 PH/40 RM_COMMAND is now used in the building process.
208 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
209 the "build-* directories that it finds.
211 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
212 address in a domain literal was a prefix of an interface address.
214 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
215 when verifying a sender address, unless rewrite_headers is false.
217 PH/44 Wrote a long comment about why errors_to addresses are verified as
218 recipients, not senders.
220 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
221 the ratelimit ACL was added.
223 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
225 PH/46 Added extra information about PostgreSQL errors to the error string.
227 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
228 avoiding the need to install special zones in a real server. This is
229 backwards compatible; if it can't find the fake resolver, it drops back.
230 Thus, both old and new test suites can be run.
232 TF/02 Added util/ratelimit.pl
234 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
235 clock is set back in time.
237 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
238 Candler <B.Candler@pobox.com>.
240 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
241 OK for addresses that are the result of redirections.
243 PH/48 A number of further additions for the benefit of the new test suite,
244 including a fake gethostbyname() that interfaces to the fake DNS resolver
247 TF/06 The fix for widen_domains has also been applied to qualify_single and
248 search_parents which are the other dnslookup options that can cause
251 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
254 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
256 TF/07 Exim produced the error message "an SRV record indicated no SMTP
257 service" if it encountered an MX record with an empty target hostname.
258 The message is now "an MX or SRV record indicated no SMTP service".
260 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
261 if the DNS of the sending site is misconfigured. This is quite a
262 common situation. This change restores the behaviour of treating a
263 helo verification defer as a failure.
265 PH/51 If self=fail was set on a router, the bounce message did not include the
266 actual error message.
272 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
274 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
275 command itself ended in error, the underlying message about the transport
276 filter timeout was being overwritten with the pipe command error. Now the
277 underlying error message should be appended to the second error message.
279 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
281 PH/02 Reduce the amount of output that "make" produces by default. Full output
282 can still be requested.
284 PH/03 The warning log line about a condition test deferring for a "warn" verb
285 was being output only once per connection, rather than after each
286 occurrence (because it was using the same function as for successful
287 "warn" verbs). This seems wrong, so I have changed it.
289 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
290 it should not have, which might have caused a crash in the right
291 circumstances, but probably never did.
293 PH/04 Installed a modified version of Tony Finch's patch to make submission
294 mode fix the return path as well as the Sender: header line, and to
295 add a /name= option so that you can make the user's friendly name appear
298 TF/03 Added the control = fakedefer ACL modifier.
300 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
301 Mark Lowes for thorough testing.
303 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
305 TK/03 Merged latest SRS patch from Miles Wilton.
307 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
308 with the definition in sysexits.h (which is #included earlier).
309 Fortunately, Exim does not actually use EX_OK. The code used to try to
310 preserve the sysexits.h value, by assumimg that macro definitions were
311 scanned for macro replacements. I have been disabused of this notion,
312 so now the code just undefines EX_OK before #including unistd.h.
314 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
315 in the smtp transport. When a block could not be written in a single
316 write() function, the timeout was being re-applied to each part-write.
317 This seems wrong - if the receiver was accepting one byte at a time it
318 would take for ever. The timeout is now adjusted when this happens. It
319 doesn't have to be particularly precise.
321 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
322 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
324 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
325 without a domain if the check to <postmaster@domain> fails.
327 SC/01 Eximstats: added -xls and the ability to specify output files
328 (patch written by Frank Heydlauf).
330 SC/02 Eximstats: use FileHandles for outputing results.
332 SC/03 Eximstats: allow any combination of xls, txt, and html output.
334 SC/04 Eximstats: fixed display of large numbers with -nvr option
336 SC/05 Eximstats: fixed merging of reports with empty tables.
338 SC/06 Eximstats: added the -include_original_destination flag
340 SC/07 Eximstats: removed tabs and trailing whitespace.
342 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
344 TK/06 MBOX spool code: Add real "From " MBOX separator line
345 so the .eml file is really in mbox format (even though
346 most programs do not really care). Patch from Alex Miller.
348 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
349 The latter is generated from $received_to and is only set if the
350 message has one envelope recipient. SA can use these headers,
351 obviously out-of-the-box. Patch from Alex Miller.
353 PH/08 The ${def test on a variable was returning false if the variable's
354 value was "0", contrary to what the specification has always said!
355 The result should be true unless the variable is empty.
357 PH/09 The syntax error of a character other than { following "${if
358 def:variable_name" (after optional whitespace) was not being diagnosed.
359 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
360 accidental colon was present, for example, could give incorrect results.
362 PH/10 Tidied the code in a number of places where the st_size field of a stat()
363 result is used (not including appendfile, where other changes are about
366 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
367 This involved changing a lot of size variables from int to off_t. It
368 should work with maildirs and everything.
370 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
371 spamd dying while we are connected to it.
373 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
376 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
377 function for BDB 4.3.
379 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
381 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
382 the sources for PCRE have been re-arranged and more files are now
385 PH/15 The code I had for printing potentially long long variables in PH/11
386 above was not the best (it lost precision). The length of off_t variables
387 is now inspected at build time, and an appropriate printing format (%ld
388 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
389 to be "long long int" or "long int". This is needed for the internal
390 formatting function string_vformat().
392 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
393 the configuration file to be ":syslog", then the script "guesses" where
394 the logs files are, rather than using the compiled in default. In our
395 case the guess is not the same as the compiled default, so the script
396 suddenly stopped working when I started to use syslog. The patch checks
397 to see if log_file_path is "". If so, it attempts to read it from exim
398 with no configuration file to get the compiled in version, before it
399 falls back to the previous guessing code."
401 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
402 implementing BATV in an Exim configuration. See NewStuff for the gory
405 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
406 Makefile that are specific to HP-UX.
408 PH/18 If the "use_postmaster" option was set for a recipient callout together
409 with the "random" option, the postmaster address was used as the MAIL
410 FROM address for the random test, but not for the subsequent recipient
411 test. It is now used for both.
413 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
414 patch removes a few documentation additions to RFC 3028, because the
415 latest draft now contains them. It adds the new en;ascii-case comparator
416 and a new error check for 8bit text in MIME parts. Comparator and
417 require names are now matched exactly. I enabled the subaddress
418 extension, but it is not well tested yet (read: it works for me)."
420 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
421 rework some of the code of TK/09 above to avoid the hardwired use of
422 "%lld" and "long long". Replaced the call to snprintf() with a call to
425 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
426 records point to non-existent hosts", "retry timeout exceeded", and
427 "retry time not reached for any host after a long failure period".
429 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
430 experimental DomainKeys support:
432 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
433 (2) On an error such as an illegally used "control", the wrong name for
434 the control was given.
436 These problems did NOT occur unless DomainKeys support was compiled.
438 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
440 PH/24 Added ${if match_ip condition.
442 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
443 databases so that it will be absolutely obvious if a crash occurs in the
444 DB library. This is a regular occurrence (often caused by mis-matched
447 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
448 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
449 picked up on a user's system that detects such things. There doesn't seem
450 to be a gcc warning option for this - only an attribute that has to be
451 put on the function's prototype. It seems that in Fedora Core 4 they have
452 set this on a number of new functions. No doubt there will be more in due
455 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
456 specify a transport. However, if an address that was verified by such a
457 router was the subject of a callout, Exim crashed because it tried to
458 read the rcpt_include_affixes from the non-existent transport. Now it
459 just assumes that the setting of that option is false. This bug was
460 introduced by 4.51/PH/31.
462 PH/28 Changed -d+all to exclude +memory, because that information is very
463 rarely of interest, but it makes the output a lot bigger. People tend to
464 do -d+all out of habit.
466 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
467 code in os-type was giving problems when libc.so lives in lib64, like on
470 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
471 aren't the modern standard, and it seems that some systems' include files
472 don't always have them. Exim was already checking for some of the newer
473 ones like T_AAAA, and defining it itself. I've added checks for all the
474 record types that Exim uses.
476 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
477 not automatically generating a new one, as it is supposed to. This
478 prevented TLS from working. If the file did exist, but contained invalid
479 data, a new version was generated, as expected. It was only the case of a
480 non-existent file that was broken.
482 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
483 with a change in libdomainkeys > 0.64.
485 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
486 from DNS. If the selector record carries the flag, it now has
487 precedence over the domain-wide flag.
489 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
491 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
492 the use of a port name that isn't defined in /etc/services) occurred, the
493 message was deferred as in a normal delivery, and thus remained on the
494 spool, instead of being failed because of the mua_wrapper setting. This
495 is now fixed, and I tidied up some of the mua_wrapper messages at the
498 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
499 the messages in a hash of arrays rather than using individual hashes.
500 This is a bit cleaner and results in dramatic memory savings, albeit
501 at a slight CPU cost.
503 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
504 as requested by Marc Sherman.
506 SC/10 Eximstats: added histograms for user specified patterns as requested
509 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
511 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
512 fopen() in the content-scanning modules that did not already have it.
518 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
519 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
521 TK/02 Fix ACL "control" statement not being available in MIME ACL.
523 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
525 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
526 to test Sieve filters that use "vacation".
528 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
529 that changes the way the GnuTLS parameters are stored in the cache file.
530 The new format can be generated externally. For backward compatibility,
531 if the data in the cache doesn't make sense, Exim assumes it has read an
532 old-format file, and it generates new data and writes a new file. This
533 means that you can't go back to an older release without removing the
536 PH/03 A redirect router that has both "unseen" and "one_time" set does not
537 work if there are any delivery delays because "one_time" forces the
538 parent to be marked "delivered", so its unseen clone is never tried
539 again. For this reason, Exim now forbids the simultaneous setting of
542 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
543 redirected to themselves ("homonym" addresses). Read the long ChangeLog
544 entry if you want to know the details. The fix, however, neglected to
545 consider the case when local delivery batching is involved. The test for
546 "previously delivered" was not happening when checking to see if an
547 address could be batched with a previous (undelivered) one; under
548 certain circumstances this could lead to multiple deliveries to the same
551 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
552 in its include files, and this causes problems building Exim.
554 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
555 header_syntax) but Exim was just ignoring anything given after a slash.
556 In particular, this caused confusion with an attempt to use "verify =
557 reverse_host_lookup/defer_ok". An error is now given when options are
558 supplied for verify items that do not have them. (Maybe reverse_host_
559 lookup should have a defer_ok option, but that's a different point.)
561 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
562 defined by RFC 821) to 2048, because there were problems with some AUTH
563 commands, and RFC 1869 says the size should be increased for extended
564 SMTP commands that take arguments.
566 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
569 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
570 "unknown" error; now it says that the functionality isn't in the binary.
572 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
573 an address' error message when a string expansion fails (syntax or
574 whatever). Otherwise the password may appear in the log. Following change
575 PH/42 below, there is no longer a chance of it appearing in a bounce
578 PH/11 Installed exipick version 20050225.0 from John Jetmore.
580 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
581 of its addresses was ever tried. (Bugzilla bug #2.)
583 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
584 the result incorrectly in the debug output. (It correctly added a newline
585 to what was transported.)
587 TF/01 Added $received_time.
589 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
590 commented out examples of how to interface to a virus scanner and to
591 SpamAssassin. Also added commented examples of av_scanner and
592 spamd_address settings.
594 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
595 and controls are allowed in which ACLs. There were a couple of minor
596 errors. Some of the entries in the conditions table (which is a table of
597 where they are NOT allowed) were getting very unwieldy; rewrote them as a
598 negation of where the condition IS allowed.
600 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
602 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
603 header file does not have a version number, so I've had to invent a new
604 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
605 API. The code is untested by me (my Linux distribution still has 0.3.2 of
606 radiusclient), but it was contributed by a Radius user.
608 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
609 files or directories, for OpenSSL.
611 PH/19 When an Exim process that is running as root has to create an Exim log
612 file, it does so in a subprocess that runs as exim:exim so as to get the
613 ownership right at creation (otherwise, other Exim processes might see
614 the file with the wrong ownership). There was no test for failure of this
615 fork() call, which would lead to the process getting stuck as it waited
616 for a non-existent subprocess. Forks do occasionally fail when resources
617 run out. I reviewed all the other calls to fork(); they all seem to check
620 PH/20 When checking for unexpected SMTP input at connect time (before writing
621 the banner), Exim was not dealing correctly with a non-positive return
622 from the read() function. If the client had disconnected by this time,
623 the result was a log entry for a synchronization error with an empty
624 string after "input=" when read() returned zero. If read() returned -1
625 (an event I could not check), uninitialized data bytes were printed.
626 There were reports of junk text (parts of files, etc) appearing after
629 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
631 PH/22 Added support for macro redefinition, and (re)definition in between
632 driver and ACL definitions.
634 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
635 forgetting to use the resulting value; it was using the unexpanded value.
637 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
638 hadn't been configured. The fix is from Juergen Kreileder, who
639 understands it better than I do:
641 "Here's what I see happening with three configured cyrus_sasl
642 authenticators configured (plain, login, cram-md5):
644 On startup auth_cyrus_sasl_init() gets called for each of these.
645 This means three calls to sasl_listmech() without a specified mech_list.
646 => SASL tests which mechs of all available mechs actually work
647 => three warnings about OTP not working
648 => the returned list contains: plain, login, cram-md5, digest-md5, ...
650 With the patch, sasl_listmech() also gets called three times. But now
651 SASL's mech_list option is set to the server_mech specified in the the
652 authenticator. Or in other words, the answer from sasl_listmech()
653 gets limited to just the mech you're testing for (which is different
655 => the return list contains just 'plain' or 'login', 'cram-md5' or
656 nothing depending on the value of ob->server_mech.
658 I've just tested the patch: Authentication still works fine,
659 unavailable mechs specified in the exim configuration are still
660 caught, and the auth.log warnings about OTP are gone."
662 PH/25 When debugging is enabled, the contents of the command line are added
663 to the debugging output, even when log_selector=+arguments is not
666 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
667 answer is "GNU", and only if the return is "GNU/something" is the answer
670 PH/27 $acl_verify_message is now set immediately after the failure of a
671 verification in an ACL, and so is available in subsequent modifiers. In
672 particular, the message can be preserved by coding like this:
674 warn !verify = sender
675 set acl_m0 = $acl_verify_message
677 Previously, $acl_verify_message was set only while expanding "message"
678 and "log_message" when a very denied access.
680 PH/28 Modified OS/os.c-Linux with
682 -#ifndef OS_LOAD_AVERAGE
683 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
685 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
686 nomenclature these days.)
688 PH/29 Installed patch from the Sieve maintainer that adds the options
689 sieve_useraddress and sieve_subaddress to the redirect router.
691 PH/30 In these circumstances:
692 . Two addresses routed to the same list of hosts;
693 . First host does not offer TLS;
694 . First host accepts first address;
695 . First host gives temporary error to second address;
696 . Second host offers TLS and a TLS session is established;
697 . Second host accepts second address.
698 Exim incorrectly logged both deliveries with the TLS parameters (cipher
699 and peerdn, if requested) that were in fact used only for the second
702 PH/31 When doing a callout as part of verifying an address, Exim was not paying
703 attention to any local part prefix or suffix that was matched by the
704 router that accepted the address. It now behaves in the same way as it
705 does for delivery: the affixes are removed from the local part unless
706 rcpt_include_affixes is set on the transport.
708 PH/32 Add the sender address, as F=<...>, to the log line when logging a
709 timeout during the DATA phase of an incoming message.
711 PH/33 Sieve envelope tests were broken for match types other than :is. I have
712 applied a patch sanctioned by the Sieve maintainer.
714 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
715 the uid or gid is negative. A case of a negative gid caused this to be
716 noticed. The fix allows for either to be negative.
718 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
719 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
722 PH/36 The change PH/12 above was broken. Fixed it.
724 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
725 the grounds that routing the same address twice would always produce the
726 same answer. This might have been true once, but it is certainly no
727 longer true now. Routing a child address may depend on the previous
728 routing that produced that child. Some complicated redirection strategies
729 went wrong when messages had multiple recipients, and made Exim's
730 behaviour dependent on the order in which the addresses were given.
732 I have moved the duplicate checking until after the routing is complete.
733 Exim scans the addresses that are assigned to local and remote
734 transports, and removes any duplicates. This means that more work will be
735 done, as duplicates will always all be routed, but duplicates are
736 presumably rare, so I don't expect this is of any significance.
738 For deliveries to pipes, files, and autoreplies, the duplicate checking
739 still happens during the routing process, since they are not going to be
742 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
743 It corrects a timeout issue with spamd. This is Ian's comment: "The
744 background is that sometimes spamd either never reads data from a
745 connection it has accepted, or it never writes response data. The exiscan
746 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
747 blindly assumes that writes won't block so it may never time out."
749 PH/39 Allow G after quota size as well as K and M.
751 PH/40 The value set for $authenticated_id in an authenticator may not contain
752 binary zeroes or newlines because the value is written to log lines and
753 to spool files. There was no check on this. Now the value is run through
754 the string_printing() function so that such characters are converted to
755 printable escape sequences.
757 PH/41 $message_linecount is a new variable that contains the total number of
758 lines in the message. Compare $body_linecount, which is the count for the
761 PH/42 Exim no longer gives details of delivery errors for specific addresses in
762 bounce and delay warning messages, except in certain special cases, which
765 (a) An SMTP error message from a remote host;
766 (b) A message specified in a :fail: redirection;
767 (c) A message specified in a "fail" command in a system filter;
768 (d) A message specified in a FAIL return from the queryprogram router;
769 (e) A message specified by the cannot_route_message router option.
771 In these cases only, Exim does include the error details in bounce and
772 warning messages. There are also a few cases where bland messages such
773 as "unrouteable address" or "local delivery error" are given.
775 PH/43 $value is now also set for the "else" part of a ${run expansion.
777 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
778 being worked on, but at least Exim now implements the latest version to
781 PH/45 In a pipe transport, although a timeout while waiting for the pipe
782 process to complete was treated as a delivery failure, a timeout while
783 writing the message to the pipe was logged, but erroneously treated as a
784 successful delivery. Such timeouts include transport filter timeouts. For
785 consistency with the overall process timeout, these timeouts are now
786 treated as errors, giving rise to delivery failures by default. However,
787 there is now a new Boolean option for the pipe transport called
788 timeout_defer, which, if set TRUE, converts the failures into defers for
789 both kinds of timeout. A transport filter timeout is now identified in
792 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
793 systems where "make" and "gmake" are different, calling "gmake" at top
794 level broke things. I've arranged for the value of $(MAKE) to be passed
795 from the Makefile to this script so that it can call the same version of
799 A note about Exim versions 4.44 and 4.50
800 ----------------------------------------
802 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
803 changes of various kinds. As a consequence, a big documentation update was
804 needed. This delayed the release for rather longer than seemed good, especially
805 in the light of a couple of (minor) security issues. Therefore, the changes
806 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
807 release. So 4.44 and 4.50 are in effect two different branches that both start
810 I have left the 4.50 change log unchanged; it contains all the changes since
811 4.43. The change log for 4.44 is below; many of its items are identical to
812 those for 4.50. This seems to be the most sensible way to preserve the
813 historical information.
819 1. Minor wording change to the doc/README.SIEVE file.
821 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
822 computation of the current number of files was incorrect.
824 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
825 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
828 4. Give more explanation in the error message when the command for a transport
829 filter fails to execute.
831 5. There are several places where Exim runs a non-Exim command in a
832 subprocess. The SIGUSR1 signal should be disabled for these processes. This
833 was being done only for the command run by the queryprogram router. It is
834 now done for all such subprocesses. The other cases are: ${run, transport
835 filters, and the commands run by the lmtp and pipe transports.
837 6. Added CONFIGURE_GROUP build-time option.
839 7. Some older OS have a limit of 256 on the maximum number of file
840 descriptors. Exim was using setrlimit() to set 1000 as a large value
841 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
842 systems. I've change it so that if it can't get 1000, it tries for 256.
844 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
845 was an oversight, and furthermore, ever since the addition of extra
846 controls (e.g. 4.43/32), the checks on when to allow different forms of
847 "control" were broken. There should now be diagnostics for all cases when a
848 control that does not make sense is encountered.
850 9. Added the /retain_sender option to "control=submission".
852 10. $recipients is now available in the predata ACL (oversight).
854 11. Tidy the search cache before the fork to do a delivery from a message
855 received from the command line. Otherwise the child will trigger a lookup
856 failure and thereby defer the delivery if it tries to use (for example) a
857 cached ldap connection that the parent has called unbind on.
859 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
860 of $address_data from the recipient verification was clobbered by the
863 13. The value of address_data from a sender verification is now available in
864 $sender_address_data in subsequent conditions in the ACL statement.
866 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
868 15. Added a new option "connect=<time>" to callout options, to set a different
871 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
872 was its contents. (It was OK if the option was not defined at all.)
874 17. A "Completed" log line is now written for messages that are removed from
875 the spool by the -Mrm option.
877 18. New variables $sender_verify_failure and $recipient_verify_failure contain
878 information about exactly what failed.
880 19. Added -dd to debug only the daemon process.
882 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
883 handles timeouts, both on the server side and network timeouts. Renamed the
884 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
886 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
887 It is now set to "smtps".
889 22. $host_address is now set to the target address during the checking of
892 23. When checking ignore_target_hosts for an ipliteral router, no host name was
893 being passed; this would have caused $sender_host_name to have been used if
894 matching the list had actually called for a host name (not very likely,
895 since this list is usually IP addresses). A host name is now passed as
898 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
899 code that specifies a non-restarting handler (typically sigaction() in
900 modern systems) in an attempt to fix a rare and obscure crash bug.
902 25. Narrowed the window for a race in the daemon that could cause it to ignore
903 SIGCHLD signals. This is not a major problem, because they are used only to
904 wake it up if nothing else does.
906 26. A malformed maildirsize file could cause Exim to calculate negative values
907 for the mailbox size or file count. Odd effects could occur as a result.
908 The maildirsize information is now recalculated if the size or filecount
911 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
912 support for a long time. Removed HAVE_SYS_VFS_H.
914 28. Installed the latest version of exipick from John Jetmore.
916 29. In an address list, if the pattern was not a regular expression, an empty
917 subject address (from a bounce message) matched only if the pattern was an
918 empty string. Non-empty patterns were not even tested. This was the wrong
919 because it is perfectly reasonable to use an empty address as part of a
920 database query. An empty address is now tested by patterns that are
921 lookups. However, all the other forms of pattern expect the subject to
922 contain a local part and a domain, and therefore, for them, an empty
923 address still always fails if the pattern is not itself empty.
925 30. Exim went into a mad DNS loop when attempting to do a callout where the
926 host was specified on an smtp transport, and looking it up yielded more
929 31. Re-factored the code for checking spool and log partition space into a
930 function that finds that data and another that does the check. The former
931 is then used to implement four new variables: $spool_space, $log_space,
932 $spool_inodes, and $log_inodes.
934 32. The RFC2047 encoding function was originally intended for short strings
935 such as real names; it was not keeping to the 75-character limit for
936 encoded words that the RFC imposes. It now respects the limit, and
937 generates multiple encoded words if necessary. To be on the safe side, I
938 have increased the buffer size for the ${rfc2047: expansion operator from
941 33. It is now permitted to omit both strings after an "if" condition; if the
942 condition is true, the result is "true". As before, when the second string
943 is omitted, a false condition yields an empty string. This makes it less
944 cumbersome to write custom ACL and router conditions.
946 34. Failure to deliver a bounce message always caused it to be frozen, even if
947 there was an errors_to setting on the router. The errors_to setting is now
950 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
951 canonical form (fully expanded) before being placed in
952 $sender_host_address.
954 36. The table in the code that translates DNS record types into text (T_A to
955 "A" for instance) was missing entries for NS and CNAME. It is just possible
956 that this could have caused confusion if both these types were looked up
957 for the same domain, because the text type is used as part of Exim's
958 per-process caching. But the chance of anyone hitting this buglet seems
961 37. The dnsdb lookup has been extended in a number of ways.
963 (1) There is a new type, "zns", which walks up the domain tree until it
964 finds some nameserver records. It should be used with care.
966 (2) There is a new type, "mxh", which is like "mx" except that it returns
967 just the host names, not the priorities.
969 (3) It is now possible to give a list of domains (or IP addresses) to be
970 looked up. The behaviour when one of the lookups defers can be
971 controlled by a keyword.
973 (4) It is now possible to specify the separator character for use when
974 multiple records are returned.
976 38. The dnslists ACL condition has been extended: it is now possible to supply
977 a list of IP addresses and/or domains to be looked up in a particular DNS
980 39. Added log_selector=+queue_time_overall.
982 40. When running the queue in the test harness, wait just a tad after forking a
983 delivery process, to get repeatability of debugging output.
985 41. Include certificate and key file names in error message when GnuTLS fails
986 to set them up, because the GnuTLS error message doesn't include the name
987 of the failing file when there is a problem reading it.
989 42. Allow both -bf and -bF in the same test run.
991 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
993 44. The "Exiscan patch" is now merged into the mainline Exim source.
995 45. Sometimes the final signoff response after QUIT could fail to get
996 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
997 < 0 before doing a fflush(). This bug looks as though it goes back to the
998 introduction of TLS in release 3.20, but "sometimes" must have been rare
999 because the tests only now provoked it.
1001 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1002 (this can affect the format of dates).
1004 47. exim_tidydb, when checking for the continued existence of a message for
1005 which it has found a message-specific retry record, was not finding
1006 messages that were in split spool directories. Consequently, it was
1007 deleting retry records that should have stayed in existence.
1009 48. Steve fixed some bugs in eximstats.
1011 49. The SPA authentication driver was not abandoning authentication and moving
1012 on to the next authenticator when an expansion was forced to fail,
1013 contradicting the general specification for all authenticators. Instead it
1014 was generating a temporary error. It now behaves as specified.
1016 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1017 (the order specifies the preference for clients). The order is now AES256,
1018 AES128, 3DES, ARCFOUR128.
1020 51. Small patch to Sieve code - explicitly set From: when generating an
1023 52. Exim crashed if a remote delivery caused a very long error message to be
1024 recorded - for instance if somebody sent an entire SpamAssassin report back
1025 as a large number of 550 error lines. This bug was coincidentally fixed by
1026 increasing the size of one of Exim's internal buffers (big_buffer) that
1027 happened as part of the Exiscan merge. However, to be on the safe side, I
1028 have made the code more robust (and fixed the comments that describe what
1031 53. Now that there can be additional text after "Completed" in log lines (if
1032 the queue_time_overall log selector is set), a one-byte patch to exigrep
1033 was needed to allow it to recognize "Completed" as not the last thing in
1036 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1037 patch that reportedly fixes this has been added. I am not expert enough to
1038 create a test for it. This is what the patch creator wrote:
1040 "I found a little strange behaviour of ldap code when working with
1041 Windows 2003 AD Domain, where users was placed in more than one
1042 Organization Units. When I tried to give exim partial DN, the exit code
1043 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1044 But simultaneously result of request was absolutely normal ldap result,
1045 so I produce this patch..."
1047 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1048 REFERENCE, so I have modified the code to exclude the patch when that macro
1051 55. Some experimental protocols are using DNS PTR records for new purposes. The
1052 keys for these records are domain names, not reversed IP addresses. The
1053 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1054 leaves it alone. Component reversal etc. now happens only for IP addresses.
1057 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1059 57. Double the size of the debug message buffer (to 2048) so that more of very
1060 long debug lines gets shown.
1062 58. The exicyclog utility now does better if the number of log files to keep
1063 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1065 59. Two changes related to the smtp_active_hostname option:
1067 (1) $smtp_active_hostname is now available as a variable.
1068 (2) The default for smtp_banner uses $smtp_active_hostname instead
1069 of $primary_hostname.
1071 60. The host_aton() function is supposed to be passed a string that is known
1072 to be a valid IP address. However, in the case of IPv6 addresses, it was
1073 not checking this. This is a hostage to fortune. Exim now panics and dies
1074 if the condition is not met. A case was found where this could be provoked
1075 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1076 components; fortuitously, this particular loophole had already been fixed
1077 by change 4.50/55 above.
1079 If there are any other similar loopholes, the new check in host_aton()
1080 itself should stop them being exploited. The report I received stated that
1081 data on the command line could provoke the exploit when Exim was running as
1082 exim, but did not say which command line option was involved. All I could
1083 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1084 running as the user.
1087 61. There was a buffer overflow vulnerability in the SPA authentication code
1088 (which came originally from the Samba project). I have added a test to the
1089 spa_base64_to_bits() function which I hope fixes it.
1092 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1093 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1095 63. The daemon start-up calls getloadavg() while still root for those OS that
1096 need the first call to be done as root, but it missed one case: when
1097 deliver_queue_load_max is set with deliver_drop_privilege. This is
1098 necessary for the benefit of the queue runner, because there is no re-exec
1099 when deliver_drop_privilege is set.
1101 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1102 This has been fixed.
1104 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1105 was in use, was not putting the data itself into the right store pool;
1106 consequently, it could be overwritten for a subsequent message in the same
1107 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1110 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1112 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1113 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1114 the function were treating the return as a boolean value, which happened to
1115 work because 0=false and not-0=true, but is not correct code.
1117 68. The host_aton() function was not handling scoped IPv6 addresses (those
1118 with, for example, "%eth0" on the end) correctly.
1120 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1121 negated items (that is, ~something) in unsigned ints. Some compilers
1122 apparently mutter when there is no cast.
1124 70. If an address verification called from an ACL failed, and did not produce a
1125 user-specific message (i.e. there was only a "system" message), nothing was
1126 put in $acl_verify_message. In this situation, it now puts the system
1129 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1130 session; change 4.31/43 added the unwanted input to the log line - except
1131 that it did not do this in the start of session case. It now does.
1133 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1134 This is wrong and can cause the other end to generate a synchronization
1135 error if it is another Exim or anything else that does the synchronization
1136 check. A QUIT command is no longer sent after a timeout.
1138 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1139 during host lookups.
1141 74. The defer_ok option of callout verification was not working if it was used
1142 when verifying addresses in header lines, that is, for this case:
1144 verify = header_sender/callout=defer_ok
1146 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1147 those file descriptors could be used for SMTP connections. If anything
1148 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1149 could be sent to the SMTP client, causing chaos. The daemon now opens
1150 stdin, stdout, and stderr to /dev/null when it puts itself into the
1153 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1154 log by default. The user can override this with suitable Perl magic.
1156 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1157 the log message when discard triggers, was not working for the DATA ACL or
1158 for the non-SMTP ACL.
1160 78. Error message wording change in sieve.c.
1162 79. If smtp_accept_max_per_host was set, the number of connections could be
1163 restricted to fewer than expected, because the daemon was trying to set up
1164 a new connection before checking whether the processes handling previous
1165 connections had finished. The check for completed processes is now done
1166 earlier. On busy systems, this bug wouldn't be noticed because something
1167 else would have woken the daemon, and it would have reaped the completed
1170 80. If a message was submitted locally by a user whose login name contained one
1171 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1172 It caused a spool format error. I have fixed the spool reading code. A
1173 related problem was that the "from" clause in the Received: line became
1174 illegal because of the space(s). It is now covered by ${quote_local_part.
1176 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1179 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1181 83. Patch from Sieve maintainer to fix unterminated string problem in
1182 "vacation" handling.
1184 84. Some minor changes to the Linux configuration files to help with other
1185 OS variants using glibc.
1187 85. One more patch for Sieve to update vacation handling to latest spec.
1190 ----------------------------------------------------
1191 See the note above about the 4.44 and 4.50 releases.
1192 ----------------------------------------------------
1198 1. Change 4.43/35 introduced a bug that caused file counts to be
1199 incorrectly computed when quota_filecount was set in an appendfile
1202 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1203 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1206 3. Give more explanation in the error message when the command for a transport
1207 filter fails to execute.
1209 4. There are several places where Exim runs a non-Exim command in a
1210 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1211 was being done only for the command run by the queryprogram router. It is
1212 now done for all such subprocesses. The other cases are: ${run, transport
1213 filters, and the commands run by the lmtp and pipe transports.
1215 5. Some older OS have a limit of 256 on the maximum number of file
1216 descriptors. Exim was using setrlimit() to set 1000 as a large value
1217 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1218 systems. I've change it so that if it can't get 1000, it tries for 256.
1220 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1221 was an oversight, and furthermore, ever since the addition of extra
1222 controls (e.g. 4.43/32), the checks on when to allow different forms of
1223 "control" were broken. There should now be diagnostics for all cases when a
1224 control that does not make sense is encountered.
1226 7. $recipients is now available in the predata ACL (oversight).
1228 8. Tidy the search cache before the fork to do a delivery from a message
1229 received from the command line. Otherwise the child will trigger a lookup
1230 failure and thereby defer the delivery if it tries to use (for example) a
1231 cached ldap connection that the parent has called unbind on.
1233 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1234 of $address_data from the recipient verification was clobbered by the
1235 sender verification.
1237 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1238 was its contents. (It was OK if the option was not defined at all.)
1240 11. A "Completed" log line is now written for messages that are removed from
1241 the spool by the -Mrm option.
1243 12. $host_address is now set to the target address during the checking of
1244 ignore_target_hosts.
1246 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1247 being passed; this would have caused $sender_host_name to have been used if
1248 matching the list had actually called for a host name (not very likely,
1249 since this list is usually IP addresses). A host name is now passed as
1252 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1253 code that specifies a non-restarting handler (typically sigaction() in
1254 modern systems) in an attempt to fix a rare and obscure crash bug.
1256 15. Narrowed the window for a race in the daemon that could cause it to ignore
1257 SIGCHLD signals. This is not a major problem, because they are used only to
1258 wake it up if nothing else does.
1260 16. A malformed maildirsize file could cause Exim to calculate negative values
1261 for the mailbox size or file count. Odd effects could occur as a result.
1262 The maildirsize information is now recalculated if the size or filecount
1265 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1266 support for a long time. Removed HAVE_SYS_VFS_H.
1268 18. Updated exipick to current release from John Jetmore.
1270 19. Allow an empty sender to be matched against a lookup in an address list.
1271 Previously the only cases considered were a regular expression, or an
1274 20. Exim went into a mad DNS lookup loop when doing a callout where the
1275 host was specified on the transport, if the DNS lookup yielded more than
1278 21. The RFC2047 encoding function was originally intended for short strings
1279 such as real names; it was not keeping to the 75-character limit for
1280 encoded words that the RFC imposes. It now respects the limit, and
1281 generates multiple encoded words if necessary. To be on the safe side, I
1282 have increased the buffer size for the ${rfc2047: expansion operator from
1285 22. Failure to deliver a bounce message always caused it to be frozen, even if
1286 there was an errors_to setting on the router. The errors_to setting is now
1289 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1290 canonical form (fully expanded) before being placed in
1291 $sender_host_address.
1293 24. Updated eximstats to version 1.33
1295 25. Include certificate and key file names in error message when GnuTLS fails
1296 to set them up, because the GnuTLS error message doesn't include the name
1297 of the failing file when there is a problem reading it.
1299 26. Expand error message when OpenSSL has problems setting up cert/key files.
1302 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1303 (this can affect the format of dates).
1305 28. exim_tidydb, when checking for the continued existence of a message for
1306 which it has found a message-specific retry record, was not finding
1307 messages that were in split spool directories. Consequently, it was
1308 deleting retry records that should have stayed in existence.
1310 29. eximstats updated to version 1.35
1311 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1312 1.35 - bugfix such that pie charts by volume are generated correctly
1314 30. The SPA authentication driver was not abandoning authentication and moving
1315 on to the next authenticator when an expansion was forced to fail,
1316 contradicting the general specification for all authenticators. Instead it
1317 was generating a temporary error. It now behaves as specified.
1319 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1320 (the order specifies the preference for clients). The order is now AES256,
1321 AES128, 3DES, ARCFOUR128.
1323 31. Small patch to Sieve code - explicitly set From: when generating an
1326 32. Exim crashed if a remote delivery caused a very long error message to be
1327 recorded - for instance if somebody sent an entire SpamAssassin report back
1328 as a large number of 550 error lines. This bug was coincidentally fixed by
1329 increasing the size of one of Exim's internal buffers (big_buffer) that
1330 happened as part of the Exiscan merge. However, to be on the safe side, I
1331 have made the code more robust (and fixed the comments that describe what
1334 33. Some experimental protocols are using DNS PTR records for new purposes. The
1335 keys for these records are domain names, not reversed IP addresses. The
1336 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1337 leaves it alone. Component reversal etc. now happens only for IP addresses.
1340 34. The host_aton() function is supposed to be passed a string that is known
1341 to be a valid IP address. However, in the case of IPv6 addresses, it was
1342 not checking this. This is a hostage to fortune. Exim now panics and dies
1343 if the condition is not met. A case was found where this could be provoked
1344 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1345 components; fortuitously, this particular loophole had already been fixed
1346 by change 4.50/55 or 4.44/33 above.
1348 If there are any other similar loopholes, the new check in host_aton()
1349 itself should stop them being exploited. The report I received stated that
1350 data on the command line could provoke the exploit when Exim was running as
1351 exim, but did not say which command line option was involved. All I could
1352 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1353 running as the user.
1356 35. There was a buffer overflow vulnerability in the SPA authentication code
1357 (which came originally from the Samba project). I have added a test to the
1358 spa_base64_to_bits() function which I hope fixes it.
1361 36. The daemon start-up calls getloadavg() while still root for those OS that
1362 need the first call to be done as root, but it missed one case: when
1363 deliver_queue_load_max is set with deliver_drop_privilege. This is
1364 necessary for the benefit of the queue runner, because there is no re-exec
1365 when deliver_drop_privilege is set.
1367 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1368 was in use, was not putting the data itself into the right store pool;
1369 consequently, it could be overwritten for a subsequent message in the same
1370 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1373 38. Sometimes the final signoff response after QUIT could fail to get
1374 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1375 < 0 before doing a fflush(). This bug looks as though it goes back to the
1376 introduction of TLS in release 3.20, but "sometimes" must have been rare
1377 because the tests only now provoked it.
1383 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1384 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1385 Now it returns an integer. A number of calls were still expecting a T/F
1386 return. Fortuitously, in all cases, the tests worked in OK situations,
1387 which is the norm. However, things would have gone wrong on any write
1388 failures on the smtp file descriptor. This function is used when sending
1389 messages over SMTP and also when doing verify callouts.
1391 2. When Exim is called to do synchronous delivery of a locally submitted
1392 message (the -odf or -odi options), it no longer closes stderr before doing
1395 3. Implemented the mua_wrapper option.
1397 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1399 5. Implemented the functions header_remove(), header_testname(),
1400 header_add_at_position(), and receive_remove_recipient(), and exported them
1403 6. If an ACL "warn" statement specified the addition of headers, Exim already
1404 inserted X-ACL-Warn: at the start if there was no header name. However, it
1405 was not making this test for the second and subsequent header lines if
1406 there were newlines in the string. This meant that an invalid header could
1407 be inserted if Exim was badly configured.
1409 7. Allow an ACL "warn" statement to add header lines at the start or after all
1410 the Received: headers, as well as at the end.
1412 8. Added the rcpt_4xx retry error code.
1414 9. Added postmaster_mailfrom=xxx to callout verification option.
1416 10. Added mailfrom=xxxx to the callout verification option, for verify=
1419 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1420 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1422 12. Inserted some casts to stop certain compilers warning when using pointer
1423 differences as field lengths or precisions in printf-type calls (mostly
1424 affecting debugging statements).
1426 13. Added optional readline() support for -be (dynamically loaded).
1428 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1429 same clock tick as a message's arrival, so that its received time was the
1430 same as the "first fail" time on the retry record, and that message
1431 remained on the queue past the ultimate address timeout, every queue runner
1432 would try a delivery (because it was past the ultimate address timeout) but
1433 after another failure, the ultimate address timeout, which should have then
1434 bounced the address, did not kick in. This was a "< instead of <=" error;
1435 in most cases the first failure would have been in the next clock tick
1436 after the received time, and all would be well.
1438 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1439 being recognized when the domain list was tested by the match_domain
1440 condition in an expansion string.
1442 16. Added the ${str2b64: operator.
1444 17. Exim was always calling setrlimit() to set a large limit for the number of
1445 processes, without checking whether the existing limit was already
1446 adequate. (It did check for the limit on file descriptors.) Furthermore,
1447 errors from getrlimit() and setrlimit() were being ignored. Now they are
1448 logged to the main and panic logs, but Exim does carry on, to try to do its
1449 job under whatever limits there are.
1451 18. Imported PCRE 5.0.
1453 19. Trivial typo in log message " temporarily refused connection" (the leading
1456 20. If the log selector return_path_on_delivery was set and an address was
1457 redirected to /dev/null, the delivery process crashed because it assumed
1458 that a return path would always be set for a "successful" delivery. In this
1459 case, the whole delivery is bypassed as an optimization, and therefore no
1462 21. Internal re-arrangement: the function for sending a challenge and reading
1463 a response while authentication was assuming a zero-terminated challenge
1464 string. It's now changed to take a pointer and a length, to allow for
1465 binary data in such strings.
1467 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1469 23. Exim was not respecting finduser_retries when seeking the login of the
1470 uid under which it was called; it was always trying 10 times. (The default
1471 setting of finduser_retries is zero.) Also, it was sleeping after the final
1472 failure, which is pointless.
1474 24. Implemented tls_on_connect_ports.
1476 25. Implemented acl_smtp_predata.
1478 26. If the domain in control=submission is set empty, Exim assumes that the
1479 authenticated id is a complete email address when it generates From: or
1480 Sender: header lines.
1482 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1483 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1484 chown and chgrp in /bin and hostname in /usr/bin.
1486 28. Exim was keeping the "process log" file open after each use, just as it
1487 does for the main log. This opens the possibility of it remaining open for
1488 long periods when the USR1 signal hits a daemon. Occasional processlog
1489 errors were reported, that could have been caused by this. Anyway, it seems
1490 much more sensible not to leave this file open at all, so that is what now
1493 29. The long-running daemon process does not normally write to the log once it
1494 has entered its main loop, and it closes the log before doing so. This is
1495 so that log files can straightforwardly be renamed and moved. However,
1496 there are a couple of unusual error situations where the daemon does write
1497 log entries, and I had neglected to close the log afterwards.
1499 30. The text of an SMTP error response that was received during a remote
1500 delivery was being truncated at 512 bytes. This is too short for some of
1501 the long messages that one sometimes sees. I've increased the limit to
1504 31. It is now possible to make retry rules that apply only when a message has a
1505 specific sender, in particular, an empty sender.
1507 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1508 it possible to be selective about when SMTP synchronization is enforced.
1510 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1512 32. Implemented hosts_connection_nolog.
1514 33. Added an ACL for QUIT.
1516 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1519 35. Added mailbox_size and mailbox_filecount to appendfile.
1521 36. Added control = no_multiline_responses to ACLs.
1523 37. There was a bug in the logic of the code that waits for the clock to tick
1524 in the case where the clock went backwards by a substantial amount such
1525 that the microsecond fraction of "now" was more than the microsecond
1526 fraction of "then" (but the whole seconds number was less).
1528 38. Added support for the libradius Radius client library this is found on
1529 FreeBSD (previously only the radiusclient library was supported).
1535 1. When certain lookups returned multiple values in the form name=value, the
1536 quoting of the values was not always being done properly. Specifically:
1537 (a) If the value started with a double quote, but contained no whitespace,
1539 (b) If the value contained whitespace other than a space character (i.e.
1540 tabs or newlines or carriage returns) it was not quoted.
1541 This fix has been applied to the mysql and pgsql lookups by writing a
1542 separate quoting function and calling it from the lookup code. The fix
1543 should probably also be applied to nisplus, ibase and oracle lookups, but
1544 since I cannot test any of those, I have not disturbed their existing code.
1546 2. A hit in the callout cache for a specific address caused a log line with no
1547 reason for rejecting RCPT. Now it says "Previous (cached) callout
1548 verification failure".
1550 3. There was an off-by-one bug in the queryprogram router. An over-long
1551 return line was truncated at 256 instead of 255 characters, thereby
1552 overflowing its buffer with the terminating zero. As well as fixing this, I
1553 have increased the buffer size to 1024 (and made a note to document this).
1555 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1556 when Exim is waiting for an SMTP response from a remote server, Exim
1557 restarts its select() call on the socket, thereby resetting its timeout.
1558 This is not a problem when such interrupts are rare. Somebody set up a cron
1559 job to run exiwhat every 2 minutes, which is less than the normal select()
1560 timeout (5 or 10 minutes). This meant that the select() timeout never
1561 kicked in because it was always reset. I have fixed this by comparing the
1562 time when an interrupt arrives with the time at the start of the first call
1563 to select(). If more time than the timeout has elapsed, the interrupt is
1564 treated as a timeout.
1566 5. Some internal re-factoring in preparation for the addition of Sieve
1567 extensions (by MH). In particular, the "personal" test is moved to a
1568 separate function, and given an option for scanning Cc: and Bcc: (which is
1569 not set for Exim filters).
1571 6. When Exim created an email address using the login of the caller as the
1572 local part (e.g. when creating a From: or Sender: header line), it was not
1573 quoting the local part when it contained special characters such as @.
1575 7. Installed new OpenBSD configuration files.
1577 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1578 try to make them clearer.
1580 9. Callout options, other than the timeout value, were being ignored when
1581 verifying sender addresses in header lines. For example, when using
1583 verify = header_sender/callout=no_cache
1585 the cache was (incorrectly) being used.
1587 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1588 only the Cygwin environment.
1590 11. When return_path_on_delivery was set as a log selector, if different remote
1591 addresses in the same message used different return paths and parallel
1592 remote delivery occurred, the wrong values would sometimes be logged.
1593 (Whenever a remote delivery process finished, the return path value from
1594 the most recently started remote delivery process was logged.)
1596 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1597 lines when AUTH and/or TLS are in use. This is the "received protocol"
1598 field. Exim used to use "asmtp" for authenticated SMTP, without any
1599 indication (in the protocol name) for TLS use. Now it follows the RFC and
1600 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1601 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1602 names appear in log lines as well as in Received: header lines.
1604 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1605 extensions, and comparison tests, and to fix some bugs.
1607 14. Changes to the "personal" filter test:
1609 (1) The test was buggy in that it was just doing the equivalent of
1610 "contains" tests on header lines. For example, if a user's address was
1611 anne@some.where, the "personal" test would incorrectly be true for
1613 To: susanne@some.where
1615 This test is now done by extracting each address from the header in turn,
1616 and checking the entire address. Other tests that are part of "personal"
1617 are now done using regular expressions (for example, to check local parts
1618 of addresses in From: header lines).
1620 (2) The list of non-personal local parts in From: addresses has been
1621 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1622 taken from the Sieve specification recommendations.
1624 (3) If the message contains any header line starting with "List-" it is
1625 treated as non-personal.
1627 (4) The test for "circular" in the Subject: header line has been removed
1628 because it now seems ill-conceived.
1630 15. Minor typos in src/EDITME comments corrected.
1632 16. Installed latest exipick from John Jetmore.
1634 17. If headers_add on a router specified a text string that was too long for
1635 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1636 of string_sprintf() is now avoided.
1638 18. $message_body_size was not set (it was always zero) when running the DATA
1639 ACL and the local_scan() function.
1641 19. For the "mail" command in an Exim filter, no default was being set for
1642 the once_repeat time, causing a random time value to be used if "once" was
1643 specified. (If the value happened to be <= 0, no repeat happened.) The
1644 default is now 0s, meaning "never repeat". The "vacation" command was OK
1645 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1646 (I found it when inspecting the code).
1648 20. There is now an overall timeout for performing a callout verification. It
1649 defaults to 4 times the callout timeout, which applies to individual SMTP
1650 commands during the callout. The overall timeout applies when there is more
1651 than one host that can be tried. The timeout is checked before trying the
1652 next host. This prevents very long delays if there are a large number of
1653 hosts and all are timing out (e.g. when the network connections are timing
1654 out). The value of the overall timeout can be changed by specifying an
1655 additional sub-option for "callout", called "maxwait". For example:
1657 verify = sender/callout=5s,maxwait=20s
1659 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1660 to the end before writing, but this should make it even safer).
1662 22. Exim was forgetting that it had advertised PIPELINING for the second and
1663 subsequent messages on an SMTP connection. It was also not resetting its
1664 memory on STARTTLS and an internal HELO.
1666 23. When Exim logs an SMTP synchronization error within a session, it now
1667 records whether PIPELINING has been advertised or not.
1669 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1670 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1671 rather than long int.
1673 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1675 26. Added the never_mail option to autoreply.
1681 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1682 crash if the getsockname() call failed; this can happen if a connection is
1683 closed very soon after it is established. The problem was simply in the
1684 order in which certain operations were done, causing Exim to try to write
1685 to the SMTP stream before it had set up the file descriptor. The bug has
1686 been fixed by making things happen in the correct order.
1692 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1693 before the connection was closed, thus losing the rejection response.
1695 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1696 some early Solaris releases, but causes trouble in current releases where
1697 socklen_t is defined.
1699 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1702 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1705 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1706 that releasing the top of it at the end releases what was used for sub-
1707 expansions (unless the block got too big). However, discard this block if
1708 the first thing is a variable or header, so that we can use its block when
1709 it is dynamic (useful for very large $message_headers, for example).
1711 6. Lookups now cache *every* query, not just the most recent. A new, separate
1712 store pool is used for this. It can be recovered when all lookup caches are
1713 flushed. Lookups now release memory at the end of their result strings.
1714 This has involved some general refactoring of the lookup sources.
1716 7. Some code has been added to the store_xxx() functions to reduce the amount
1717 of flapping under certain conditions.
1719 8. log_incoming_interface used to affect only the <= reception log lines. Now
1720 it causes the local interface and port to be added to several more SMTP log
1721 lines, for example "SMTP connection from", and rejection lines.
1723 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1725 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1727 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1728 could be overwritten at the end of the current message (or the start of a
1729 new message if it was set in a HELO ACL). The value is now preserved for
1730 the duration of the SMTP connection.
1732 12. If a transport had a headers_rewrite setting, and a matching header line
1733 contained an unqualified address, that address was qualified, even if it
1734 did not match any rewriting rules. The underlying bug was that the values
1735 of the flags that permit the existence of unqualified sender and recipient
1736 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1737 non-local messages, and by -bnq for local messages) were not being
1738 preserved with the message after it was received.
1740 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1741 "next input=" as part of the text comprising the host identity instead of
1742 the correct text. The code was using the same buffer for two different
1743 strings. However, depending on which order the printing function evaluated
1744 its arguments, the bug did not always show up. Under Linux, for example, my
1745 test suite worked just fine.
1747 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1748 doesn't work with some older versions of Perl. It has been changed to "my",
1749 which in any case is probably the better facility to use.
1751 15. A really picky compiler found some instances of statements for creating
1752 error messages that either had too many or two few arguments for the format
1755 16. The size of the buffer for calls to the DNS resolver has been increased
1756 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1757 for addresses that have a lot of PTR records. This alleviates a problem; it
1758 does not fully solve it.
1760 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1761 buffer now truncates the list and logs the incident, which is the same
1762 action as happens when Exim is looking up a host name and its aliases.
1763 Previously in this situation something unpredictable would happen;
1764 sometimes it was "internal error: store_reset failed".
1766 18. If a server dropped the connection unexpectedly when an Exim client was
1767 using GnuTLS and trying to read a response, the client delivery process
1768 crashed while trying to generate an error log message.
1770 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1771 string, for example:
1773 warn message = H1: something\nH2: something
1775 the text was added as a single header line from Exim's point of view
1776 though it ended up OK in the delivered message. However, searching for the
1777 second and subsequent header lines using $h_h2: did not work. This has been
1778 fixed. Similarly, if a system filter added multiple headers in this way,
1779 the routers could not see them.
1781 20. Expanded the error message when iplsearch is called with an invalid key to
1782 suggest using net-iplsearch in a host list.
1784 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1785 ACLs are no longer actually imposed (and a message to that effect is
1788 22. If a "gecos" field in a passwd entry contained escaped characters, in
1789 particular, if it contained a \" sequence, Exim got it wrong when building
1790 a From: or a Sender: header from that name. A second bug also caused
1791 incorrect handling when an unquoted " was present following a character
1792 that needed quoting.
1794 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1795 was not being matched caselessly.
1797 24. Arranged for all hyphens in the exim.8 source to be escaped with
1800 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1801 properly. Recipient callout cache records were still being keyed to include
1802 the sender, even when use_sender was set false. This led to far more
1803 callouts that were necessary. The sender is no longer included in the key
1804 when use_sender is false.
1806 26. Added "control = submission" modifier to ACLs.
1808 27. Added the ${base62d: operator to decode base 62 numbers.
1810 28. dnsdb lookups can now access SRV records.
1812 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1813 the configuration file.
1815 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1816 (-v) mode. This makes the output for a verbose queue run more intelligible.
1818 31. Added a use_postmaster feature to recipient callouts.
1820 32. Added the $body_zerocount variable, containing the number of binary zero
1821 bytes in the message body.
1823 33. The time of last modification of the "new" subdirectory is now used as the
1824 "mailbox time last read" when there is a quota error for a maildir
1827 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1829 35. Added +ignore_unknown as a special item in host lists.
1831 36. Code for decoding IPv6 addresses in host lists is now included, even if
1832 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1833 address was recognized as an IP address, but was then not correctly decoded
1834 into binary, causing unexpected and incorrect effects when compared with
1841 1. Very minor rewording of debugging text in manualroute to say "list of
1842 hosts" instead of "hostlist".
1844 2. If verify=header_syntax was set, and a header line with an unqualified
1845 address (no domain) and a large number of spaces between the end of the
1846 name and the colon was received, the reception process suffered a buffer
1847 overflow, and (when I tested it) crashed. This was caused by some obsolete
1848 code that should have been removed. The fix is to remove it!
1850 3. When running in the test harness, delay a bit after writing a bounce
1851 message to get a bit more predictability in the log output.
1853 4. Added a call to search_tidyup() just before forking a reception process. In
1854 theory, someone could use a lookup in the expansion of smtp_accept_max_
1855 per_host which, without the tidyup, could leave open a database connection.
1857 5. Added the variables $recipient_data and $sender_data which get set from a
1858 lookup success in an ACL "recipients" or "senders" condition, or a router
1859 "senders" option, similar to $domain_data and $local_part_data.
1861 6. Moved the writing of debug_print from before to after the "senders" test
1864 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1865 problems for message scanning, either using a data ACL, or using
1866 local_scan() because the Received: header was not generated till after they
1867 were called (in order to set the time as the time of reception completion).
1868 I have revised the way this works. The header is now generated after the
1869 body is received, but before the ACL or local_scan() are called. After they
1870 are run, the timestamp in the header is updated.
1876 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1877 before starting a queue runner without re-exec. This happened only when
1878 deliver_drop_privilege was set or when the Exim user was set to root. The
1879 effect of the bug was that timeouts during subsequent deliveries caused
1880 crashes instead of being properly handled. The handler is now left at its
1881 default (and expected) setting.
1883 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1884 message, again when deliver_drop_privilege is set or Exim is run as root.
1885 The bug described in (1) was not present in this case, but the tidying up
1886 of the other signals was missing. I have made the two cases consistent.
1888 3. The ignore_target_hosts setting on a manualroute router was being ignored
1889 for hosts that were looked up using the /MX notation.
1891 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1894 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1895 operated on the sender address. After changing the $sender_address to <>
1896 for the sender address verify, Exim was re-instated it as the original
1897 (before rewriting) address, but remembering that it had rewritten it, so it
1898 wasn't rewriting it again. This bug also had the effect of breaking the
1899 sender address verification caching when the sender address was rewritten.
1901 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1902 This has been changed so that if the ip literal address matches
1903 ignore_target_hosts, the router declines.
1905 7. Added expansion conditions match_domain, match_address, and match_local_
1906 part (NOT match_host).
1908 8. The placeholder for the Received: header didn't have a length field set.
1910 9. Added code to Exim itself and to exim_lock to test for a specific race
1911 condition that could lead to file corruption when using MBX delivery. The
1912 issue is with the lockfile that is created in /tmp. If this file is removed
1913 after a process has opened it but before that process has acquired a lock,
1914 there is the potential for a second process to recreate the file and also
1915 acquire a lock. This could lead to two Exim processes writing to the file
1916 at the same time. The added code performs the same test as UW imapd; it
1917 checks after acquiring the lock that its file descriptor still refers to
1918 the same named file.
1920 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1921 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1924 11. Added the smtp_active_hostname option. If used, this will typically be made
1925 to depend on the incoming interface address. Because $interface_address is
1926 not set up until the daemon has forked a reception process, error responses
1927 that can happen earlier (such as "too many connections") no longer contain
1930 12. If an expansion in a condition on a "warn" statement fails because a lookup
1931 defers, the "warn" statement is abandoned, and the next ACL statement is
1932 processed. Previously this caused the whole ACL to be aborted.
1934 13. Added the iplsearch lookup type.
1936 14. Added ident_timeout as a log selector.
1938 15. Added tls_certificate_verified as a log selector.
1940 16. Added a global option tls_require_ciphers (compare the smtp transport
1941 option of the same name). This controls incoming TLS connections.
1943 17. I finally figured out how to make tls_require_ciphers do a similar thing
1944 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1945 before starting the TLS session.
1947 18. Tabs are now shown as \t in -bP output.
1949 19. If the log selector return_path_on_delivery was set, Exim crashed when
1950 bouncing a message because it had too many Received: header lines.
1952 20. If two routers both had headers_remove settings, and the first one included
1953 a superfluous trailing colon, the final name in the first list and the
1954 first name in the second list were incorrectly joined into one item (with a
1955 colon in the middle).
1961 1. Added -C and -D options to the exinext utility, mainly to make it easier
1962 to include in the automated testing, but these could be helpful when
1963 multiple configurations are in use.
1965 2. The exinext utility was not formatting the output nicely when there was
1966 an alternate port involved in the retry record key, nor when there was a
1967 message id as well (for retries that were specific to a specific message
1968 and a specific host). It was also confused by IPv6 addresses, because of
1969 the additional colons they contain. I have fixed the IPv4 problem, and
1970 patched it up to do a reasonable job for IPv6.
1972 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1973 delivery, the log line now contains "pipelined" if PIPELINING was used.
1975 4. An SMTP transport process used to panic and die if the bind() call to set
1976 an explicit outgoing interface failed. This has been changed; it is now
1977 treated in the same way as a connect() failure.
1979 5. A reference to $sender_host_name in the part of a conditional expansion
1980 that was being skipped was still causing a DNS lookup. This no longer
1983 6. The def: expansion condition was not recognizing references to header lines
1984 that used bh_ and bheader_.
1986 7. Added the _cache feature to named lists.
1988 8. The code for checking quota_filecount in the appendfile transport was
1989 allowing one more file than it should have been.
1991 9. For compatibility with Sendmail, the command line option
1999 and sets the incoming protocol and host name (for trusted callers). The
2000 host name and its colon can be omitted when only the protocol is to be set.
2001 Note the Exim already has two private options, -pd and -ps, that refer to
2002 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2003 "s", but I don't think that's a major issue.
2005 10. A number of refactoring changes to the code, none of which should affect
2008 (a) The number of logging options was getting close to filling up the
2009 32-bit word that was used as a bit map. I have split them into two classes:
2010 those that are passed in the argument to log_write(), and those that are
2011 only ever tested independently outside of that function. These are now in
2012 separate 32-bit words, so there is plenty of room for expansion again.
2013 There is no change in the user interface or the logging behaviour.
2015 (b) When building, for example, log lines, the code previously used a
2016 macro that called string_cat() twice, in order to add two strings. This is
2017 not really sufficiently general. Furthermore, there was one instance where
2018 it was actually wrong because one of the argument was used twice, and in
2019 one call a function was used. (As it happened, calling the function twice
2020 did not affect the overall behaviour.) The macro has been replaced by a
2021 function that can join an arbitrary number of extra strings onto a growing
2024 (c) The code for expansion conditions now uses a table and a binary chop
2025 instead of a serial search (which was left over from when there were very
2026 few conditions). Also, it now recognizes conditions like "pam" even when
2027 the relevant support is not compiled in: a suitably worded error message is
2028 given if an attempt is made to use such a condition.
2030 11. Added ${time_interval:xxxxx}.
2032 12. A bug was causing one of the ddress fields not to be passed back correctly
2033 from remote delivery subprocesses. The field in question was not being
2034 subsequently used, so this caused to problems in practice.
2036 13. Added new log selectors queue_time and deliver_time.
2038 14. Might have fixed a bug in maildirsizefile handling that threw up
2039 "unexpected character" debug warnings, and recalculated the data
2040 unnecessarily. In any case, I expanded the warning message to give more
2043 15. Added the message "Restricted characters in address" to the statements in
2044 the default ACL that block characters like @ and % in local parts.
2046 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2047 Three changes have been made:
2049 (a) There was a serious bug; a negative response to MAIL caused the whole
2050 recipient domain to be cached as invalid, thereby blocking all messages
2051 to all local parts at the same domain, from all senders. This bug has
2052 been fixed. The domain is no longer cached after a negative response to
2053 MAIL if the sender used is not empty.
2055 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2058 (c) A new callout option, "use_sender" has been added for people who want
2059 the modified behaviour.
2065 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2068 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2069 indeed breaks things for older releases.
2071 3. Added additional logging to the case where there is a problem reading data
2072 from a filter that is running in a subprocess using a pipe, in order to
2073 try to track down a specific problem.
2075 4. Testing facility fudge: when running in the test harness and attempting
2076 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2077 getting "No route to host". Convert this to a timeout.
2079 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2082 6. Some OS don't have socklen_t but use size_t instead. This affects the
2083 fifth argument of getsockopt() amongst other things. This is now
2084 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2085 can be set for individual OS. I have set it for SunOS5, OSF1, and
2086 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2087 some earlier ones do not.
2089 7. Change 4.30/15 was not doing the test caselessly.
2091 8. The standard form for an IPv6 address literal was being rejected by address
2092 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2093 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2094 this, as well as the form without the "IPv6" on the front (but only when
2095 address literals are enabled, of course).
2097 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2099 10. Exim crashed if a message with an empty sender address specified by -f
2100 encountered a router with an errors_to setting. This could be provoked only
2101 by a command such as
2105 where an empty string was supplied; "<>" did not hit this bug.
2107 11. Installed PCRE release 4.5.
2109 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2110 remained set. It is now erased.
2112 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2113 times from message ids (which are base 36 rather than the normal 62).
2115 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2116 were being counted as actual protocol errors, and logged if the log
2117 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2118 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2119 and DATA following a set of rejected RCPTs do not count as protocol errors.
2120 In other words, Exim assumes they were pipelined, though this may not
2121 actually be the case. Of course, in all cases the client gets an
2122 appropriate error code.
2124 15. If a lookup fails in an ACL condition, a message about the failure may
2125 be available; it is used if testing the ACL cannot continue, because most
2126 such messages specify what the cause of the deferral is. However, some
2127 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2128 that caused an old message to be retained and used if a later statement
2129 caused a defer, replacing the real cause of the deferral.
2131 16. If an IP address had so many PTR records that the DNS lookup buffer
2132 was not large enough to hold them, Exim could crash while trying to process
2133 the truncated data. It now detects and logs this case.
2135 17. Further to 4.21/58, another change has been made: if (and only if) the
2136 first line of a message (the first header line) ends with CRLF, a bare LF
2137 in a subsequent header line has a space inserted after it, so as not to
2138 terminate the header.
2140 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2141 unnecessarily, used atoi() instead of strtol(), and didn't check the
2142 termination when getting file sizes from file names by regex.
2144 19. Completely re-implemented the support for maildirsize files, in the light
2145 of a number of problems with the previous contributed implementation
2146 (4.30/29). In particular:
2148 . If the quota is zero, the maildirsize file is maintained, but no quota is
2151 . If the maildir directory does not exist, it is created before any attempt
2152 to write a maildirsize file.
2154 . The quota value in the file is just a cache; if the quota is changed in
2155 the transport, the new value overrides.
2157 . A regular expression is available for excluding directories from the
2160 20. The autoreply transport checks the characters in options that define the
2161 message's headers; it allows continued headers, but it was checking with
2162 isspace() after an embedded newline instead of explicitly looking for a
2165 21. If all the "regular" hosts to which an address was routed had passed their
2166 expiry times, and had not reached their retry times, the address was
2167 bounced, even if fallback hosts were defined. Now Exim should go on to try
2170 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2171 equivalent code in the SMTP transport. Some hosts send humungous responses
2172 to HELO/EHLO, more than 1024 it seems.
2174 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2175 gives compiler warnings in some environments. I've now done it "properly",
2178 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2179 (because of problems with the built-in one) was declared to return uschar *
2180 instead of char *, causing compiler failure.
2182 25. Fixed a file descriptor leak when processing alias/forward files.
2184 26. Fixed a minor format string issue in dbfn.c.
2186 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2188 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2189 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2192 29. When a sender address is verified, it is cached, to save repeating the test
2193 when there is more than one recipient in a message. However, when the
2194 verification involves a callout, it is possible for different callout
2195 options to be set for different recipients. It is too complicated to keep
2196 track of this in the cache, so now Exim always runs a verification when a
2197 callout is required, relying on the callout cache for the optimization.
2198 The overhead is duplication of the address routing, but this should not be
2201 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2202 to be verified with callout=postmaster, and the main callout worked but the
2203 postmaster check failed, the verification correctly failed. However, if a
2204 subsequent RCPT command asked for sender verification *without* the
2205 postmaster check, incorrect caching caused this verification also to fail,
2208 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2209 it was not caching the DNS options (qualify_single, search_parents) that
2210 were used when the lookup failed. A subsequent lookup with different
2211 options therefore always gave the same answer, though there were cases
2212 where it should not have. (Example: a "domains = !$mx_any" option on a
2213 dnslookup router: the "domains" option is always processed without any
2214 widening, but the router might have qualify_single set.) Now Exim uses the
2215 cached value only when the same options are set.
2217 32. Added John Jetmore's "exipick" utility to the distribution.
2219 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2220 than a timeout (e.g. a certificate is required, and is not provided), an
2221 Exim server now closes the connection immediately. Previously it waited for
2222 the client to close - but if the client is SSL, it seems that they each
2223 wait for each other, leading to a delay before one of them times out.
2225 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2226 maintained 0.8.x compatibility because I don't think many are using it, and
2227 it is clearly obsolete.
2229 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2232 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2233 client certificate was expired. A simple patch fixes this, though I don't
2234 understand the full logic of why the verify callback is called multiple
2237 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2238 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2239 which causes problems with some clients (such as the Certicom SSL Plus
2240 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2241 disables the coutermeasure allowing Eudora to connect."
2243 38. Exim was not checking that a write() to a log file succeeded. This could
2244 lead to Bad Things if a log got too big, in particular if it hit a file
2245 size limit. Exim now panics and dies if it cannot write to a log file, just
2246 as it does if it cannot open a log file.
2248 39. Modified OS/Makefile-Linux so that it now contains
2250 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2252 The two -D definitions ensure that Exim is compiled with large file
2253 support, which makes it possible to handle log files that are bigger than
2256 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2257 instance) a domain was checked against a named list that involved a lookup,
2258 causing $domain_data to be set, then another domain was checked against the
2259 same list, then the first domain was re-checked, the value of $domain_data
2260 after the final check could be wrong. In particular, if the second check
2261 failed, it could be set empty. This bug probably also applied to
2264 41. The strip_trailing_dot option was not being applied to the address given
2265 with the -f command-line option.
2267 42. The code for reading a message's header from the spool was incrementing
2268 $received_count, but never initializing it. This meant that the value was
2269 incorrect (doubled) while delivering a message in the same process in which
2270 it was received. In the most common configuration of Exim, this never
2271 happens - a fresh exec is done - but it can happen when
2272 deliver_drop_privilege is set.
2274 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2275 it now includes up to 150 characters of the unexpected data in the log
2278 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2279 and building data strings. The size of both of these buffers was 10 000
2280 bytes - far larger than anybody would *ever* want, thought I. Needless to
2281 say, somebody hit the limit. I have increased the maximum line length to
2282 20 000 and the maximum data length of concatenated lines to 100 000. I have
2283 also fixed two bugs, because there was no checking on these buffers. Tsk,
2284 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2285 buffer is too small.
2287 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2288 lsearch lookups. Now it does.
2290 46. When parsing a route_list item in a manualroute router, a fixed-length
2291 buffer was used for the list of hosts. I made this 1024 bytes long,
2292 thinking that nobody would ever have a list of hosts that long. Wrong.
2293 Somebody had a whole pile of complicated expansion conditions, and the
2294 string was silently truncated, leading to an expansion error. It turns out
2295 that it is easier to change to an unlimited length (owing to other changes
2296 that have happened since this code was originally written) than to build
2297 structure for giving a limitation error. The length of the item that
2298 expands into the list of hosts is now unlimited.
2300 47. The lsearch lookup could not handle data where the length of text line was
2301 more than 4095 characters. Such lines were truncated, leading to shortened
2302 data being returned. It should now handle lines of any length.
2304 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2305 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2308 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2309 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2310 no longer happens in comment lines. A list of replacements is now placed
2311 at the head of all of the source files, except those whose only change is
2312 to replace PERL_COMMAND in the very first #! line.
2314 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2315 messages on the queue, with a bottom-up merge sort, using code contributed
2316 by Michael Haardt. This should make operations like -bp somewhat faster on
2317 large queues. It won't affect queue runners, except when queue_run_in_order
2320 51. Installed eximstats 1.31 in the distribution.
2322 52. Added support for SRV lookups to the dnslookup router.
2324 53. If an ACL referred to $message_body or $message_body_end, the value was not
2325 reset for any messages that followed in the same SMTP session.
2327 54. The store-handling optimization for building very long strings was not
2328 differentiating between the different store pools. I don't think this
2329 actually made any difference in practice, but I've tidied it.
2331 55. While running the routers to verify a sender address, $sender_address
2332 was still set to the sender address. This is wrong, because when routing to
2333 send a bounce to the sender, it would be empty. Therefore, I have changed
2334 it so that, while verifying a sender address, $sender_address is set to <>.
2335 (There is no change to what happens when verifying a recipient address.)
2337 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2338 target A or AAAA records (if not already returned) without resetting the
2339 qualify_single or search_parents options of the DNS resolver. These are
2340 inappropriate in this case because the targets of MX and SRV records must
2341 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2342 target that, when qualified, matched something in the local domain. These
2343 two options are now turned off when doing these lookups.
2345 57. It seems that at least some releases of Reiserfs (which does not have the
2346 concept of a fixed number of inodes) returns zero and not -1 for the
2347 number of available inodes. This interacted badly with check_spool_inodes,
2348 which assumed that -1 was the "no such thing" setting. What I have done is
2349 to check that the total number of inodes is greater than zero before doing
2350 the test of how many are available.
2352 58. When a "warn" ACL statement has a log_message modifier, the message is
2353 remembered, and not repeated. This is to avoid a lot of repetition when a
2354 message has many recipients that cause the same warning to be written.
2355 Howewer, Exim was preserving the list of already written lines for an
2356 entire SMTP session, which doesn't seem right. The memory is now reset if a
2357 new message is started.
2359 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2360 debugging output unless log_rewrite was also set.
2362 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2363 of (int)(handle) when we know that handle contains (void *)(-1).
2365 61. The Exim daemon panic-logs an error return when it closes the incoming
2366 connection. However "connection reset by peer" seems to be common, and
2367 isn't really an error worthy of noting specially, so that particular error
2370 62. When Exim is trying to find all the local interfaces, it used to panic and
2371 die if the ioctl to get the interface flags failed. However, it seems that
2372 on at least one OS (Solaris 9) it is possible to have an interface that is
2373 included in the list of interfaces, but for which you get a failure error
2374 for this call. This happens when the interface is not "plumbed" into a
2375 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2376 failure of the "get flags" call assumes that the interface is down.
2378 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2379 makes life easier for people who are doing arithmetic on fields extracted
2380 from dates, where you often get leading zeros that should not be
2381 interpreted as octal.
2383 64. Added qualify_domain to the redirect router, to override the global
2386 65. If a pathologically long header line contained very many addresses (the
2387 report of this problem mentioned 10 000) and each of them was rewritten,
2388 Exim could use up a very large amount of memory. (It kept on making new
2389 copies of the header line as it rewrote, and never released the old ones.)
2390 At the expense of a bit more processing, the header rewriting function has
2391 been changed so that it no longer eats memory in this way.
2393 66. The generation of the Received: header has been moved from the time that a
2394 message starts to be received, to the time that it finishes. The timestamp
2395 in the Received: header should now be very close to that of the <= log
2396 line. There are two side-effects of this change:
2398 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2399 logged header lines no longer include the local Received: line, because
2400 it has not yet been created. The same applies to a copy of the message
2401 that is returned to a non-SMTP sender when a message is rejected.
2403 (b) When a filter file is tested using -bf, no additional Received: header
2404 is added to the test message. After some thought, I decided that this
2407 This change does not affect the value of $received_for. It is still set
2408 after address rewriting, but before local_scan() is called.
2410 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2412 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2413 gave an unhelpful panic error message, and a defer error. I have managed to
2414 change this behaviour so that it now rejects any supplied certificate,
2415 which seems right, as the list of acceptable certificates is empty.
2417 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2418 gave an unhelpful defer error. I have not managed to make this reject any
2419 supplied certificates, but the error message it gives is "no certificate
2420 supplied", which is not helpful.
2422 70. exigrep's output now also includes lines that are not associated with any
2423 message, but which match the given pattern. Implemented by a patch from
2424 Martin Sluka, which also tidied up the Perl a bit.
2426 71. Recipient callout verification, like sender verification, was using <> in
2427 the MAIL FROM command. This isn't really the right thing, since the actual
2428 sender may affect whether the remote host accepts the recipient or not. I
2429 have changed it to use the actual sender in the callout; this means that
2430 the cache record is now keyed on a recipient/sender pair, not just the
2431 recipient address. There doesn't seem to be a real danger of callout loops,
2432 since a callout by the remote host to check the sender would use <>.
2433 [SEE ABOVE: changed after hitting problems.]
2435 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2436 temporary errors. However, in the case of such a code being given after
2437 the end of a data transmission (i.e. after ".") Exim was failing to write
2438 a retry record for the message. (Yes, there was some broken host that was
2439 actually sending 8xx at this point.)
2441 73. An unknown lookup type in a host list could cause Exim to panic-die when
2442 the list was checked. (An example that provoked this was putting <; in the
2443 middle of a list instead of at the start.) If this happened during a DATA
2444 ACL check, a -D file could be left lying around. This kind of configuration
2445 error no longer causes Exim to die; instead it causes a defer errror. The
2446 incident is still logged to the main and panic logs.
2448 74. Buglet left over from Exim 3 conversion. The message "too many messages
2449 in one connection" was written to the rejectlog but not the mainlog, except
2450 when address rewriting (yes!) was being logged.
2452 75. Added write_rejectlog option.
2454 76. When a system filter was run not as root (that is, when system_filter_user
2455 was set), the values of the $n variables were not being returned to the
2456 main process; thus, they were not subsequently available in the $sn
2459 77. Added +return_path_on_delivery log selector.
2461 78. A connection timeout was being treated differently from recipients deferred
2462 when testing hosts_max_try with a message that was older than the host's
2463 retry timeout. (The host should not be counted, thus allowing all hosts to
2464 be tried at least once before bouncing.) This may have been the cause of an
2465 occasionally reported bug whereby a message would remain on the queue
2466 longer than the retry timeout, but would be bounced if a delivery was
2467 forced. I say "may" because I never totally pinned down the problem;
2468 setting up timeout/retry tests is difficult. See also the next item.
2470 79. The ultimate address timeout was not being applied to errors that involved
2471 a combination of host plus message (for example, a timeout on a MAIL
2472 command). When an address resolved to a number of possible hosts, and they
2473 were not all tried for each delivery (e.g. because of hosts_max_try), a
2474 message could remain on the queue longer than the retry timeout.
2476 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2479 81. Fixed an obscure SMTP outgoing bug which required at least the following
2480 conditions: (a) there was another message waiting for the same server;
2481 (b) the server returned 5xx to all RCPT commands in the first message so
2482 that the message was not completed; (c) the server dropped the connection
2483 or gave a negative response to the RSET that Exim sends to abort the
2484 transaction. The observed case was a dropped connection after DATA that had
2485 been sent in pipelining mode. That is, the server had advertised PIPELINING
2486 but was not implementing it correctly. The effect of the bug was incorrect
2487 behaviour, such as trying another host, and this could lead to a crash.
2493 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2494 and daemon.c were passed as pointers to ints; they should have been
2495 pointers to socklen_t variables (which are typically unsigned ints).
2497 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2500 3. Fixed a really odd bug that affected only the testing scheme; patching a
2501 certain fixed string in the binary changed the value of another string that
2502 happened to be identical to the end of the original first string.
2504 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2505 name", it returns that address as the IP address. On some operating
2506 systems (e.g. Solaris), it also passes back the IP address string as the
2507 "host name". However, on others (e.g. Linux), it passes back an empty
2508 string. Exim wasn't checking for this, and was changing the host name to an
2509 empty string, assuming it had been canonicized.
2511 5. Although rare, it is permitted to have more than one PTR record for a given
2512 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2513 all the names associated with an address, because they do in Solaris.
2514 However, it seems that they do not in Linux for data that comes from the
2515 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2516 I found this out when I moved to a new Linux workstation and tried to run
2517 the Exim test suite.
2519 To get round this problem I have changed the code so that it now does its
2520 own call to the DNS to look up PTR records when searching for a host name.
2521 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2522 addresses that are only in /etc/hosts are still found.
2524 This behaviour is, however, controlled by an option called host_lookup_
2525 order, which defaults to "bydns:byaddr". If people want to use the other
2526 order, or indeed, just use one or the other means of lookup, they can
2527 specify it in this variable.
2529 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2530 some operating systems, this comes back from gethostbyaddr() as an empty
2531 string, and this is what Exim used to test for. However, it seems that in
2532 other systems, "." is yielded. Exim now tests for this case too.
2534 7. The values of check_spool_space and check_log_space are now held internally
2535 as a number of kilobytes instead of an absolute number of bytes. If a
2536 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2537 kilobyte. This means that much larger values can be stored.
2539 8. Exim monitor: an attempt to get the action menu when not actually pointing
2540 at a message produces an empty menu entitled "No message selected". This
2541 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2542 no entries in it ("Shell widget menu has zero width and/or height"). So I
2543 have added a single, blank menu entry in this case.
2545 9. Added ${quote_local_part.
2547 10. MIME decoding is now applied to the contents of Subject: header lines when
2550 11. Now that a reference to $sender_host_address automatically causes a reverse
2551 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2552 host lookup before query-style lookups in lists that might use this
2553 variable. This has therefore been abolished, and the "net-" prefix is no
2554 longer necessary for query-style lookups.
2556 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2557 have been a typo for LFLAGS, so it has been changed.
2559 13. The install script calls Exim with "-C /dev/null" in order to find the
2560 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2561 to be output. Howeve, since Exim outputs its version number before the
2562 error, it didn't break the script. It just looked ugly. I fixed this by
2563 always allowing "-C /dev/null" if the caller is root.
2565 14. Ignore overlarge ACL variable number when reading spool file - insurance
2566 against a later release with more variables having written the file.
2568 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2569 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2570 this, as well as the form without the "IPv6" on the front.
2572 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2573 OS/Makefile-Darwin file.
2575 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2576 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2578 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2579 variable. However, some OS use ints for inodes. I've added cast to long int
2580 to get rid of the compiler warning.
2582 19. I had forgotten to lock out "/../" in configuration file names when
2583 ALT_CONFIG_PREFIX was set.
2585 20. Routers used for verification do not need to specify transports. However,
2586 if such a router generated a host list, and callout was configured, Exim
2587 crashed, because it could not find a port number from the (non-existent)
2588 transport. It now assumes port 25 in this circumstance.
2590 21. Added the -t option to exigrep.
2592 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2593 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2594 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2595 them is set, LOOKUP_LSEARCH is forced.
2597 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2599 24. Added sender and host information to the "rejected by local_scan()" log
2600 line; previously there was no indication of these.
2602 25. Added .include_if_exists.
2604 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2605 Linux. It turns out that not all file systems support this. Apparently some
2606 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2607 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2608 not supported on the file descriptor, is now ignored when Exim is trying to
2609 sync a directory. This applies only to Linux.
2611 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2613 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2614 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2617 29. Added support for maildirsize files from supplied patch (modified a bit).
2619 30. The use of :fail: followed by an empty string could lead Exim to respond to
2620 sender verification failures with (e.g.):
2622 550 Verification failed for <xxx>
2623 550 Sender verify failed
2625 where the first response line was missing the '-' that indicates it is not
2626 the final line of the response.
2628 31. The loop for finding the name of the user that called Exim had a hardwired
2629 limit of 10; it now uses the value of finduser_retries, which is used for
2630 all other user lookups.
2632 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2635 33. Exim was neglecting to zero errno before one call of strtol() when
2636 expanding a string and expecting an integer value. On some systems this
2637 resulted in spurious "integer overflow" errors. Also, it was casting the
2638 result into an int without checking.
2640 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2641 did not work. The code looks as if it has *never* worked, though it appears
2642 to have been documented since at least releast 1.62. I have made it work.
2644 35. The "timeout_DNS" error in retry rules, also documented since at least
2645 1.62, also never worked. As it isn't clear exactly what this means, and
2646 clearly it isn't a major issue, I have abolished the feature by treating it
2647 as "timeout", and writing a warning to the main and panic logs.
2649 36. The display of retry rules for -brt wasn't always showing the error code
2652 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2653 timeout_connect_A, timeout_connect_MX.
2655 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2656 to the empty sender.
2658 39. The daemon was not analysing the content of -oX till after it had closed
2659 stderr and disconnected from the controlling terminal. This meant that any
2660 syntax errors were only noted on the panic log, and the return code from
2661 the command was 0. By re-arranging the code a little, I've made the
2662 decoding happen first, so such errors now appear on stderr, and the return
2663 code is 1. However, the actual setting up of the sockets still happens in
2664 the disconnected process, so errors there are still only recorded on the
2667 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2668 connections (as happens on some IP stacks) was logged at start up time as
2669 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2670 it from "IPv6 and IPv4", which means that two separate sockets are being
2673 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2674 says whether AF_INET or AF_INET6 was passed as an argument.
2676 42. Exiwhat output was messed up when time zones were included in log
2679 43. Exiwhat now gives more information about the daemon's listening ports,
2680 and whether -tls-on-connect was used.
2682 44. The "port" option of the smtp transport is now expanded.
2684 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2685 silently ignored. Now an error message is written to the main and panic
2688 46. There's a new ACL modifier called "logwrite" which writes to a log file
2689 as soon as it is encountered.
2691 47. Added $local_user_uid and $local_user_gid at routing time.
2693 48. Exim crashed when trying to verify a sender address that was being
2696 49. Exim was recognizing only a space character after ".include". It now also
2697 recognizes a tab character.
2699 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2700 extracting the relevant information from the specification. The man page no
2701 longer contains scrambled data for the -d option, and I've added a section
2702 at the front about calling Exim under different names.
2704 51. Added "extra_headers" argument to the "mail" command in filter files.
2706 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2709 53. Installed eximstats 1.29.
2711 54. Added transport_filter_timeout as a generic transport option.
2713 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2714 Cc: header lines. This was required by RFC 822, but it not required by RFC
2717 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2718 incoming messages that did not have them. Now it does so only if the
2719 message originates locally, that is, if there is no associated remote host
2720 address. When Resent- header lines are present, this applies to the Resent-
2721 lines rather than the non-Resent- lines.
2723 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2724 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2726 58. Messages for configuration errors now include the name of the main
2727 configuration file - useful now that there may be more than one file in a
2728 list (.included file names were always shown).
2730 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2731 for those rare installations that do not start the daemon as root or run it
2732 setuid root. I've cut out the call to initgroups() if the daemon is not
2735 60. The Exim user and group can now be bound into the binary as text strings
2736 that are looked up at the start of Exim's processing.
2738 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2740 62. Added $mailstore_basename variable.
2742 63. Installed patch to sieve.c from Michael Haardt.
2744 64. When Exim failed to open the panic log after failing to open the main log,
2745 the original message it was trying to log was written to stderr and debug
2746 output, but if they were not available (the usual case in production), it
2747 was lost. Now it is written to syslog before the two lines that record the
2748 failures to open the logs.
2750 65. Users' Exim filters run in subprocesses under the user's uid. It is
2751 possible for a "deliver" command or an alias in a "personal" command to
2752 provoke an address rewrite. If logging of address rewriting is configured,
2753 this fails because the process is not running as root or exim. There may be
2754 a better way of dealing with this, but for the moment (because 4.30 needs
2755 to be released), I have disabled address rewrite logging when running a
2756 filter in a non-root, non-exim process.
2762 1. The buildconfig auxiliary program wasn't quoting the value set for
2763 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2764 not defined. This bug was masked in 4.22 by the effect that was fixed in
2767 2. Some messages that were rejected after a message id was allocated were
2768 shown as "incomplete" by exigrep. It no longer does this for messages that
2769 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2771 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2772 have allow_domain_literals set, the ID did not get logged in the <= line.
2773 Domain literals are now always recognized in Message-ID: header lines.
2775 4. The first argument for a ${extract expansion item is the key name or field
2776 number. Leading and trailing spaces in this item were not being ignored,
2777 causing some misleading effects.
2779 5. When deliver_drop_privilege was set, single queue runner processes started
2780 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2781 same command in the process it spins off) were not dropping privilege.
2783 6. When the daemon running as "exim" started a queue runner, it always
2784 re-executed Exim in the spun-off process. This is a waste of effort when
2785 deliver_drop_privilege is set. The new process now just calls the
2786 queue-runner function directly.
2792 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2795 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2796 ignored. Though the use of -oP was forcing the writing of a pid file, it
2797 was always written to the default place.
2799 3. If the message "no IP address found for host xxxx" is generated during
2800 incoming verification, it is now followed by identification of the incoming
2801 connection (so you can more easily find what provoked it).
2803 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2805 5. Added some features to "harden" Exim a bit more against certain attacks:
2807 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2808 be put in Local/Makefile. This is like the never_users runtime option,
2809 but it cannot be overridden. The default setting is "root".
2811 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2812 prefix string with which any file named in a -C command line option
2815 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2816 is retained for -C and -D only if the caller of Exim is root. Without
2817 it, the exim user may also use -C and -D and retain privilege.
2819 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2820 command line option is disabled.
2822 6. Macro names set by the -D option must start with an upper case letter, just
2823 like macro names defined in the configuration file.
2825 7. Added "dereference=" facility to LDAP.
2827 8. Two instances of the typo "uknown" in the source files are fixed.
2829 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2830 the Configure-Makefile script screwed up while processing it.
2832 10. Incorporated PCRE 4.4.
2834 11. The SMTP synchronization check was not operating right at the start of an
2835 SMTP session. For example, it could not catch a HELO sent before the client
2836 waited for the greeting. There is now a check for outstanding input at the
2837 point when the greeting is written. Because of the duplex, asynchronous
2838 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2839 way, but not yet received, when the check is performed.
2841 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2842 on TCP/IP sockets, because this apparently causes some broken clients to
2845 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2846 unchanged) from the Cygwin maintainer.
2848 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2849 was supported instead of testing for mbx. Effectively a typo.
2851 15. The spa authenticator server code was not checking that the input it
2852 received was valid base64.
2854 16. The debug output line for the "set" modifier in ACLs was not showing the
2855 name of the variable that was being set.
2857 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2859 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2860 lookup on its own. Something else was needed to trigger the lookup. For
2861 example, a match in host_lookup or the need for a host name in a host list.
2862 Now, if $sender_host_name is referenced and the host name has not yet been
2863 looked up, a lookup is performed. If the lookup fails, the variable remains
2864 empty, and $host_lookup_failed is set to "1".
2866 19. Added "eqi" as a case-independent comparison operator.
2868 20. The saslauthd authentication condition could segfault if neither service
2869 nor realm was specified.
2871 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2872 error message that was logged was misleading, and incoming SMTP
2873 connections were dropped. The message is now more accurate, and temporary
2874 errors are given to SMTP connections.
2876 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2877 (except RSET) with a 421 error, until QUIT is received. However, it was
2878 failing to send a response to QUIT.
2880 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2881 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2882 ACL is now run after the helo_try_verify_hosts code.
2884 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2885 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2886 case-independent, so other case variants are also recognized). Apparently
2887 some systems use these upper case variants.
2889 25. If more than two messages were waiting for the same host, and a transport
2890 filter was specified for the transport, Exim sent two messages over the
2891 same TCP/IP connection, and then failed with "socket operation on non-
2892 socket" when it tried to send the third.
2894 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2896 27. The extern definition of crypt16() in expand.c was not being excluded when
2897 the OS had its own crypt16() function.
2899 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2900 as a preferred synonym for return_size_limit, both as an option and as an
2903 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2905 30. Changed the default configuration ACL to relax the local part checking rule
2906 for addresses that are not in any local domains. For these addresses,
2907 slashes and pipe symbols are allowed within local parts, but the sequence
2908 /../ is explicitly forbidden.
2910 31. SPA server authentication was not clearing the challenge buffer before
2913 32. log_message in a "warn" ACL statement was writing to the reject log as
2914 well as to the main log, which contradicts the documentation and doesn't
2915 seem right (because no rejection is happening). So I have stopped it.
2917 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2918 However, I am unable to do any testing of this.
2920 34. Fixed an infelicity in the appendfile transport. When checking directories
2921 for a mailbox, to see if any needed to be created, it was accidentally
2922 using path names with one or more superfluous leading slashes; tracing
2923 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2925 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2926 contents are added to the log line that is written for every discarded
2927 recipient. (Previously a log_message setting was ignored.)
2929 36. The ${quote: operator now quotes the string if it is empty.
2931 37. The install script runs exim in order to find its version number. If for
2932 some reason other than non-existence or emptiness, which it checks, it
2933 could not run './exim', it was installing it with an empty version number,
2934 i.e. as "exim-". This error state is now caught, and the installation is
2937 38. An argument was missing from the function that creates an error message
2938 when Exim fails to connect to the socket for saslauthd authentication.
2939 This could cause Exim to crash, or give a corrupted message.
2941 39. Added isip, isip4, and isip6 to ${if conditions.
2943 40. The ACL variables $acl_xx are now saved with the message, and can be
2944 accessed later in routers, transports, and filters.
2946 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2947 strings in the file are not string-expanded.
2949 42. If a MAIL command specified a SIZE value that was too large to fit into an
2950 int variable, the check against message_size_limit failed. Such values are
2951 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2952 day this will have to be increased, but I don't think I want to be around
2953 when emails are that large.
2960 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2961 iconv() is not standard in FreeBSD.
2963 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2964 IPv6 enabled. The observed symptom was a segmentation fault on return from
2965 the function os_common_find_running_interfaces() in src/os.c.
2967 3. In the check_special_case() function in daemon.c I had used "errno" as an
2968 argument name, which causes warnings on some systems. This was basically a
2969 typo, since it was named "eno" in the comments!
2971 4. The code that waits for the clock to tick (at a resolution of some fraction
2972 of a second) so as to ensure message-id uniqueness was always waiting for
2973 at least one whole tick, when it could have waited for less. [This is
2974 almost certainly not relevant at current processor speeds, where it is
2975 unlikely to ever wait at all. But we try to future-proof.]
2977 5. The function that sleeps for a time interval that includes fractions of a
2978 second contained a race. It did not block SIGALRM between setting the
2979 timer, and suspending (a couple of lines later). If the interval was short
2980 and the sigsuspend() was delayed until after it had expired, the suspension
2981 never ended. On busy systems this could lead to processes getting stuck for
2984 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2985 process, before it forks any delivery processes. The open lookup caching
2986 mechanism meant that the open file or database connection was passed into
2987 the delivery process. The problem was that delivery processes always tidy
2988 up cached lookup data. This could cause a problem for the next delivery
2989 process started by the queue runner, because the external queue runner
2990 process does not know about the closure. So the next delivery process
2991 still has data in the lookup cache. In the case of a file lookup, there was
2992 no problem because closing a file descriptor in a subprocess doesn't affect
2993 the parent. However, if the lookup was caching a connection to a database,
2994 the connection was closed, and the second delivery process was likely to
2995 see errors such as "PGSQL: query failed: server closed the connection
2996 unexpectedly". The problem has been fixed by closing all cached lookups
2997 in a queue runner before running a delivery process.
2999 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3000 seem to have the "const" qualifier which it has on other OS. I've
3003 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3004 *of the same type* (client or server) with the same public name that an
3005 error should be diagnosed.
3007 9. When Exim looked up a host name for an IP address, but failed to find the
3008 original IP address when looking up the host name (a safety check), it
3009 output the message "<ip address> does not match any IP for NULL", which was
3010 confusing, to say the least. The bug was that the host name should have
3011 appeared instead of "NULL".
3013 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3014 user that is built into the binary, and the -C or -D options is used, root
3015 privilege is dropped before the configuration file is read. In addition,
3016 logging is switched to stderr instead of the normal log files. If the
3017 configuration then re-defines the Exim user, the unprivileged environment
3018 is probably not what is expected, so Exim logs a panic warning message (but
3021 However, if deliver_drop_privilege is set, the unprivileged state may well
3022 be exactly what is intended, so the warning has been cut out in that case,
3023 and Exim is allowed to try to write to its normal log files.
3029 1. smtp_return_error_details was not giving details for temporary sender
3030 or receiver verification errors.
3032 2. Diagnose a configuration error if two authenticators have the same public
3035 3. Exim used not to create the message log file for a message until the first
3036 delivery attempt. This could be confusing when incoming messages were held
3037 for policy or load reasons. The message log file is now created at the time
3038 the message is received, and an initial "Received" line is written to it.
3040 4. The automatically generated man page for command line options had a minor
3041 bug that caused no ill effects; however, a more serious problem was that
3042 the procedure for building the man page automatically didn't always
3043 operate. Consequently, release 4.20 contains an out-of-date version. This
3044 shouldn't happen again.
3046 5. When building Exim with embedded Perl support, the script that builds the
3047 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3048 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3050 6. The freeze_tell option was not being used for messages that were frozen on
3051 arrival, either by an ACL or by local_scan().
3053 7. Added the smtp_incomplete_transaction log selector.
3055 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3056 was accepting AUTH without a new EHLO.
3058 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3059 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3062 10. Logging of TCP/IP connections (when configured) now happens in the main
3063 daemon process instead of the child process, so that the TCP/IP connection
3064 count is more accurate (but it can never be perfect).
3066 11. The use of "drop" in a nested ACL was not being handled correctly in the
3067 outer ACL. Now, if condition failure induced by the nested "drop" causes
3068 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3069 or "require"), the connection is dropped.
3071 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3072 that yield "discard" can now be used with an "accept" or a "discard" verb,
3073 but an error is generated for any others (because I can't see a useful way
3074 to define what should happen).
3076 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3077 are now processed in the same way as lines in the Exim configuration file.
3078 In particular, continuation lines are supported.
3080 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3082 15. Added -ti meaning -t -i.
3084 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3085 domains, and warn by logging if others are found.
3087 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3088 returned from GIFCONF when Exim is trying to find the list of interfaces on
3089 a host. The code in os.c has been modified to copy each ifreq to an aligned
3090 structure in all cases.
3092 Also, in some cases, the returned ifreq's were being copied to a 'struct
3093 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3094 means the last couple of bytes of an IPv6 address could be chopped if the
3095 ifreq contained only a normal sockaddr (14 bytes storage).
3097 18. Named domain lists were not supported in the hosts_treat_as_local option.
3098 An entry such as +xxxx was not recognized, and was treated as a literal
3101 19. Ensure that header lines added by a DATA ACL are included in the reject log
3102 if the ACL subsequently rejects the message.
3104 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3105 MD5 (which is deprecated).
3107 21. When testing a filter file using -bf, Exim was writing a message when it
3108 took the sender from a "From " line in the message, but it was not doing so
3109 when it took $return_path from a Return-Path: header line. It now does.
3111 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3112 with a valid header line field name (a series of printing characters
3113 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3115 23. Changed "disc" in the source to "disk" to conform to the documentation and
3116 the book and for uniformity.
3118 24. Ignore Sendmail's -Ooption=value command line item.
3120 25. When execve() failed while trying to run a command in a pipe transport,
3121 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3122 could be confused with a return value of 69 from the command itself. This
3123 has been changed to 127, the value the shell returns if it is asked to run
3124 a non-existent command. The wording for the related log line suggests a
3125 non-existent command as the problem.
3127 26. If received_header_text expands to an empty string, do not add a Received:
3128 header line to the message. (Well, it adds a token one on the spool, but
3129 marks it "old" so that it doesn't get used or transmitted.)
3131 27. Installed eximstats 1.28 (addition of -nt option).
3133 28. There was no check for failure on the call to getsockname() in the daemon
3134 code. This can fail if there is a shortage of resources on the system, with
3135 ENOMEM, for example. A temporary error is now given on failure.
3137 29. Contrary to the C standard, it seems that in some environments, the
3138 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3139 program. Exim now does this explicitly; it affects the formatting of
3140 timestamps using strftime().
3142 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3143 complaints. I've now initialized all the variables, to avoid this.
3145 32. Header lines added by a system filter were not being "seen" during
3146 transport-time rewrites.
3148 33. The info_callback() function passed to OpenSSL is set up with type void
3149 (*)(SSL *, int, int), as described somewhere. However, when calling the
3150 function (actually a macro) that sets it up, the type void(*)() is
3151 expected. I've put in a cast to prevent warnings from picky compilers.
3153 34. If a DNS black list lookup found a CNAME record, but there were no A
3154 records associated with the domain it pointed at, Exim crashed.
3156 35. If a DNS black list lookup returned more than one A record, Exim ignored
3157 all but the first. It now scans all returned addresses if a particular IP
3158 value is being sought. In this situation, the contents of the
3159 $dnslist_value variable are a list of all the addresses, separated by a
3162 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3163 to accept a host name and all its aliases if the forward lookup for any of
3164 them yielded the IP address of the incoming connection. Now it accepts only
3165 those names whose forward lookup yields the correct IP address. Any other
3166 names are discarded. This closes a loophole whereby a rogue DNS
3167 administrator could create reverse DNS records to break through a
3168 wildcarded host restriction in an ACL.
3170 37. If a user filter or a system filter that ran in a subprocess used any of
3171 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3172 the wrong values were passed to the pipe command ($thisaddress had the
3173 value of $0, $0 had the value of $1, etc). This bug was introduced by
3174 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3176 38. Improved the line breaking for long SMTP error messages from ACLs.
3177 Previously, if there was no break point between 40 and 75 characters, Exim
3178 left the rest of the message alone. Two changes have been made: (a) I've
3179 reduced the minimum length to 35 characters; (b) if it can't find a break
3180 point between 35 and 75 characters, it looks ahead and uses the first one
3181 that it finds. This may give the occasional overlong line, but at least the
3182 remaining text gets split now.
3184 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3185 file descriptors might be low, and that setting 1000 would always raise it.
3186 It turns out that in some environments, the limit is already over 1000 and
3187 that lowering it causes trouble. So now Exim takes care not to decrease it.
3189 40. When delivering a message, the value of $return_path is set to $sender_
3190 address at the start of routing (routers may change the value). By an
3191 oversight, this default was not being set up when an address was tested by
3192 -bt or -bv, which affected the outcome if any router or filter referred to
3195 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3196 log only when "message" or "log_message" are set. However, if one of the
3197 conditions was an address verification, or a call to a nested ACL, the
3198 messages generated by the underlying test were being passed through. This
3199 no longer happens. The underlying message is available in $acl_verify_
3200 message for both "message" and "log_message" expansions, so it can be
3201 passed through if needed.
3203 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3204 new expansion $bh_ to give the encoded byte string without charset
3205 translation. Translation happens only if iconv() is available; HAVE_ICONV
3206 indicates this at build time. HEADERS_CHARSET gives the charset to
3207 translate to; headers_charset can change it in the configuration, and
3208 "headers charset" can change it in an individual filter file.
3210 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3211 that creates RFC 2047 encoded "words" labels them as that charset instead
3212 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3213 expansion operator; (ii) when Exim creates a From: line for a local
3214 message; (iii) when a header line is rewritten to include a "phrase" part.
3216 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3217 buggy, causing it to skip the first lines of messages whose message ID
3218 ended in 'D'. This would not have bitten before Exim release 4.14, because
3219 message IDs were unlikely to end in 'D' before then. The effect was to have
3220 incorrect size information for certain domains.
3222 45. #include "config.h" was missing at the start of the crypt16.c module. This
3223 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3226 46. If there was a timeout during a "random" callout check, Exim treated it as
3227 a failure of the random address, and carried on sending RSET and the real
3228 address. If the delay was just some slowness somewhere, the response to the
3229 original RCPT would be taken as a response to RSET and so on, causing
3230 mayhem of various kinds.
3232 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3233 when I implemented it. It didn't allow for the fact that some option values
3234 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3235 the right test for positive values.
3237 48. Domain names in DNS records are case-independent. Exim always looks them up
3238 in lower case. Some resolvers return domain names in exactly the case they
3239 appear in the zone file, that is, they may contain uppercase letters. Not
3240 all resolvers do this - some return always lower case. Exim was treating a
3241 change of case by a resolver as a change of domain, similar to a widening
3242 of a domain abbreviation. This triggered its re-routing code and so it was
3243 trying to route what was effectively the same domain again. This normally
3244 caused routing to fail (because the router wouldn't handle the domain
3245 twice). Now Exim checks for this case specially, and just changes the
3246 casing of the domain that it ultimately uses when it transmits the message
3249 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3252 50. If a filter generated a file delivery with a non-absolute name (possible if
3253 no home directory exists for the router), the forbid_file option was not
3256 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3257 the existing equality matching.
3259 52. Exim was using ints instead of ino_t variables in some places where it was
3260 dealing with inode numbers.
3262 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3263 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3264 called TMPDIR, and if it finds it is different, it changes its value.
3266 54. The smtp_printf() function is now made available to local_scan() so
3267 additional output lines can be written before returning. There is also an
3268 smtp_fflush() function to enable the detection of a dropped connection.
3269 The variables smtp_input and smtp_batched_input are exported to
3272 55. Changed the default runtime configuration: the message "Unknown user"
3273 has been removed from the ACL, and instead placed on the localuser router,
3274 using the cannot_route_message feature. This means that any verification
3275 failures that generate their own messages won't get overridden. Similarly,
3276 the "Unrouteable address" message that was in the ACL for unverifiable
3277 relay addresses has also been removed.
3279 56. Added hosts_avoid_esmtp to the smtp transport.
3281 57. The exicyclog script was not checking for the esoteric option
3282 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3283 will work only if exicyclog is run under the appropriate euid.
3285 58. Following a discussion on the list, the rules by which Exim recognises line
3286 endings on incoming messages have been changed. The -dropcr and drop_cr
3287 options are now no-ops, retained only for backwards compatibility. The
3288 following line terminators are recognized: LF CRLF CR. However, special
3289 processing applies to CR:
3291 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3292 nor a local message in the state where . is a terminator.
3294 (ii) If a bare CR is encountered in a header line, an extra space is added
3295 after the line terminator so as not to end the header. The reasoning
3296 behind this is that bare CRs in header lines are most likely either
3297 to be mistakes, or people trying to play silly games.
3299 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3300 was being incorrectly given as 18 bytes larger than it should have been.
3301 This is a VOB (very old bug).
3303 60. This may never have affected anything current, but just in case it has:
3304 When the local host is found other than at the start of a list of hosts,
3305 the local host, those with the same MX, and any that follow, are discarded.
3306 When the list in question was part of a longer list of hosts, the following
3307 hosts (not currently being processed) were also being discarded. This no
3308 longer happens. I'm not sure if this situation could ever has previously
3311 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3314 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3315 header. This is something that is recommended in a new Internet Draft, and
3316 is something that is documented as being done by Sendmail. There are two
3317 possible values. For messages generated by the autoreply transport, Exim
3320 Auto-Submitted: auto-replied
3322 whereas for all other generated messages (e.g. bounces) it adds
3324 Auto-Submitted: auto-generated
3326 63. The "personal" condition in filters now includes a test for the
3327 Auto-Submitted: header. If it contains the string "auto-" the message it
3328 not considered personal.
3330 64. Added rcpt_include_affixes as a generic transport option.
3332 65. Added queue_only_override (default true).
3334 66. Added the syslog_duplication option.
3336 67. If what should have been the first header line of a message consisted of
3337 a space followed by a colon, Exim was mis-interpreting it as a header line.
3338 It isn't of course - it is syntactically invalid and should therefore be
3339 treated as the start of the message body. The misbehaviour could have
3340 caused a number of strange effects, including loss of data in subsequent
3341 header lines, and spool format errors.
3343 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3344 client host had authenticated. This control can now be exercised by an ACL
3345 for more flexibility.
3347 69. By default, callouts do not happen when testing with -bh. There is now a
3348 variant, -bhc, which does actually run the callout code, including
3349 consulting and updating the callout cache.
3351 70. Added support for saslauthd authentication, courtesy of Alexander
3354 71. If statvfs() failed on the spool or log directories while checking their
3355 size for availability, Exim confusingly gave the error "space shortage".
3356 Furthermore, in debugging mode it crashed with a floating point exception.
3357 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3358 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3359 serious problem, Exim now writes to the main and panic logs when this
3360 happens, with details of the failure. It then refuses to accept the
3361 incoming message, giving the message "spool directory problem" or "log
3362 directory problem" with a 421 code for SMTP messages.
3364 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3365 0, 1, and 2 exist, because some OS complain for execs without them (see
3366 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3367 descriptors. However, the code omitted to check that the open succeeded,
3368 causing mysterious errors if for some reason the permissions on /dev/null
3369 got screwed. Now Exim writes a message to the main and panic logs, and
3370 bombs out if it can't open /dev/null.
3372 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3373 interact so that it is all more flexible. It is supposed to remain
3374 backwards compatible. Also added extra_local_interfaces.
3376 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3377 to bomb out with an assertion failure - to the client this appears as a
3378 connection drop. This problem occurs in the part of the code that was taken
3379 from the Samba project. Fortunately, the assertion is in a very simple
3380 function, so I have fixed this by reproducing the function inline in the
3381 one place where it is called, and arranging for authentication to fail
3382 instead of killing the process with assert().
3384 75. The SPA client code was not working when the server requested OEM rather
3385 than Unicode encoding.
3387 76. Added code to make require_files with a specific uid setting more usable in
3388 the case where statting the file as root fails - usually a non-root-mounted
3389 NFS file system. When this happens and the failure is EACCES, Exim now
3390 forks a subprocess and does the per-uid checking as the relevant uid.
3392 77. Added process_log_path.
3394 78. If log_file_path was not explicitly set, a setting of check_log_space or
3395 check_log_inodes was ignored.
3397 79. If a space check for the spool or log partitions fails, the incident is now
3398 logged. Of course, in the latter case the data may get lost...
3400 80. Added the %p formatting code to string_format() so that it can be used to
3401 print addresses in debug_print(). Adjusted all the address printing in the
3402 debugging in store.c to use %p rather than %d.
3404 81. There was a concern that a line of code in smtp_in.c could overflow a
3405 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3406 initially expressed, the concern was not well-founded, because trailing
3407 spaces are removed early. However, if the trailing spaces were followed by
3408 a NULL, they did not get removed, so the overflow was possible. Two fixes
3411 (a) I re-wrote the offending code in a cleaner fashion.
3412 (b) If an incoming SMTP command contains a NULL character, it is rejected
3415 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3416 runs initgroups(), so that if the Exim user is in any additional groups,
3417 they will be used during message reception.
3423 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob