1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.221 2005/09/12 13:50:03 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76 PH/20 If a delivery was routed to a non-standard port by means of an SRV
77 record, the port was not correctly logged when the outgoing_port log
78 selector was set (it logged the transort's default port).
80 PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
87 authentication with radiusclient 0.4.9:
89 - Error returned from rc_read_config was caught wrongly
90 - Username/password not passed on to radius server due to wrong length.
92 The presumption is that some radiusclient API changes for 4.51/PH/17
93 were not taken care of correctly. The code is still untested by me (my
94 Linux distribution still has 0.3.2 of radiusclient), but it was
95 contributed by a Radius user.
97 PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
101 while reading a MIME header. Thanks to Tom Hughes for a patch.
103 PH/24 Include config.h inside local_scan.h so that configuration settings are
106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
107 that in an ACL for RCPT (for example), you can examine exactly what was
110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
111 commands, but it was not correctly comparing the address with the actual
112 client host address. Thus, it would show the EHLO address in Received:
113 header lines when this was not necessary.
115 PH/27 Added the % operator to ${eval:}.
117 PH/28 Exim tries to create and chdir to its spool directory when it starts;
118 it should be ignoring failures (because with -C, for example, it has lost
119 privilege). It wasn't ignoring creation failures other than "already
122 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
125 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
126 because an input line was too long, either on its own, or by virtue of
127 too many continuations, the temporary file was not being removed, and the
128 return code was incorrect.
130 PH/31 Missing "BOOL" in function definition in filtertest.c.
132 PH/32 Applied Sieve patches from the maintainer.
134 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
136 PH/33 Added "verify = not_blind".
138 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
139 Local/Makefile (with some defaults set). These are used in built scripts
140 such as exicyclog, but they have never been used in the exim_install
141 script (though there are many overriding facilities there). I have
142 arranged that the exim_install script now takes note of these two
145 PH/35 Installed configuration files for Dragonfly.
147 PH/36 When a locally submitted message by a trusted user did not contain a
148 From: header, and the sender address was obtained from -f or from an SMTP
149 MAIL command, and the trusted user did not use -F to supply a sender
150 name, $originator_name was incorrectly used when constructing a From:
151 header. Furthermore, $originator_name was used for submission mode
152 messages from external hosts without From: headers in a similar way,
153 which is clearly wrong.
155 PH/37 Added control=suppress_local_fixups.
157 PH/38 When log_selector = +received_sender was set, and the addition of the
158 sender made the log line's construction buffer exactly full, or one byte
159 less than full, an overflow happened when the terminating "\n" was
162 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
163 when the result of a list match is failure because a DNS lookup failed.
165 PH/40 RM_COMMAND is now used in the building process.
171 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
173 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
174 command itself ended in error, the underlying message about the transport
175 filter timeout was being overwritten with the pipe command error. Now the
176 underlying error message should be appended to the second error message.
178 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
180 PH/02 Reduce the amount of output that "make" produces by default. Full output
181 can still be requested.
183 PH/03 The warning log line about a condition test deferring for a "warn" verb
184 was being output only once per connection, rather than after each
185 occurrence (because it was using the same function as for successful
186 "warn" verbs). This seems wrong, so I have changed it.
188 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
189 it should not have, which might have caused a crash in the right
190 circumstances, but probably never did.
192 PH/04 Installed a modified version of Tony Finch's patch to make submission
193 mode fix the return path as well as the Sender: header line, and to
194 add a /name= option so that you can make the user's friendly name appear
197 TF/03 Added the control = fakedefer ACL modifier.
199 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
200 Mark Lowes for thorough testing.
202 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
204 TK/03 Merged latest SRS patch from Miles Wilton.
206 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
207 with the definition in sysexits.h (which is #included earlier).
208 Fortunately, Exim does not actually use EX_OK. The code used to try to
209 preserve the sysexits.h value, by assumimg that macro definitions were
210 scanned for macro replacements. I have been disabused of this notion,
211 so now the code just undefines EX_OK before #including unistd.h.
213 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
214 in the smtp transport. When a block could not be written in a single
215 write() function, the timeout was being re-applied to each part-write.
216 This seems wrong - if the receiver was accepting one byte at a time it
217 would take for ever. The timeout is now adjusted when this happens. It
218 doesn't have to be particularly precise.
220 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
221 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
223 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
224 without a domain if the check to <postmaster@domain> fails.
226 SC/01 Eximstats: added -xls and the ability to specify output files
227 (patch written by Frank Heydlauf).
229 SC/02 Eximstats: use FileHandles for outputing results.
231 SC/03 Eximstats: allow any combination of xls, txt, and html output.
233 SC/04 Eximstats: fixed display of large numbers with -nvr option
235 SC/05 Eximstats: fixed merging of reports with empty tables.
237 SC/06 Eximstats: added the -include_original_destination flag
239 SC/07 Eximstats: removed tabs and trailing whitespace.
241 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
243 TK/06 MBOX spool code: Add real "From " MBOX separator line
244 so the .eml file is really in mbox format (even though
245 most programs do not really care). Patch from Alex Miller.
247 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
248 The latter is generated from $received_to and is only set if the
249 message has one envelope recipient. SA can use these headers,
250 obviously out-of-the-box. Patch from Alex Miller.
252 PH/08 The ${def test on a variable was returning false if the variable's
253 value was "0", contrary to what the specification has always said!
254 The result should be true unless the variable is empty.
256 PH/09 The syntax error of a character other than { following "${if
257 def:variable_name" (after optional whitespace) was not being diagnosed.
258 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
259 accidental colon was present, for example, could give incorrect results.
261 PH/10 Tidied the code in a number of places where the st_size field of a stat()
262 result is used (not including appendfile, where other changes are about
265 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
266 This involved changing a lot of size variables from int to off_t. It
267 should work with maildirs and everything.
269 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
270 spamd dying while we are connected to it.
272 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
275 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
276 function for BDB 4.3.
278 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
280 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
281 the sources for PCRE have been re-arranged and more files are now
284 PH/15 The code I had for printing potentially long long variables in PH/11
285 above was not the best (it lost precision). The length of off_t variables
286 is now inspected at build time, and an appropriate printing format (%ld
287 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
288 to be "long long int" or "long int". This is needed for the internal
289 formatting function string_vformat().
291 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
292 the configuration file to be ":syslog", then the script "guesses" where
293 the logs files are, rather than using the compiled in default. In our
294 case the guess is not the same as the compiled default, so the script
295 suddenly stopped working when I started to use syslog. The patch checks
296 to see if log_file_path is "". If so, it attempts to read it from exim
297 with no configuration file to get the compiled in version, before it
298 falls back to the previous guessing code."
300 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
301 implementing BATV in an Exim configuration. See NewStuff for the gory
304 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
305 Makefile that are specific to HP-UX.
307 PH/18 If the "use_postmaster" option was set for a recipient callout together
308 with the "random" option, the postmaster address was used as the MAIL
309 FROM address for the random test, but not for the subsequent recipient
310 test. It is now used for both.
312 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
313 patch removes a few documentation additions to RFC 3028, because the
314 latest draft now contains them. It adds the new en;ascii-case comparator
315 and a new error check for 8bit text in MIME parts. Comparator and
316 require names are now matched exactly. I enabled the subaddress
317 extension, but it is not well tested yet (read: it works for me)."
319 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
320 rework some of the code of TK/09 above to avoid the hardwired use of
321 "%lld" and "long long". Replaced the call to snprintf() with a call to
324 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
325 records point to non-existent hosts", "retry timeout exceeded", and
326 "retry time not reached for any host after a long failure period".
328 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
329 experimental DomainKeys support:
331 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
332 (2) On an error such as an illegally used "control", the wrong name for
333 the control was given.
335 These problems did NOT occur unless DomainKeys support was compiled.
337 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
339 PH/24 Added ${if match_ip condition.
341 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
342 databases so that it will be absolutely obvious if a crash occurs in the
343 DB library. This is a regular occurrence (often caused by mis-matched
346 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
347 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
348 picked up on a user's system that detects such things. There doesn't seem
349 to be a gcc warning option for this - only an attribute that has to be
350 put on the function's prototype. It seems that in Fedora Core 4 they have
351 set this on a number of new functions. No doubt there will be more in due
354 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
355 specify a transport. However, if an address that was verified by such a
356 router was the subject of a callout, Exim crashed because it tried to
357 read the rcpt_include_affixes from the non-existent transport. Now it
358 just assumes that the setting of that option is false. This bug was
359 introduced by 4.51/PH/31.
361 PH/28 Changed -d+all to exclude +memory, because that information is very
362 rarely of interest, but it makes the output a lot bigger. People tend to
363 do -d+all out of habit.
365 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
366 code in os-type was giving problems when libc.so lives in lib64, like on
369 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
370 aren't the modern standard, and it seems that some systems' include files
371 don't always have them. Exim was already checking for some of the newer
372 ones like T_AAAA, and defining it itself. I've added checks for all the
373 record types that Exim uses.
375 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
376 not automatically generating a new one, as it is supposed to. This
377 prevented TLS from working. If the file did exist, but contained invalid
378 data, a new version was generated, as expected. It was only the case of a
379 non-existent file that was broken.
381 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
382 with a change in libdomainkeys > 0.64.
384 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
385 from DNS. If the selector record carries the flag, it now has
386 precedence over the domain-wide flag.
388 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
390 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
391 the use of a port name that isn't defined in /etc/services) occurred, the
392 message was deferred as in a normal delivery, and thus remained on the
393 spool, instead of being failed because of the mua_wrapper setting. This
394 is now fixed, and I tidied up some of the mua_wrapper messages at the
397 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
398 the messages in a hash of arrays rather than using individual hashes.
399 This is a bit cleaner and results in dramatic memory savings, albeit
400 at a slight CPU cost.
402 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
403 as requested by Marc Sherman.
405 SC/10 Eximstats: added histograms for user specified patterns as requested
408 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
410 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
411 fopen() in the content-scanning modules that did not already have it.
417 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
418 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
420 TK/02 Fix ACL "control" statement not being available in MIME ACL.
422 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
424 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
425 to test Sieve filters that use "vacation".
427 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
428 that changes the way the GnuTLS parameters are stored in the cache file.
429 The new format can be generated externally. For backward compatibility,
430 if the data in the cache doesn't make sense, Exim assumes it has read an
431 old-format file, and it generates new data and writes a new file. This
432 means that you can't go back to an older release without removing the
435 PH/03 A redirect router that has both "unseen" and "one_time" set does not
436 work if there are any delivery delays because "one_time" forces the
437 parent to be marked "delivered", so its unseen clone is never tried
438 again. For this reason, Exim now forbids the simultaneous setting of
441 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
442 redirected to themselves ("homonym" addresses). Read the long ChangeLog
443 entry if you want to know the details. The fix, however, neglected to
444 consider the case when local delivery batching is involved. The test for
445 "previously delivered" was not happening when checking to see if an
446 address could be batched with a previous (undelivered) one; under
447 certain circumstances this could lead to multiple deliveries to the same
450 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
451 in its include files, and this causes problems building Exim.
453 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
454 header_syntax) but Exim was just ignoring anything given after a slash.
455 In particular, this caused confusion with an attempt to use "verify =
456 reverse_host_lookup/defer_ok". An error is now given when options are
457 supplied for verify items that do not have them. (Maybe reverse_host_
458 lookup should have a defer_ok option, but that's a different point.)
460 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
461 defined by RFC 821) to 2048, because there were problems with some AUTH
462 commands, and RFC 1869 says the size should be increased for extended
463 SMTP commands that take arguments.
465 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
468 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
469 "unknown" error; now it says that the functionality isn't in the binary.
471 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
472 an address' error message when a string expansion fails (syntax or
473 whatever). Otherwise the password may appear in the log. Following change
474 PH/42 below, there is no longer a chance of it appearing in a bounce
477 PH/11 Installed exipick version 20050225.0 from John Jetmore.
479 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
480 of its addresses was ever tried. (Bugzilla bug #2.)
482 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
483 the result incorrectly in the debug output. (It correctly added a newline
484 to what was transported.)
486 TF/01 Added $received_time.
488 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
489 commented out examples of how to interface to a virus scanner and to
490 SpamAssassin. Also added commented examples of av_scanner and
491 spamd_address settings.
493 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
494 and controls are allowed in which ACLs. There were a couple of minor
495 errors. Some of the entries in the conditions table (which is a table of
496 where they are NOT allowed) were getting very unwieldy; rewrote them as a
497 negation of where the condition IS allowed.
499 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
501 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
502 header file does not have a version number, so I've had to invent a new
503 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
504 API. The code is untested by me (my Linux distribution still has 0.3.2 of
505 radiusclient), but it was contributed by a Radius user.
507 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
508 files or directories, for OpenSSL.
510 PH/19 When an Exim process that is running as root has to create an Exim log
511 file, it does so in a subprocess that runs as exim:exim so as to get the
512 ownership right at creation (otherwise, other Exim processes might see
513 the file with the wrong ownership). There was no test for failure of this
514 fork() call, which would lead to the process getting stuck as it waited
515 for a non-existent subprocess. Forks do occasionally fail when resources
516 run out. I reviewed all the other calls to fork(); they all seem to check
519 PH/20 When checking for unexpected SMTP input at connect time (before writing
520 the banner), Exim was not dealing correctly with a non-positive return
521 from the read() function. If the client had disconnected by this time,
522 the result was a log entry for a synchronization error with an empty
523 string after "input=" when read() returned zero. If read() returned -1
524 (an event I could not check), uninitialized data bytes were printed.
525 There were reports of junk text (parts of files, etc) appearing after
528 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
530 PH/22 Added support for macro redefinition, and (re)definition in between
531 driver and ACL definitions.
533 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
534 forgetting to use the resulting value; it was using the unexpanded value.
536 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
537 hadn't been configured. The fix is from Juergen Kreileder, who
538 understands it better than I do:
540 "Here's what I see happening with three configured cyrus_sasl
541 authenticators configured (plain, login, cram-md5):
543 On startup auth_cyrus_sasl_init() gets called for each of these.
544 This means three calls to sasl_listmech() without a specified mech_list.
545 => SASL tests which mechs of all available mechs actually work
546 => three warnings about OTP not working
547 => the returned list contains: plain, login, cram-md5, digest-md5, ...
549 With the patch, sasl_listmech() also gets called three times. But now
550 SASL's mech_list option is set to the server_mech specified in the the
551 authenticator. Or in other words, the answer from sasl_listmech()
552 gets limited to just the mech you're testing for (which is different
554 => the return list contains just 'plain' or 'login', 'cram-md5' or
555 nothing depending on the value of ob->server_mech.
557 I've just tested the patch: Authentication still works fine,
558 unavailable mechs specified in the exim configuration are still
559 caught, and the auth.log warnings about OTP are gone."
561 PH/25 When debugging is enabled, the contents of the command line are added
562 to the debugging output, even when log_selector=+arguments is not
565 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
566 answer is "GNU", and only if the return is "GNU/something" is the answer
569 PH/27 $acl_verify_message is now set immediately after the failure of a
570 verification in an ACL, and so is available in subsequent modifiers. In
571 particular, the message can be preserved by coding like this:
573 warn !verify = sender
574 set acl_m0 = $acl_verify_message
576 Previously, $acl_verify_message was set only while expanding "message"
577 and "log_message" when a very denied access.
579 PH/28 Modified OS/os.c-Linux with
581 -#ifndef OS_LOAD_AVERAGE
582 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
584 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
585 nomenclature these days.)
587 PH/29 Installed patch from the Sieve maintainer that adds the options
588 sieve_useraddress and sieve_subaddress to the redirect router.
590 PH/30 In these circumstances:
591 . Two addresses routed to the same list of hosts;
592 . First host does not offer TLS;
593 . First host accepts first address;
594 . First host gives temporary error to second address;
595 . Second host offers TLS and a TLS session is established;
596 . Second host accepts second address.
597 Exim incorrectly logged both deliveries with the TLS parameters (cipher
598 and peerdn, if requested) that were in fact used only for the second
601 PH/31 When doing a callout as part of verifying an address, Exim was not paying
602 attention to any local part prefix or suffix that was matched by the
603 router that accepted the address. It now behaves in the same way as it
604 does for delivery: the affixes are removed from the local part unless
605 rcpt_include_affixes is set on the transport.
607 PH/32 Add the sender address, as F=<...>, to the log line when logging a
608 timeout during the DATA phase of an incoming message.
610 PH/33 Sieve envelope tests were broken for match types other than :is. I have
611 applied a patch sanctioned by the Sieve maintainer.
613 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
614 the uid or gid is negative. A case of a negative gid caused this to be
615 noticed. The fix allows for either to be negative.
617 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
618 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
621 PH/36 The change PH/12 above was broken. Fixed it.
623 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
624 the grounds that routing the same address twice would always produce the
625 same answer. This might have been true once, but it is certainly no
626 longer true now. Routing a child address may depend on the previous
627 routing that produced that child. Some complicated redirection strategies
628 went wrong when messages had multiple recipients, and made Exim's
629 behaviour dependent on the order in which the addresses were given.
631 I have moved the duplicate checking until after the routing is complete.
632 Exim scans the addresses that are assigned to local and remote
633 transports, and removes any duplicates. This means that more work will be
634 done, as duplicates will always all be routed, but duplicates are
635 presumably rare, so I don't expect this is of any significance.
637 For deliveries to pipes, files, and autoreplies, the duplicate checking
638 still happens during the routing process, since they are not going to be
641 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
642 It corrects a timeout issue with spamd. This is Ian's comment: "The
643 background is that sometimes spamd either never reads data from a
644 connection it has accepted, or it never writes response data. The exiscan
645 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
646 blindly assumes that writes won't block so it may never time out."
648 PH/39 Allow G after quota size as well as K and M.
650 PH/40 The value set for $authenticated_id in an authenticator may not contain
651 binary zeroes or newlines because the value is written to log lines and
652 to spool files. There was no check on this. Now the value is run through
653 the string_printing() function so that such characters are converted to
654 printable escape sequences.
656 PH/41 $message_linecount is a new variable that contains the total number of
657 lines in the message. Compare $body_linecount, which is the count for the
660 PH/42 Exim no longer gives details of delivery errors for specific addresses in
661 bounce and delay warning messages, except in certain special cases, which
664 (a) An SMTP error message from a remote host;
665 (b) A message specified in a :fail: redirection;
666 (c) A message specified in a "fail" command in a system filter;
667 (d) A message specified in a FAIL return from the queryprogram router;
668 (e) A message specified by the cannot_route_message router option.
670 In these cases only, Exim does include the error details in bounce and
671 warning messages. There are also a few cases where bland messages such
672 as "unrouteable address" or "local delivery error" are given.
674 PH/43 $value is now also set for the "else" part of a ${run expansion.
676 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
677 being worked on, but at least Exim now implements the latest version to
680 PH/45 In a pipe transport, although a timeout while waiting for the pipe
681 process to complete was treated as a delivery failure, a timeout while
682 writing the message to the pipe was logged, but erroneously treated as a
683 successful delivery. Such timeouts include transport filter timeouts. For
684 consistency with the overall process timeout, these timeouts are now
685 treated as errors, giving rise to delivery failures by default. However,
686 there is now a new Boolean option for the pipe transport called
687 timeout_defer, which, if set TRUE, converts the failures into defers for
688 both kinds of timeout. A transport filter timeout is now identified in
691 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
692 systems where "make" and "gmake" are different, calling "gmake" at top
693 level broke things. I've arranged for the value of $(MAKE) to be passed
694 from the Makefile to this script so that it can call the same version of
698 A note about Exim versions 4.44 and 4.50
699 ----------------------------------------
701 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
702 changes of various kinds. As a consequence, a big documentation update was
703 needed. This delayed the release for rather longer than seemed good, especially
704 in the light of a couple of (minor) security issues. Therefore, the changes
705 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
706 release. So 4.44 and 4.50 are in effect two different branches that both start
709 I have left the 4.50 change log unchanged; it contains all the changes since
710 4.43. The change log for 4.44 is below; many of its items are identical to
711 those for 4.50. This seems to be the most sensible way to preserve the
712 historical information.
718 1. Minor wording change to the doc/README.SIEVE file.
720 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
721 computation of the current number of files was incorrect.
723 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
724 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
727 4. Give more explanation in the error message when the command for a transport
728 filter fails to execute.
730 5. There are several places where Exim runs a non-Exim command in a
731 subprocess. The SIGUSR1 signal should be disabled for these processes. This
732 was being done only for the command run by the queryprogram router. It is
733 now done for all such subprocesses. The other cases are: ${run, transport
734 filters, and the commands run by the lmtp and pipe transports.
736 6. Added CONFIGURE_GROUP build-time option.
738 7. Some older OS have a limit of 256 on the maximum number of file
739 descriptors. Exim was using setrlimit() to set 1000 as a large value
740 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
741 systems. I've change it so that if it can't get 1000, it tries for 256.
743 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
744 was an oversight, and furthermore, ever since the addition of extra
745 controls (e.g. 4.43/32), the checks on when to allow different forms of
746 "control" were broken. There should now be diagnostics for all cases when a
747 control that does not make sense is encountered.
749 9. Added the /retain_sender option to "control=submission".
751 10. $recipients is now available in the predata ACL (oversight).
753 11. Tidy the search cache before the fork to do a delivery from a message
754 received from the command line. Otherwise the child will trigger a lookup
755 failure and thereby defer the delivery if it tries to use (for example) a
756 cached ldap connection that the parent has called unbind on.
758 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
759 of $address_data from the recipient verification was clobbered by the
762 13. The value of address_data from a sender verification is now available in
763 $sender_address_data in subsequent conditions in the ACL statement.
765 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
767 15. Added a new option "connect=<time>" to callout options, to set a different
770 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
771 was its contents. (It was OK if the option was not defined at all.)
773 17. A "Completed" log line is now written for messages that are removed from
774 the spool by the -Mrm option.
776 18. New variables $sender_verify_failure and $recipient_verify_failure contain
777 information about exactly what failed.
779 19. Added -dd to debug only the daemon process.
781 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
782 handles timeouts, both on the server side and network timeouts. Renamed the
783 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
785 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
786 It is now set to "smtps".
788 22. $host_address is now set to the target address during the checking of
791 23. When checking ignore_target_hosts for an ipliteral router, no host name was
792 being passed; this would have caused $sender_host_name to have been used if
793 matching the list had actually called for a host name (not very likely,
794 since this list is usually IP addresses). A host name is now passed as
797 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
798 code that specifies a non-restarting handler (typically sigaction() in
799 modern systems) in an attempt to fix a rare and obscure crash bug.
801 25. Narrowed the window for a race in the daemon that could cause it to ignore
802 SIGCHLD signals. This is not a major problem, because they are used only to
803 wake it up if nothing else does.
805 26. A malformed maildirsize file could cause Exim to calculate negative values
806 for the mailbox size or file count. Odd effects could occur as a result.
807 The maildirsize information is now recalculated if the size or filecount
810 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
811 support for a long time. Removed HAVE_SYS_VFS_H.
813 28. Installed the latest version of exipick from John Jetmore.
815 29. In an address list, if the pattern was not a regular expression, an empty
816 subject address (from a bounce message) matched only if the pattern was an
817 empty string. Non-empty patterns were not even tested. This was the wrong
818 because it is perfectly reasonable to use an empty address as part of a
819 database query. An empty address is now tested by patterns that are
820 lookups. However, all the other forms of pattern expect the subject to
821 contain a local part and a domain, and therefore, for them, an empty
822 address still always fails if the pattern is not itself empty.
824 30. Exim went into a mad DNS loop when attempting to do a callout where the
825 host was specified on an smtp transport, and looking it up yielded more
828 31. Re-factored the code for checking spool and log partition space into a
829 function that finds that data and another that does the check. The former
830 is then used to implement four new variables: $spool_space, $log_space,
831 $spool_inodes, and $log_inodes.
833 32. The RFC2047 encoding function was originally intended for short strings
834 such as real names; it was not keeping to the 75-character limit for
835 encoded words that the RFC imposes. It now respects the limit, and
836 generates multiple encoded words if necessary. To be on the safe side, I
837 have increased the buffer size for the ${rfc2047: expansion operator from
840 33. It is now permitted to omit both strings after an "if" condition; if the
841 condition is true, the result is "true". As before, when the second string
842 is omitted, a false condition yields an empty string. This makes it less
843 cumbersome to write custom ACL and router conditions.
845 34. Failure to deliver a bounce message always caused it to be frozen, even if
846 there was an errors_to setting on the router. The errors_to setting is now
849 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
850 canonical form (fully expanded) before being placed in
851 $sender_host_address.
853 36. The table in the code that translates DNS record types into text (T_A to
854 "A" for instance) was missing entries for NS and CNAME. It is just possible
855 that this could have caused confusion if both these types were looked up
856 for the same domain, because the text type is used as part of Exim's
857 per-process caching. But the chance of anyone hitting this buglet seems
860 37. The dnsdb lookup has been extended in a number of ways.
862 (1) There is a new type, "zns", which walks up the domain tree until it
863 finds some nameserver records. It should be used with care.
865 (2) There is a new type, "mxh", which is like "mx" except that it returns
866 just the host names, not the priorities.
868 (3) It is now possible to give a list of domains (or IP addresses) to be
869 looked up. The behaviour when one of the lookups defers can be
870 controlled by a keyword.
872 (4) It is now possible to specify the separator character for use when
873 multiple records are returned.
875 38. The dnslists ACL condition has been extended: it is now possible to supply
876 a list of IP addresses and/or domains to be looked up in a particular DNS
879 39. Added log_selector=+queue_time_overall.
881 40. When running the queue in the test harness, wait just a tad after forking a
882 delivery process, to get repeatability of debugging output.
884 41. Include certificate and key file names in error message when GnuTLS fails
885 to set them up, because the GnuTLS error message doesn't include the name
886 of the failing file when there is a problem reading it.
888 42. Allow both -bf and -bF in the same test run.
890 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
892 44. The "Exiscan patch" is now merged into the mainline Exim source.
894 45. Sometimes the final signoff response after QUIT could fail to get
895 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
896 < 0 before doing a fflush(). This bug looks as though it goes back to the
897 introduction of TLS in release 3.20, but "sometimes" must have been rare
898 because the tests only now provoked it.
900 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
901 (this can affect the format of dates).
903 47. exim_tidydb, when checking for the continued existence of a message for
904 which it has found a message-specific retry record, was not finding
905 messages that were in split spool directories. Consequently, it was
906 deleting retry records that should have stayed in existence.
908 48. Steve fixed some bugs in eximstats.
910 49. The SPA authentication driver was not abandoning authentication and moving
911 on to the next authenticator when an expansion was forced to fail,
912 contradicting the general specification for all authenticators. Instead it
913 was generating a temporary error. It now behaves as specified.
915 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
916 (the order specifies the preference for clients). The order is now AES256,
917 AES128, 3DES, ARCFOUR128.
919 51. Small patch to Sieve code - explicitly set From: when generating an
922 52. Exim crashed if a remote delivery caused a very long error message to be
923 recorded - for instance if somebody sent an entire SpamAssassin report back
924 as a large number of 550 error lines. This bug was coincidentally fixed by
925 increasing the size of one of Exim's internal buffers (big_buffer) that
926 happened as part of the Exiscan merge. However, to be on the safe side, I
927 have made the code more robust (and fixed the comments that describe what
930 53. Now that there can be additional text after "Completed" in log lines (if
931 the queue_time_overall log selector is set), a one-byte patch to exigrep
932 was needed to allow it to recognize "Completed" as not the last thing in
935 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
936 patch that reportedly fixes this has been added. I am not expert enough to
937 create a test for it. This is what the patch creator wrote:
939 "I found a little strange behaviour of ldap code when working with
940 Windows 2003 AD Domain, where users was placed in more than one
941 Organization Units. When I tried to give exim partial DN, the exit code
942 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
943 But simultaneously result of request was absolutely normal ldap result,
944 so I produce this patch..."
946 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
947 REFERENCE, so I have modified the code to exclude the patch when that macro
950 55. Some experimental protocols are using DNS PTR records for new purposes. The
951 keys for these records are domain names, not reversed IP addresses. The
952 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
953 leaves it alone. Component reversal etc. now happens only for IP addresses.
956 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
958 57. Double the size of the debug message buffer (to 2048) so that more of very
959 long debug lines gets shown.
961 58. The exicyclog utility now does better if the number of log files to keep
962 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
964 59. Two changes related to the smtp_active_hostname option:
966 (1) $smtp_active_hostname is now available as a variable.
967 (2) The default for smtp_banner uses $smtp_active_hostname instead
968 of $primary_hostname.
970 60. The host_aton() function is supposed to be passed a string that is known
971 to be a valid IP address. However, in the case of IPv6 addresses, it was
972 not checking this. This is a hostage to fortune. Exim now panics and dies
973 if the condition is not met. A case was found where this could be provoked
974 from a dnsdb PTR lookup with an IPv6 address that had more than 8
975 components; fortuitously, this particular loophole had already been fixed
976 by change 4.50/55 above.
978 If there are any other similar loopholes, the new check in host_aton()
979 itself should stop them being exploited. The report I received stated that
980 data on the command line could provoke the exploit when Exim was running as
981 exim, but did not say which command line option was involved. All I could
982 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
986 61. There was a buffer overflow vulnerability in the SPA authentication code
987 (which came originally from the Samba project). I have added a test to the
988 spa_base64_to_bits() function which I hope fixes it.
991 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
992 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
994 63. The daemon start-up calls getloadavg() while still root for those OS that
995 need the first call to be done as root, but it missed one case: when
996 deliver_queue_load_max is set with deliver_drop_privilege. This is
997 necessary for the benefit of the queue runner, because there is no re-exec
998 when deliver_drop_privilege is set.
1000 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1001 This has been fixed.
1003 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1004 was in use, was not putting the data itself into the right store pool;
1005 consequently, it could be overwritten for a subsequent message in the same
1006 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1009 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1011 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1012 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1013 the function were treating the return as a boolean value, which happened to
1014 work because 0=false and not-0=true, but is not correct code.
1016 68. The host_aton() function was not handling scoped IPv6 addresses (those
1017 with, for example, "%eth0" on the end) correctly.
1019 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1020 negated items (that is, ~something) in unsigned ints. Some compilers
1021 apparently mutter when there is no cast.
1023 70. If an address verification called from an ACL failed, and did not produce a
1024 user-specific message (i.e. there was only a "system" message), nothing was
1025 put in $acl_verify_message. In this situation, it now puts the system
1028 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1029 session; change 4.31/43 added the unwanted input to the log line - except
1030 that it did not do this in the start of session case. It now does.
1032 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1033 This is wrong and can cause the other end to generate a synchronization
1034 error if it is another Exim or anything else that does the synchronization
1035 check. A QUIT command is no longer sent after a timeout.
1037 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1038 during host lookups.
1040 74. The defer_ok option of callout verification was not working if it was used
1041 when verifying addresses in header lines, that is, for this case:
1043 verify = header_sender/callout=defer_ok
1045 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1046 those file descriptors could be used for SMTP connections. If anything
1047 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1048 could be sent to the SMTP client, causing chaos. The daemon now opens
1049 stdin, stdout, and stderr to /dev/null when it puts itself into the
1052 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1053 log by default. The user can override this with suitable Perl magic.
1055 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1056 the log message when discard triggers, was not working for the DATA ACL or
1057 for the non-SMTP ACL.
1059 78. Error message wording change in sieve.c.
1061 79. If smtp_accept_max_per_host was set, the number of connections could be
1062 restricted to fewer than expected, because the daemon was trying to set up
1063 a new connection before checking whether the processes handling previous
1064 connections had finished. The check for completed processes is now done
1065 earlier. On busy systems, this bug wouldn't be noticed because something
1066 else would have woken the daemon, and it would have reaped the completed
1069 80. If a message was submitted locally by a user whose login name contained one
1070 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1071 It caused a spool format error. I have fixed the spool reading code. A
1072 related problem was that the "from" clause in the Received: line became
1073 illegal because of the space(s). It is now covered by ${quote_local_part.
1075 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1078 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1080 83. Patch from Sieve maintainer to fix unterminated string problem in
1081 "vacation" handling.
1083 84. Some minor changes to the Linux configuration files to help with other
1084 OS variants using glibc.
1086 85. One more patch for Sieve to update vacation handling to latest spec.
1089 ----------------------------------------------------
1090 See the note above about the 4.44 and 4.50 releases.
1091 ----------------------------------------------------
1097 1. Change 4.43/35 introduced a bug that caused file counts to be
1098 incorrectly computed when quota_filecount was set in an appendfile
1101 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1102 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1105 3. Give more explanation in the error message when the command for a transport
1106 filter fails to execute.
1108 4. There are several places where Exim runs a non-Exim command in a
1109 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1110 was being done only for the command run by the queryprogram router. It is
1111 now done for all such subprocesses. The other cases are: ${run, transport
1112 filters, and the commands run by the lmtp and pipe transports.
1114 5. Some older OS have a limit of 256 on the maximum number of file
1115 descriptors. Exim was using setrlimit() to set 1000 as a large value
1116 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1117 systems. I've change it so that if it can't get 1000, it tries for 256.
1119 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1120 was an oversight, and furthermore, ever since the addition of extra
1121 controls (e.g. 4.43/32), the checks on when to allow different forms of
1122 "control" were broken. There should now be diagnostics for all cases when a
1123 control that does not make sense is encountered.
1125 7. $recipients is now available in the predata ACL (oversight).
1127 8. Tidy the search cache before the fork to do a delivery from a message
1128 received from the command line. Otherwise the child will trigger a lookup
1129 failure and thereby defer the delivery if it tries to use (for example) a
1130 cached ldap connection that the parent has called unbind on.
1132 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1133 of $address_data from the recipient verification was clobbered by the
1134 sender verification.
1136 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1137 was its contents. (It was OK if the option was not defined at all.)
1139 11. A "Completed" log line is now written for messages that are removed from
1140 the spool by the -Mrm option.
1142 12. $host_address is now set to the target address during the checking of
1143 ignore_target_hosts.
1145 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1146 being passed; this would have caused $sender_host_name to have been used if
1147 matching the list had actually called for a host name (not very likely,
1148 since this list is usually IP addresses). A host name is now passed as
1151 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1152 code that specifies a non-restarting handler (typically sigaction() in
1153 modern systems) in an attempt to fix a rare and obscure crash bug.
1155 15. Narrowed the window for a race in the daemon that could cause it to ignore
1156 SIGCHLD signals. This is not a major problem, because they are used only to
1157 wake it up if nothing else does.
1159 16. A malformed maildirsize file could cause Exim to calculate negative values
1160 for the mailbox size or file count. Odd effects could occur as a result.
1161 The maildirsize information is now recalculated if the size or filecount
1164 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1165 support for a long time. Removed HAVE_SYS_VFS_H.
1167 18. Updated exipick to current release from John Jetmore.
1169 19. Allow an empty sender to be matched against a lookup in an address list.
1170 Previously the only cases considered were a regular expression, or an
1173 20. Exim went into a mad DNS lookup loop when doing a callout where the
1174 host was specified on the transport, if the DNS lookup yielded more than
1177 21. The RFC2047 encoding function was originally intended for short strings
1178 such as real names; it was not keeping to the 75-character limit for
1179 encoded words that the RFC imposes. It now respects the limit, and
1180 generates multiple encoded words if necessary. To be on the safe side, I
1181 have increased the buffer size for the ${rfc2047: expansion operator from
1184 22. Failure to deliver a bounce message always caused it to be frozen, even if
1185 there was an errors_to setting on the router. The errors_to setting is now
1188 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1189 canonical form (fully expanded) before being placed in
1190 $sender_host_address.
1192 24. Updated eximstats to version 1.33
1194 25. Include certificate and key file names in error message when GnuTLS fails
1195 to set them up, because the GnuTLS error message doesn't include the name
1196 of the failing file when there is a problem reading it.
1198 26. Expand error message when OpenSSL has problems setting up cert/key files.
1201 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1202 (this can affect the format of dates).
1204 28. exim_tidydb, when checking for the continued existence of a message for
1205 which it has found a message-specific retry record, was not finding
1206 messages that were in split spool directories. Consequently, it was
1207 deleting retry records that should have stayed in existence.
1209 29. eximstats updated to version 1.35
1210 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1211 1.35 - bugfix such that pie charts by volume are generated correctly
1213 30. The SPA authentication driver was not abandoning authentication and moving
1214 on to the next authenticator when an expansion was forced to fail,
1215 contradicting the general specification for all authenticators. Instead it
1216 was generating a temporary error. It now behaves as specified.
1218 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1219 (the order specifies the preference for clients). The order is now AES256,
1220 AES128, 3DES, ARCFOUR128.
1222 31. Small patch to Sieve code - explicitly set From: when generating an
1225 32. Exim crashed if a remote delivery caused a very long error message to be
1226 recorded - for instance if somebody sent an entire SpamAssassin report back
1227 as a large number of 550 error lines. This bug was coincidentally fixed by
1228 increasing the size of one of Exim's internal buffers (big_buffer) that
1229 happened as part of the Exiscan merge. However, to be on the safe side, I
1230 have made the code more robust (and fixed the comments that describe what
1233 33. Some experimental protocols are using DNS PTR records for new purposes. The
1234 keys for these records are domain names, not reversed IP addresses. The
1235 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1236 leaves it alone. Component reversal etc. now happens only for IP addresses.
1239 34. The host_aton() function is supposed to be passed a string that is known
1240 to be a valid IP address. However, in the case of IPv6 addresses, it was
1241 not checking this. This is a hostage to fortune. Exim now panics and dies
1242 if the condition is not met. A case was found where this could be provoked
1243 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1244 components; fortuitously, this particular loophole had already been fixed
1245 by change 4.50/55 or 4.44/33 above.
1247 If there are any other similar loopholes, the new check in host_aton()
1248 itself should stop them being exploited. The report I received stated that
1249 data on the command line could provoke the exploit when Exim was running as
1250 exim, but did not say which command line option was involved. All I could
1251 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1252 running as the user.
1255 35. There was a buffer overflow vulnerability in the SPA authentication code
1256 (which came originally from the Samba project). I have added a test to the
1257 spa_base64_to_bits() function which I hope fixes it.
1260 36. The daemon start-up calls getloadavg() while still root for those OS that
1261 need the first call to be done as root, but it missed one case: when
1262 deliver_queue_load_max is set with deliver_drop_privilege. This is
1263 necessary for the benefit of the queue runner, because there is no re-exec
1264 when deliver_drop_privilege is set.
1266 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1267 was in use, was not putting the data itself into the right store pool;
1268 consequently, it could be overwritten for a subsequent message in the same
1269 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1272 38. Sometimes the final signoff response after QUIT could fail to get
1273 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1274 < 0 before doing a fflush(). This bug looks as though it goes back to the
1275 introduction of TLS in release 3.20, but "sometimes" must have been rare
1276 because the tests only now provoked it.
1282 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1283 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1284 Now it returns an integer. A number of calls were still expecting a T/F
1285 return. Fortuitously, in all cases, the tests worked in OK situations,
1286 which is the norm. However, things would have gone wrong on any write
1287 failures on the smtp file descriptor. This function is used when sending
1288 messages over SMTP and also when doing verify callouts.
1290 2. When Exim is called to do synchronous delivery of a locally submitted
1291 message (the -odf or -odi options), it no longer closes stderr before doing
1294 3. Implemented the mua_wrapper option.
1296 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1298 5. Implemented the functions header_remove(), header_testname(),
1299 header_add_at_position(), and receive_remove_recipient(), and exported them
1302 6. If an ACL "warn" statement specified the addition of headers, Exim already
1303 inserted X-ACL-Warn: at the start if there was no header name. However, it
1304 was not making this test for the second and subsequent header lines if
1305 there were newlines in the string. This meant that an invalid header could
1306 be inserted if Exim was badly configured.
1308 7. Allow an ACL "warn" statement to add header lines at the start or after all
1309 the Received: headers, as well as at the end.
1311 8. Added the rcpt_4xx retry error code.
1313 9. Added postmaster_mailfrom=xxx to callout verification option.
1315 10. Added mailfrom=xxxx to the callout verification option, for verify=
1318 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1319 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1321 12. Inserted some casts to stop certain compilers warning when using pointer
1322 differences as field lengths or precisions in printf-type calls (mostly
1323 affecting debugging statements).
1325 13. Added optional readline() support for -be (dynamically loaded).
1327 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1328 same clock tick as a message's arrival, so that its received time was the
1329 same as the "first fail" time on the retry record, and that message
1330 remained on the queue past the ultimate address timeout, every queue runner
1331 would try a delivery (because it was past the ultimate address timeout) but
1332 after another failure, the ultimate address timeout, which should have then
1333 bounced the address, did not kick in. This was a "< instead of <=" error;
1334 in most cases the first failure would have been in the next clock tick
1335 after the received time, and all would be well.
1337 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1338 being recognized when the domain list was tested by the match_domain
1339 condition in an expansion string.
1341 16. Added the ${str2b64: operator.
1343 17. Exim was always calling setrlimit() to set a large limit for the number of
1344 processes, without checking whether the existing limit was already
1345 adequate. (It did check for the limit on file descriptors.) Furthermore,
1346 errors from getrlimit() and setrlimit() were being ignored. Now they are
1347 logged to the main and panic logs, but Exim does carry on, to try to do its
1348 job under whatever limits there are.
1350 18. Imported PCRE 5.0.
1352 19. Trivial typo in log message " temporarily refused connection" (the leading
1355 20. If the log selector return_path_on_delivery was set and an address was
1356 redirected to /dev/null, the delivery process crashed because it assumed
1357 that a return path would always be set for a "successful" delivery. In this
1358 case, the whole delivery is bypassed as an optimization, and therefore no
1361 21. Internal re-arrangement: the function for sending a challenge and reading
1362 a response while authentication was assuming a zero-terminated challenge
1363 string. It's now changed to take a pointer and a length, to allow for
1364 binary data in such strings.
1366 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1368 23. Exim was not respecting finduser_retries when seeking the login of the
1369 uid under which it was called; it was always trying 10 times. (The default
1370 setting of finduser_retries is zero.) Also, it was sleeping after the final
1371 failure, which is pointless.
1373 24. Implemented tls_on_connect_ports.
1375 25. Implemented acl_smtp_predata.
1377 26. If the domain in control=submission is set empty, Exim assumes that the
1378 authenticated id is a complete email address when it generates From: or
1379 Sender: header lines.
1381 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1382 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1383 chown and chgrp in /bin and hostname in /usr/bin.
1385 28. Exim was keeping the "process log" file open after each use, just as it
1386 does for the main log. This opens the possibility of it remaining open for
1387 long periods when the USR1 signal hits a daemon. Occasional processlog
1388 errors were reported, that could have been caused by this. Anyway, it seems
1389 much more sensible not to leave this file open at all, so that is what now
1392 29. The long-running daemon process does not normally write to the log once it
1393 has entered its main loop, and it closes the log before doing so. This is
1394 so that log files can straightforwardly be renamed and moved. However,
1395 there are a couple of unusual error situations where the daemon does write
1396 log entries, and I had neglected to close the log afterwards.
1398 30. The text of an SMTP error response that was received during a remote
1399 delivery was being truncated at 512 bytes. This is too short for some of
1400 the long messages that one sometimes sees. I've increased the limit to
1403 31. It is now possible to make retry rules that apply only when a message has a
1404 specific sender, in particular, an empty sender.
1406 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1407 it possible to be selective about when SMTP synchronization is enforced.
1409 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1411 32. Implemented hosts_connection_nolog.
1413 33. Added an ACL for QUIT.
1415 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1418 35. Added mailbox_size and mailbox_filecount to appendfile.
1420 36. Added control = no_multiline_responses to ACLs.
1422 37. There was a bug in the logic of the code that waits for the clock to tick
1423 in the case where the clock went backwards by a substantial amount such
1424 that the microsecond fraction of "now" was more than the microsecond
1425 fraction of "then" (but the whole seconds number was less).
1427 38. Added support for the libradius Radius client library this is found on
1428 FreeBSD (previously only the radiusclient library was supported).
1434 1. When certain lookups returned multiple values in the form name=value, the
1435 quoting of the values was not always being done properly. Specifically:
1436 (a) If the value started with a double quote, but contained no whitespace,
1438 (b) If the value contained whitespace other than a space character (i.e.
1439 tabs or newlines or carriage returns) it was not quoted.
1440 This fix has been applied to the mysql and pgsql lookups by writing a
1441 separate quoting function and calling it from the lookup code. The fix
1442 should probably also be applied to nisplus, ibase and oracle lookups, but
1443 since I cannot test any of those, I have not disturbed their existing code.
1445 2. A hit in the callout cache for a specific address caused a log line with no
1446 reason for rejecting RCPT. Now it says "Previous (cached) callout
1447 verification failure".
1449 3. There was an off-by-one bug in the queryprogram router. An over-long
1450 return line was truncated at 256 instead of 255 characters, thereby
1451 overflowing its buffer with the terminating zero. As well as fixing this, I
1452 have increased the buffer size to 1024 (and made a note to document this).
1454 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1455 when Exim is waiting for an SMTP response from a remote server, Exim
1456 restarts its select() call on the socket, thereby resetting its timeout.
1457 This is not a problem when such interrupts are rare. Somebody set up a cron
1458 job to run exiwhat every 2 minutes, which is less than the normal select()
1459 timeout (5 or 10 minutes). This meant that the select() timeout never
1460 kicked in because it was always reset. I have fixed this by comparing the
1461 time when an interrupt arrives with the time at the start of the first call
1462 to select(). If more time than the timeout has elapsed, the interrupt is
1463 treated as a timeout.
1465 5. Some internal re-factoring in preparation for the addition of Sieve
1466 extensions (by MH). In particular, the "personal" test is moved to a
1467 separate function, and given an option for scanning Cc: and Bcc: (which is
1468 not set for Exim filters).
1470 6. When Exim created an email address using the login of the caller as the
1471 local part (e.g. when creating a From: or Sender: header line), it was not
1472 quoting the local part when it contained special characters such as @.
1474 7. Installed new OpenBSD configuration files.
1476 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1477 try to make them clearer.
1479 9. Callout options, other than the timeout value, were being ignored when
1480 verifying sender addresses in header lines. For example, when using
1482 verify = header_sender/callout=no_cache
1484 the cache was (incorrectly) being used.
1486 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1487 only the Cygwin environment.
1489 11. When return_path_on_delivery was set as a log selector, if different remote
1490 addresses in the same message used different return paths and parallel
1491 remote delivery occurred, the wrong values would sometimes be logged.
1492 (Whenever a remote delivery process finished, the return path value from
1493 the most recently started remote delivery process was logged.)
1495 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1496 lines when AUTH and/or TLS are in use. This is the "received protocol"
1497 field. Exim used to use "asmtp" for authenticated SMTP, without any
1498 indication (in the protocol name) for TLS use. Now it follows the RFC and
1499 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1500 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1501 names appear in log lines as well as in Received: header lines.
1503 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1504 extensions, and comparison tests, and to fix some bugs.
1506 14. Changes to the "personal" filter test:
1508 (1) The test was buggy in that it was just doing the equivalent of
1509 "contains" tests on header lines. For example, if a user's address was
1510 anne@some.where, the "personal" test would incorrectly be true for
1512 To: susanne@some.where
1514 This test is now done by extracting each address from the header in turn,
1515 and checking the entire address. Other tests that are part of "personal"
1516 are now done using regular expressions (for example, to check local parts
1517 of addresses in From: header lines).
1519 (2) The list of non-personal local parts in From: addresses has been
1520 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1521 taken from the Sieve specification recommendations.
1523 (3) If the message contains any header line starting with "List-" it is
1524 treated as non-personal.
1526 (4) The test for "circular" in the Subject: header line has been removed
1527 because it now seems ill-conceived.
1529 15. Minor typos in src/EDITME comments corrected.
1531 16. Installed latest exipick from John Jetmore.
1533 17. If headers_add on a router specified a text string that was too long for
1534 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1535 of string_sprintf() is now avoided.
1537 18. $message_body_size was not set (it was always zero) when running the DATA
1538 ACL and the local_scan() function.
1540 19. For the "mail" command in an Exim filter, no default was being set for
1541 the once_repeat time, causing a random time value to be used if "once" was
1542 specified. (If the value happened to be <= 0, no repeat happened.) The
1543 default is now 0s, meaning "never repeat". The "vacation" command was OK
1544 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1545 (I found it when inspecting the code).
1547 20. There is now an overall timeout for performing a callout verification. It
1548 defaults to 4 times the callout timeout, which applies to individual SMTP
1549 commands during the callout. The overall timeout applies when there is more
1550 than one host that can be tried. The timeout is checked before trying the
1551 next host. This prevents very long delays if there are a large number of
1552 hosts and all are timing out (e.g. when the network connections are timing
1553 out). The value of the overall timeout can be changed by specifying an
1554 additional sub-option for "callout", called "maxwait". For example:
1556 verify = sender/callout=5s,maxwait=20s
1558 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1559 to the end before writing, but this should make it even safer).
1561 22. Exim was forgetting that it had advertised PIPELINING for the second and
1562 subsequent messages on an SMTP connection. It was also not resetting its
1563 memory on STARTTLS and an internal HELO.
1565 23. When Exim logs an SMTP synchronization error within a session, it now
1566 records whether PIPELINING has been advertised or not.
1568 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1569 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1570 rather than long int.
1572 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1574 26. Added the never_mail option to autoreply.
1580 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1581 crash if the getsockname() call failed; this can happen if a connection is
1582 closed very soon after it is established. The problem was simply in the
1583 order in which certain operations were done, causing Exim to try to write
1584 to the SMTP stream before it had set up the file descriptor. The bug has
1585 been fixed by making things happen in the correct order.
1591 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1592 before the connection was closed, thus losing the rejection response.
1594 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1595 some early Solaris releases, but causes trouble in current releases where
1596 socklen_t is defined.
1598 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1601 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1604 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1605 that releasing the top of it at the end releases what was used for sub-
1606 expansions (unless the block got too big). However, discard this block if
1607 the first thing is a variable or header, so that we can use its block when
1608 it is dynamic (useful for very large $message_headers, for example).
1610 6. Lookups now cache *every* query, not just the most recent. A new, separate
1611 store pool is used for this. It can be recovered when all lookup caches are
1612 flushed. Lookups now release memory at the end of their result strings.
1613 This has involved some general refactoring of the lookup sources.
1615 7. Some code has been added to the store_xxx() functions to reduce the amount
1616 of flapping under certain conditions.
1618 8. log_incoming_interface used to affect only the <= reception log lines. Now
1619 it causes the local interface and port to be added to several more SMTP log
1620 lines, for example "SMTP connection from", and rejection lines.
1622 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1624 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1626 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1627 could be overwritten at the end of the current message (or the start of a
1628 new message if it was set in a HELO ACL). The value is now preserved for
1629 the duration of the SMTP connection.
1631 12. If a transport had a headers_rewrite setting, and a matching header line
1632 contained an unqualified address, that address was qualified, even if it
1633 did not match any rewriting rules. The underlying bug was that the values
1634 of the flags that permit the existence of unqualified sender and recipient
1635 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1636 non-local messages, and by -bnq for local messages) were not being
1637 preserved with the message after it was received.
1639 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1640 "next input=" as part of the text comprising the host identity instead of
1641 the correct text. The code was using the same buffer for two different
1642 strings. However, depending on which order the printing function evaluated
1643 its arguments, the bug did not always show up. Under Linux, for example, my
1644 test suite worked just fine.
1646 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1647 doesn't work with some older versions of Perl. It has been changed to "my",
1648 which in any case is probably the better facility to use.
1650 15. A really picky compiler found some instances of statements for creating
1651 error messages that either had too many or two few arguments for the format
1654 16. The size of the buffer for calls to the DNS resolver has been increased
1655 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1656 for addresses that have a lot of PTR records. This alleviates a problem; it
1657 does not fully solve it.
1659 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1660 buffer now truncates the list and logs the incident, which is the same
1661 action as happens when Exim is looking up a host name and its aliases.
1662 Previously in this situation something unpredictable would happen;
1663 sometimes it was "internal error: store_reset failed".
1665 18. If a server dropped the connection unexpectedly when an Exim client was
1666 using GnuTLS and trying to read a response, the client delivery process
1667 crashed while trying to generate an error log message.
1669 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1670 string, for example:
1672 warn message = H1: something\nH2: something
1674 the text was added as a single header line from Exim's point of view
1675 though it ended up OK in the delivered message. However, searching for the
1676 second and subsequent header lines using $h_h2: did not work. This has been
1677 fixed. Similarly, if a system filter added multiple headers in this way,
1678 the routers could not see them.
1680 20. Expanded the error message when iplsearch is called with an invalid key to
1681 suggest using net-iplsearch in a host list.
1683 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1684 ACLs are no longer actually imposed (and a message to that effect is
1687 22. If a "gecos" field in a passwd entry contained escaped characters, in
1688 particular, if it contained a \" sequence, Exim got it wrong when building
1689 a From: or a Sender: header from that name. A second bug also caused
1690 incorrect handling when an unquoted " was present following a character
1691 that needed quoting.
1693 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1694 was not being matched caselessly.
1696 24. Arranged for all hyphens in the exim.8 source to be escaped with
1699 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1700 properly. Recipient callout cache records were still being keyed to include
1701 the sender, even when use_sender was set false. This led to far more
1702 callouts that were necessary. The sender is no longer included in the key
1703 when use_sender is false.
1705 26. Added "control = submission" modifier to ACLs.
1707 27. Added the ${base62d: operator to decode base 62 numbers.
1709 28. dnsdb lookups can now access SRV records.
1711 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1712 the configuration file.
1714 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1715 (-v) mode. This makes the output for a verbose queue run more intelligible.
1717 31. Added a use_postmaster feature to recipient callouts.
1719 32. Added the $body_zerocount variable, containing the number of binary zero
1720 bytes in the message body.
1722 33. The time of last modification of the "new" subdirectory is now used as the
1723 "mailbox time last read" when there is a quota error for a maildir
1726 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1728 35. Added +ignore_unknown as a special item in host lists.
1730 36. Code for decoding IPv6 addresses in host lists is now included, even if
1731 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1732 address was recognized as an IP address, but was then not correctly decoded
1733 into binary, causing unexpected and incorrect effects when compared with
1740 1. Very minor rewording of debugging text in manualroute to say "list of
1741 hosts" instead of "hostlist".
1743 2. If verify=header_syntax was set, and a header line with an unqualified
1744 address (no domain) and a large number of spaces between the end of the
1745 name and the colon was received, the reception process suffered a buffer
1746 overflow, and (when I tested it) crashed. This was caused by some obsolete
1747 code that should have been removed. The fix is to remove it!
1749 3. When running in the test harness, delay a bit after writing a bounce
1750 message to get a bit more predictability in the log output.
1752 4. Added a call to search_tidyup() just before forking a reception process. In
1753 theory, someone could use a lookup in the expansion of smtp_accept_max_
1754 per_host which, without the tidyup, could leave open a database connection.
1756 5. Added the variables $recipient_data and $sender_data which get set from a
1757 lookup success in an ACL "recipients" or "senders" condition, or a router
1758 "senders" option, similar to $domain_data and $local_part_data.
1760 6. Moved the writing of debug_print from before to after the "senders" test
1763 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1764 problems for message scanning, either using a data ACL, or using
1765 local_scan() because the Received: header was not generated till after they
1766 were called (in order to set the time as the time of reception completion).
1767 I have revised the way this works. The header is now generated after the
1768 body is received, but before the ACL or local_scan() are called. After they
1769 are run, the timestamp in the header is updated.
1775 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1776 before starting a queue runner without re-exec. This happened only when
1777 deliver_drop_privilege was set or when the Exim user was set to root. The
1778 effect of the bug was that timeouts during subsequent deliveries caused
1779 crashes instead of being properly handled. The handler is now left at its
1780 default (and expected) setting.
1782 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1783 message, again when deliver_drop_privilege is set or Exim is run as root.
1784 The bug described in (1) was not present in this case, but the tidying up
1785 of the other signals was missing. I have made the two cases consistent.
1787 3. The ignore_target_hosts setting on a manualroute router was being ignored
1788 for hosts that were looked up using the /MX notation.
1790 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1793 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1794 operated on the sender address. After changing the $sender_address to <>
1795 for the sender address verify, Exim was re-instated it as the original
1796 (before rewriting) address, but remembering that it had rewritten it, so it
1797 wasn't rewriting it again. This bug also had the effect of breaking the
1798 sender address verification caching when the sender address was rewritten.
1800 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1801 This has been changed so that if the ip literal address matches
1802 ignore_target_hosts, the router declines.
1804 7. Added expansion conditions match_domain, match_address, and match_local_
1805 part (NOT match_host).
1807 8. The placeholder for the Received: header didn't have a length field set.
1809 9. Added code to Exim itself and to exim_lock to test for a specific race
1810 condition that could lead to file corruption when using MBX delivery. The
1811 issue is with the lockfile that is created in /tmp. If this file is removed
1812 after a process has opened it but before that process has acquired a lock,
1813 there is the potential for a second process to recreate the file and also
1814 acquire a lock. This could lead to two Exim processes writing to the file
1815 at the same time. The added code performs the same test as UW imapd; it
1816 checks after acquiring the lock that its file descriptor still refers to
1817 the same named file.
1819 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1820 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1823 11. Added the smtp_active_hostname option. If used, this will typically be made
1824 to depend on the incoming interface address. Because $interface_address is
1825 not set up until the daemon has forked a reception process, error responses
1826 that can happen earlier (such as "too many connections") no longer contain
1829 12. If an expansion in a condition on a "warn" statement fails because a lookup
1830 defers, the "warn" statement is abandoned, and the next ACL statement is
1831 processed. Previously this caused the whole ACL to be aborted.
1833 13. Added the iplsearch lookup type.
1835 14. Added ident_timeout as a log selector.
1837 15. Added tls_certificate_verified as a log selector.
1839 16. Added a global option tls_require_ciphers (compare the smtp transport
1840 option of the same name). This controls incoming TLS connections.
1842 17. I finally figured out how to make tls_require_ciphers do a similar thing
1843 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1844 before starting the TLS session.
1846 18. Tabs are now shown as \t in -bP output.
1848 19. If the log selector return_path_on_delivery was set, Exim crashed when
1849 bouncing a message because it had too many Received: header lines.
1851 20. If two routers both had headers_remove settings, and the first one included
1852 a superfluous trailing colon, the final name in the first list and the
1853 first name in the second list were incorrectly joined into one item (with a
1854 colon in the middle).
1860 1. Added -C and -D options to the exinext utility, mainly to make it easier
1861 to include in the automated testing, but these could be helpful when
1862 multiple configurations are in use.
1864 2. The exinext utility was not formatting the output nicely when there was
1865 an alternate port involved in the retry record key, nor when there was a
1866 message id as well (for retries that were specific to a specific message
1867 and a specific host). It was also confused by IPv6 addresses, because of
1868 the additional colons they contain. I have fixed the IPv4 problem, and
1869 patched it up to do a reasonable job for IPv6.
1871 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1872 delivery, the log line now contains "pipelined" if PIPELINING was used.
1874 4. An SMTP transport process used to panic and die if the bind() call to set
1875 an explicit outgoing interface failed. This has been changed; it is now
1876 treated in the same way as a connect() failure.
1878 5. A reference to $sender_host_name in the part of a conditional expansion
1879 that was being skipped was still causing a DNS lookup. This no longer
1882 6. The def: expansion condition was not recognizing references to header lines
1883 that used bh_ and bheader_.
1885 7. Added the _cache feature to named lists.
1887 8. The code for checking quota_filecount in the appendfile transport was
1888 allowing one more file than it should have been.
1890 9. For compatibility with Sendmail, the command line option
1898 and sets the incoming protocol and host name (for trusted callers). The
1899 host name and its colon can be omitted when only the protocol is to be set.
1900 Note the Exim already has two private options, -pd and -ps, that refer to
1901 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1902 "s", but I don't think that's a major issue.
1904 10. A number of refactoring changes to the code, none of which should affect
1907 (a) The number of logging options was getting close to filling up the
1908 32-bit word that was used as a bit map. I have split them into two classes:
1909 those that are passed in the argument to log_write(), and those that are
1910 only ever tested independently outside of that function. These are now in
1911 separate 32-bit words, so there is plenty of room for expansion again.
1912 There is no change in the user interface or the logging behaviour.
1914 (b) When building, for example, log lines, the code previously used a
1915 macro that called string_cat() twice, in order to add two strings. This is
1916 not really sufficiently general. Furthermore, there was one instance where
1917 it was actually wrong because one of the argument was used twice, and in
1918 one call a function was used. (As it happened, calling the function twice
1919 did not affect the overall behaviour.) The macro has been replaced by a
1920 function that can join an arbitrary number of extra strings onto a growing
1923 (c) The code for expansion conditions now uses a table and a binary chop
1924 instead of a serial search (which was left over from when there were very
1925 few conditions). Also, it now recognizes conditions like "pam" even when
1926 the relevant support is not compiled in: a suitably worded error message is
1927 given if an attempt is made to use such a condition.
1929 11. Added ${time_interval:xxxxx}.
1931 12. A bug was causing one of the ddress fields not to be passed back correctly
1932 from remote delivery subprocesses. The field in question was not being
1933 subsequently used, so this caused to problems in practice.
1935 13. Added new log selectors queue_time and deliver_time.
1937 14. Might have fixed a bug in maildirsizefile handling that threw up
1938 "unexpected character" debug warnings, and recalculated the data
1939 unnecessarily. In any case, I expanded the warning message to give more
1942 15. Added the message "Restricted characters in address" to the statements in
1943 the default ACL that block characters like @ and % in local parts.
1945 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1946 Three changes have been made:
1948 (a) There was a serious bug; a negative response to MAIL caused the whole
1949 recipient domain to be cached as invalid, thereby blocking all messages
1950 to all local parts at the same domain, from all senders. This bug has
1951 been fixed. The domain is no longer cached after a negative response to
1952 MAIL if the sender used is not empty.
1954 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1957 (c) A new callout option, "use_sender" has been added for people who want
1958 the modified behaviour.
1964 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1967 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1968 indeed breaks things for older releases.
1970 3. Added additional logging to the case where there is a problem reading data
1971 from a filter that is running in a subprocess using a pipe, in order to
1972 try to track down a specific problem.
1974 4. Testing facility fudge: when running in the test harness and attempting
1975 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1976 getting "No route to host". Convert this to a timeout.
1978 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1981 6. Some OS don't have socklen_t but use size_t instead. This affects the
1982 fifth argument of getsockopt() amongst other things. This is now
1983 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1984 can be set for individual OS. I have set it for SunOS5, OSF1, and
1985 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1986 some earlier ones do not.
1988 7. Change 4.30/15 was not doing the test caselessly.
1990 8. The standard form for an IPv6 address literal was being rejected by address
1991 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1992 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1993 this, as well as the form without the "IPv6" on the front (but only when
1994 address literals are enabled, of course).
1996 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1998 10. Exim crashed if a message with an empty sender address specified by -f
1999 encountered a router with an errors_to setting. This could be provoked only
2000 by a command such as
2004 where an empty string was supplied; "<>" did not hit this bug.
2006 11. Installed PCRE release 4.5.
2008 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2009 remained set. It is now erased.
2011 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2012 times from message ids (which are base 36 rather than the normal 62).
2014 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2015 were being counted as actual protocol errors, and logged if the log
2016 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2017 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2018 and DATA following a set of rejected RCPTs do not count as protocol errors.
2019 In other words, Exim assumes they were pipelined, though this may not
2020 actually be the case. Of course, in all cases the client gets an
2021 appropriate error code.
2023 15. If a lookup fails in an ACL condition, a message about the failure may
2024 be available; it is used if testing the ACL cannot continue, because most
2025 such messages specify what the cause of the deferral is. However, some
2026 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2027 that caused an old message to be retained and used if a later statement
2028 caused a defer, replacing the real cause of the deferral.
2030 16. If an IP address had so many PTR records that the DNS lookup buffer
2031 was not large enough to hold them, Exim could crash while trying to process
2032 the truncated data. It now detects and logs this case.
2034 17. Further to 4.21/58, another change has been made: if (and only if) the
2035 first line of a message (the first header line) ends with CRLF, a bare LF
2036 in a subsequent header line has a space inserted after it, so as not to
2037 terminate the header.
2039 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2040 unnecessarily, used atoi() instead of strtol(), and didn't check the
2041 termination when getting file sizes from file names by regex.
2043 19. Completely re-implemented the support for maildirsize files, in the light
2044 of a number of problems with the previous contributed implementation
2045 (4.30/29). In particular:
2047 . If the quota is zero, the maildirsize file is maintained, but no quota is
2050 . If the maildir directory does not exist, it is created before any attempt
2051 to write a maildirsize file.
2053 . The quota value in the file is just a cache; if the quota is changed in
2054 the transport, the new value overrides.
2056 . A regular expression is available for excluding directories from the
2059 20. The autoreply transport checks the characters in options that define the
2060 message's headers; it allows continued headers, but it was checking with
2061 isspace() after an embedded newline instead of explicitly looking for a
2064 21. If all the "regular" hosts to which an address was routed had passed their
2065 expiry times, and had not reached their retry times, the address was
2066 bounced, even if fallback hosts were defined. Now Exim should go on to try
2069 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2070 equivalent code in the SMTP transport. Some hosts send humungous responses
2071 to HELO/EHLO, more than 1024 it seems.
2073 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2074 gives compiler warnings in some environments. I've now done it "properly",
2077 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2078 (because of problems with the built-in one) was declared to return uschar *
2079 instead of char *, causing compiler failure.
2081 25. Fixed a file descriptor leak when processing alias/forward files.
2083 26. Fixed a minor format string issue in dbfn.c.
2085 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2087 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2088 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2091 29. When a sender address is verified, it is cached, to save repeating the test
2092 when there is more than one recipient in a message. However, when the
2093 verification involves a callout, it is possible for different callout
2094 options to be set for different recipients. It is too complicated to keep
2095 track of this in the cache, so now Exim always runs a verification when a
2096 callout is required, relying on the callout cache for the optimization.
2097 The overhead is duplication of the address routing, but this should not be
2100 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2101 to be verified with callout=postmaster, and the main callout worked but the
2102 postmaster check failed, the verification correctly failed. However, if a
2103 subsequent RCPT command asked for sender verification *without* the
2104 postmaster check, incorrect caching caused this verification also to fail,
2107 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2108 it was not caching the DNS options (qualify_single, search_parents) that
2109 were used when the lookup failed. A subsequent lookup with different
2110 options therefore always gave the same answer, though there were cases
2111 where it should not have. (Example: a "domains = !$mx_any" option on a
2112 dnslookup router: the "domains" option is always processed without any
2113 widening, but the router might have qualify_single set.) Now Exim uses the
2114 cached value only when the same options are set.
2116 32. Added John Jetmore's "exipick" utility to the distribution.
2118 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2119 than a timeout (e.g. a certificate is required, and is not provided), an
2120 Exim server now closes the connection immediately. Previously it waited for
2121 the client to close - but if the client is SSL, it seems that they each
2122 wait for each other, leading to a delay before one of them times out.
2124 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2125 maintained 0.8.x compatibility because I don't think many are using it, and
2126 it is clearly obsolete.
2128 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2131 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2132 client certificate was expired. A simple patch fixes this, though I don't
2133 understand the full logic of why the verify callback is called multiple
2136 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2137 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2138 which causes problems with some clients (such as the Certicom SSL Plus
2139 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2140 disables the coutermeasure allowing Eudora to connect."
2142 38. Exim was not checking that a write() to a log file succeeded. This could
2143 lead to Bad Things if a log got too big, in particular if it hit a file
2144 size limit. Exim now panics and dies if it cannot write to a log file, just
2145 as it does if it cannot open a log file.
2147 39. Modified OS/Makefile-Linux so that it now contains
2149 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2151 The two -D definitions ensure that Exim is compiled with large file
2152 support, which makes it possible to handle log files that are bigger than
2155 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2156 instance) a domain was checked against a named list that involved a lookup,
2157 causing $domain_data to be set, then another domain was checked against the
2158 same list, then the first domain was re-checked, the value of $domain_data
2159 after the final check could be wrong. In particular, if the second check
2160 failed, it could be set empty. This bug probably also applied to
2163 41. The strip_trailing_dot option was not being applied to the address given
2164 with the -f command-line option.
2166 42. The code for reading a message's header from the spool was incrementing
2167 $received_count, but never initializing it. This meant that the value was
2168 incorrect (doubled) while delivering a message in the same process in which
2169 it was received. In the most common configuration of Exim, this never
2170 happens - a fresh exec is done - but it can happen when
2171 deliver_drop_privilege is set.
2173 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2174 it now includes up to 150 characters of the unexpected data in the log
2177 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2178 and building data strings. The size of both of these buffers was 10 000
2179 bytes - far larger than anybody would *ever* want, thought I. Needless to
2180 say, somebody hit the limit. I have increased the maximum line length to
2181 20 000 and the maximum data length of concatenated lines to 100 000. I have
2182 also fixed two bugs, because there was no checking on these buffers. Tsk,
2183 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2184 buffer is too small.
2186 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2187 lsearch lookups. Now it does.
2189 46. When parsing a route_list item in a manualroute router, a fixed-length
2190 buffer was used for the list of hosts. I made this 1024 bytes long,
2191 thinking that nobody would ever have a list of hosts that long. Wrong.
2192 Somebody had a whole pile of complicated expansion conditions, and the
2193 string was silently truncated, leading to an expansion error. It turns out
2194 that it is easier to change to an unlimited length (owing to other changes
2195 that have happened since this code was originally written) than to build
2196 structure for giving a limitation error. The length of the item that
2197 expands into the list of hosts is now unlimited.
2199 47. The lsearch lookup could not handle data where the length of text line was
2200 more than 4095 characters. Such lines were truncated, leading to shortened
2201 data being returned. It should now handle lines of any length.
2203 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2204 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2207 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2208 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2209 no longer happens in comment lines. A list of replacements is now placed
2210 at the head of all of the source files, except those whose only change is
2211 to replace PERL_COMMAND in the very first #! line.
2213 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2214 messages on the queue, with a bottom-up merge sort, using code contributed
2215 by Michael Haardt. This should make operations like -bp somewhat faster on
2216 large queues. It won't affect queue runners, except when queue_run_in_order
2219 51. Installed eximstats 1.31 in the distribution.
2221 52. Added support for SRV lookups to the dnslookup router.
2223 53. If an ACL referred to $message_body or $message_body_end, the value was not
2224 reset for any messages that followed in the same SMTP session.
2226 54. The store-handling optimization for building very long strings was not
2227 differentiating between the different store pools. I don't think this
2228 actually made any difference in practice, but I've tidied it.
2230 55. While running the routers to verify a sender address, $sender_address
2231 was still set to the sender address. This is wrong, because when routing to
2232 send a bounce to the sender, it would be empty. Therefore, I have changed
2233 it so that, while verifying a sender address, $sender_address is set to <>.
2234 (There is no change to what happens when verifying a recipient address.)
2236 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2237 target A or AAAA records (if not already returned) without resetting the
2238 qualify_single or search_parents options of the DNS resolver. These are
2239 inappropriate in this case because the targets of MX and SRV records must
2240 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2241 target that, when qualified, matched something in the local domain. These
2242 two options are now turned off when doing these lookups.
2244 57. It seems that at least some releases of Reiserfs (which does not have the
2245 concept of a fixed number of inodes) returns zero and not -1 for the
2246 number of available inodes. This interacted badly with check_spool_inodes,
2247 which assumed that -1 was the "no such thing" setting. What I have done is
2248 to check that the total number of inodes is greater than zero before doing
2249 the test of how many are available.
2251 58. When a "warn" ACL statement has a log_message modifier, the message is
2252 remembered, and not repeated. This is to avoid a lot of repetition when a
2253 message has many recipients that cause the same warning to be written.
2254 Howewer, Exim was preserving the list of already written lines for an
2255 entire SMTP session, which doesn't seem right. The memory is now reset if a
2256 new message is started.
2258 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2259 debugging output unless log_rewrite was also set.
2261 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2262 of (int)(handle) when we know that handle contains (void *)(-1).
2264 61. The Exim daemon panic-logs an error return when it closes the incoming
2265 connection. However "connection reset by peer" seems to be common, and
2266 isn't really an error worthy of noting specially, so that particular error
2269 62. When Exim is trying to find all the local interfaces, it used to panic and
2270 die if the ioctl to get the interface flags failed. However, it seems that
2271 on at least one OS (Solaris 9) it is possible to have an interface that is
2272 included in the list of interfaces, but for which you get a failure error
2273 for this call. This happens when the interface is not "plumbed" into a
2274 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2275 failure of the "get flags" call assumes that the interface is down.
2277 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2278 makes life easier for people who are doing arithmetic on fields extracted
2279 from dates, where you often get leading zeros that should not be
2280 interpreted as octal.
2282 64. Added qualify_domain to the redirect router, to override the global
2285 65. If a pathologically long header line contained very many addresses (the
2286 report of this problem mentioned 10 000) and each of them was rewritten,
2287 Exim could use up a very large amount of memory. (It kept on making new
2288 copies of the header line as it rewrote, and never released the old ones.)
2289 At the expense of a bit more processing, the header rewriting function has
2290 been changed so that it no longer eats memory in this way.
2292 66. The generation of the Received: header has been moved from the time that a
2293 message starts to be received, to the time that it finishes. The timestamp
2294 in the Received: header should now be very close to that of the <= log
2295 line. There are two side-effects of this change:
2297 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2298 logged header lines no longer include the local Received: line, because
2299 it has not yet been created. The same applies to a copy of the message
2300 that is returned to a non-SMTP sender when a message is rejected.
2302 (b) When a filter file is tested using -bf, no additional Received: header
2303 is added to the test message. After some thought, I decided that this
2306 This change does not affect the value of $received_for. It is still set
2307 after address rewriting, but before local_scan() is called.
2309 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2311 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2312 gave an unhelpful panic error message, and a defer error. I have managed to
2313 change this behaviour so that it now rejects any supplied certificate,
2314 which seems right, as the list of acceptable certificates is empty.
2316 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2317 gave an unhelpful defer error. I have not managed to make this reject any
2318 supplied certificates, but the error message it gives is "no certificate
2319 supplied", which is not helpful.
2321 70. exigrep's output now also includes lines that are not associated with any
2322 message, but which match the given pattern. Implemented by a patch from
2323 Martin Sluka, which also tidied up the Perl a bit.
2325 71. Recipient callout verification, like sender verification, was using <> in
2326 the MAIL FROM command. This isn't really the right thing, since the actual
2327 sender may affect whether the remote host accepts the recipient or not. I
2328 have changed it to use the actual sender in the callout; this means that
2329 the cache record is now keyed on a recipient/sender pair, not just the
2330 recipient address. There doesn't seem to be a real danger of callout loops,
2331 since a callout by the remote host to check the sender would use <>.
2332 [SEE ABOVE: changed after hitting problems.]
2334 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2335 temporary errors. However, in the case of such a code being given after
2336 the end of a data transmission (i.e. after ".") Exim was failing to write
2337 a retry record for the message. (Yes, there was some broken host that was
2338 actually sending 8xx at this point.)
2340 73. An unknown lookup type in a host list could cause Exim to panic-die when
2341 the list was checked. (An example that provoked this was putting <; in the
2342 middle of a list instead of at the start.) If this happened during a DATA
2343 ACL check, a -D file could be left lying around. This kind of configuration
2344 error no longer causes Exim to die; instead it causes a defer errror. The
2345 incident is still logged to the main and panic logs.
2347 74. Buglet left over from Exim 3 conversion. The message "too many messages
2348 in one connection" was written to the rejectlog but not the mainlog, except
2349 when address rewriting (yes!) was being logged.
2351 75. Added write_rejectlog option.
2353 76. When a system filter was run not as root (that is, when system_filter_user
2354 was set), the values of the $n variables were not being returned to the
2355 main process; thus, they were not subsequently available in the $sn
2358 77. Added +return_path_on_delivery log selector.
2360 78. A connection timeout was being treated differently from recipients deferred
2361 when testing hosts_max_try with a message that was older than the host's
2362 retry timeout. (The host should not be counted, thus allowing all hosts to
2363 be tried at least once before bouncing.) This may have been the cause of an
2364 occasionally reported bug whereby a message would remain on the queue
2365 longer than the retry timeout, but would be bounced if a delivery was
2366 forced. I say "may" because I never totally pinned down the problem;
2367 setting up timeout/retry tests is difficult. See also the next item.
2369 79. The ultimate address timeout was not being applied to errors that involved
2370 a combination of host plus message (for example, a timeout on a MAIL
2371 command). When an address resolved to a number of possible hosts, and they
2372 were not all tried for each delivery (e.g. because of hosts_max_try), a
2373 message could remain on the queue longer than the retry timeout.
2375 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2378 81. Fixed an obscure SMTP outgoing bug which required at least the following
2379 conditions: (a) there was another message waiting for the same server;
2380 (b) the server returned 5xx to all RCPT commands in the first message so
2381 that the message was not completed; (c) the server dropped the connection
2382 or gave a negative response to the RSET that Exim sends to abort the
2383 transaction. The observed case was a dropped connection after DATA that had
2384 been sent in pipelining mode. That is, the server had advertised PIPELINING
2385 but was not implementing it correctly. The effect of the bug was incorrect
2386 behaviour, such as trying another host, and this could lead to a crash.
2392 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2393 and daemon.c were passed as pointers to ints; they should have been
2394 pointers to socklen_t variables (which are typically unsigned ints).
2396 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2399 3. Fixed a really odd bug that affected only the testing scheme; patching a
2400 certain fixed string in the binary changed the value of another string that
2401 happened to be identical to the end of the original first string.
2403 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2404 name", it returns that address as the IP address. On some operating
2405 systems (e.g. Solaris), it also passes back the IP address string as the
2406 "host name". However, on others (e.g. Linux), it passes back an empty
2407 string. Exim wasn't checking for this, and was changing the host name to an
2408 empty string, assuming it had been canonicized.
2410 5. Although rare, it is permitted to have more than one PTR record for a given
2411 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2412 all the names associated with an address, because they do in Solaris.
2413 However, it seems that they do not in Linux for data that comes from the
2414 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2415 I found this out when I moved to a new Linux workstation and tried to run
2416 the Exim test suite.
2418 To get round this problem I have changed the code so that it now does its
2419 own call to the DNS to look up PTR records when searching for a host name.
2420 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2421 addresses that are only in /etc/hosts are still found.
2423 This behaviour is, however, controlled by an option called host_lookup_
2424 order, which defaults to "bydns:byaddr". If people want to use the other
2425 order, or indeed, just use one or the other means of lookup, they can
2426 specify it in this variable.
2428 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2429 some operating systems, this comes back from gethostbyaddr() as an empty
2430 string, and this is what Exim used to test for. However, it seems that in
2431 other systems, "." is yielded. Exim now tests for this case too.
2433 7. The values of check_spool_space and check_log_space are now held internally
2434 as a number of kilobytes instead of an absolute number of bytes. If a
2435 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2436 kilobyte. This means that much larger values can be stored.
2438 8. Exim monitor: an attempt to get the action menu when not actually pointing
2439 at a message produces an empty menu entitled "No message selected". This
2440 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2441 no entries in it ("Shell widget menu has zero width and/or height"). So I
2442 have added a single, blank menu entry in this case.
2444 9. Added ${quote_local_part.
2446 10. MIME decoding is now applied to the contents of Subject: header lines when
2449 11. Now that a reference to $sender_host_address automatically causes a reverse
2450 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2451 host lookup before query-style lookups in lists that might use this
2452 variable. This has therefore been abolished, and the "net-" prefix is no
2453 longer necessary for query-style lookups.
2455 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2456 have been a typo for LFLAGS, so it has been changed.
2458 13. The install script calls Exim with "-C /dev/null" in order to find the
2459 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2460 to be output. Howeve, since Exim outputs its version number before the
2461 error, it didn't break the script. It just looked ugly. I fixed this by
2462 always allowing "-C /dev/null" if the caller is root.
2464 14. Ignore overlarge ACL variable number when reading spool file - insurance
2465 against a later release with more variables having written the file.
2467 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2468 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2469 this, as well as the form without the "IPv6" on the front.
2471 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2472 OS/Makefile-Darwin file.
2474 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2475 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2477 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2478 variable. However, some OS use ints for inodes. I've added cast to long int
2479 to get rid of the compiler warning.
2481 19. I had forgotten to lock out "/../" in configuration file names when
2482 ALT_CONFIG_PREFIX was set.
2484 20. Routers used for verification do not need to specify transports. However,
2485 if such a router generated a host list, and callout was configured, Exim
2486 crashed, because it could not find a port number from the (non-existent)
2487 transport. It now assumes port 25 in this circumstance.
2489 21. Added the -t option to exigrep.
2491 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2492 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2493 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2494 them is set, LOOKUP_LSEARCH is forced.
2496 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2498 24. Added sender and host information to the "rejected by local_scan()" log
2499 line; previously there was no indication of these.
2501 25. Added .include_if_exists.
2503 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2504 Linux. It turns out that not all file systems support this. Apparently some
2505 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2506 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2507 not supported on the file descriptor, is now ignored when Exim is trying to
2508 sync a directory. This applies only to Linux.
2510 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2512 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2513 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2516 29. Added support for maildirsize files from supplied patch (modified a bit).
2518 30. The use of :fail: followed by an empty string could lead Exim to respond to
2519 sender verification failures with (e.g.):
2521 550 Verification failed for <xxx>
2522 550 Sender verify failed
2524 where the first response line was missing the '-' that indicates it is not
2525 the final line of the response.
2527 31. The loop for finding the name of the user that called Exim had a hardwired
2528 limit of 10; it now uses the value of finduser_retries, which is used for
2529 all other user lookups.
2531 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2534 33. Exim was neglecting to zero errno before one call of strtol() when
2535 expanding a string and expecting an integer value. On some systems this
2536 resulted in spurious "integer overflow" errors. Also, it was casting the
2537 result into an int without checking.
2539 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2540 did not work. The code looks as if it has *never* worked, though it appears
2541 to have been documented since at least releast 1.62. I have made it work.
2543 35. The "timeout_DNS" error in retry rules, also documented since at least
2544 1.62, also never worked. As it isn't clear exactly what this means, and
2545 clearly it isn't a major issue, I have abolished the feature by treating it
2546 as "timeout", and writing a warning to the main and panic logs.
2548 36. The display of retry rules for -brt wasn't always showing the error code
2551 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2552 timeout_connect_A, timeout_connect_MX.
2554 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2555 to the empty sender.
2557 39. The daemon was not analysing the content of -oX till after it had closed
2558 stderr and disconnected from the controlling terminal. This meant that any
2559 syntax errors were only noted on the panic log, and the return code from
2560 the command was 0. By re-arranging the code a little, I've made the
2561 decoding happen first, so such errors now appear on stderr, and the return
2562 code is 1. However, the actual setting up of the sockets still happens in
2563 the disconnected process, so errors there are still only recorded on the
2566 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2567 connections (as happens on some IP stacks) was logged at start up time as
2568 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2569 it from "IPv6 and IPv4", which means that two separate sockets are being
2572 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2573 says whether AF_INET or AF_INET6 was passed as an argument.
2575 42. Exiwhat output was messed up when time zones were included in log
2578 43. Exiwhat now gives more information about the daemon's listening ports,
2579 and whether -tls-on-connect was used.
2581 44. The "port" option of the smtp transport is now expanded.
2583 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2584 silently ignored. Now an error message is written to the main and panic
2587 46. There's a new ACL modifier called "logwrite" which writes to a log file
2588 as soon as it is encountered.
2590 47. Added $local_user_uid and $local_user_gid at routing time.
2592 48. Exim crashed when trying to verify a sender address that was being
2595 49. Exim was recognizing only a space character after ".include". It now also
2596 recognizes a tab character.
2598 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2599 extracting the relevant information from the specification. The man page no
2600 longer contains scrambled data for the -d option, and I've added a section
2601 at the front about calling Exim under different names.
2603 51. Added "extra_headers" argument to the "mail" command in filter files.
2605 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2608 53. Installed eximstats 1.29.
2610 54. Added transport_filter_timeout as a generic transport option.
2612 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2613 Cc: header lines. This was required by RFC 822, but it not required by RFC
2616 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2617 incoming messages that did not have them. Now it does so only if the
2618 message originates locally, that is, if there is no associated remote host
2619 address. When Resent- header lines are present, this applies to the Resent-
2620 lines rather than the non-Resent- lines.
2622 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2623 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2625 58. Messages for configuration errors now include the name of the main
2626 configuration file - useful now that there may be more than one file in a
2627 list (.included file names were always shown).
2629 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2630 for those rare installations that do not start the daemon as root or run it
2631 setuid root. I've cut out the call to initgroups() if the daemon is not
2634 60. The Exim user and group can now be bound into the binary as text strings
2635 that are looked up at the start of Exim's processing.
2637 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2639 62. Added $mailstore_basename variable.
2641 63. Installed patch to sieve.c from Michael Haardt.
2643 64. When Exim failed to open the panic log after failing to open the main log,
2644 the original message it was trying to log was written to stderr and debug
2645 output, but if they were not available (the usual case in production), it
2646 was lost. Now it is written to syslog before the two lines that record the
2647 failures to open the logs.
2649 65. Users' Exim filters run in subprocesses under the user's uid. It is
2650 possible for a "deliver" command or an alias in a "personal" command to
2651 provoke an address rewrite. If logging of address rewriting is configured,
2652 this fails because the process is not running as root or exim. There may be
2653 a better way of dealing with this, but for the moment (because 4.30 needs
2654 to be released), I have disabled address rewrite logging when running a
2655 filter in a non-root, non-exim process.
2661 1. The buildconfig auxiliary program wasn't quoting the value set for
2662 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2663 not defined. This bug was masked in 4.22 by the effect that was fixed in
2666 2. Some messages that were rejected after a message id was allocated were
2667 shown as "incomplete" by exigrep. It no longer does this for messages that
2668 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2670 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2671 have allow_domain_literals set, the ID did not get logged in the <= line.
2672 Domain literals are now always recognized in Message-ID: header lines.
2674 4. The first argument for a ${extract expansion item is the key name or field
2675 number. Leading and trailing spaces in this item were not being ignored,
2676 causing some misleading effects.
2678 5. When deliver_drop_privilege was set, single queue runner processes started
2679 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2680 same command in the process it spins off) were not dropping privilege.
2682 6. When the daemon running as "exim" started a queue runner, it always
2683 re-executed Exim in the spun-off process. This is a waste of effort when
2684 deliver_drop_privilege is set. The new process now just calls the
2685 queue-runner function directly.
2691 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2694 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2695 ignored. Though the use of -oP was forcing the writing of a pid file, it
2696 was always written to the default place.
2698 3. If the message "no IP address found for host xxxx" is generated during
2699 incoming verification, it is now followed by identification of the incoming
2700 connection (so you can more easily find what provoked it).
2702 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2704 5. Added some features to "harden" Exim a bit more against certain attacks:
2706 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2707 be put in Local/Makefile. This is like the never_users runtime option,
2708 but it cannot be overridden. The default setting is "root".
2710 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2711 prefix string with which any file named in a -C command line option
2714 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2715 is retained for -C and -D only if the caller of Exim is root. Without
2716 it, the exim user may also use -C and -D and retain privilege.
2718 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2719 command line option is disabled.
2721 6. Macro names set by the -D option must start with an upper case letter, just
2722 like macro names defined in the configuration file.
2724 7. Added "dereference=" facility to LDAP.
2726 8. Two instances of the typo "uknown" in the source files are fixed.
2728 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2729 the Configure-Makefile script screwed up while processing it.
2731 10. Incorporated PCRE 4.4.
2733 11. The SMTP synchronization check was not operating right at the start of an
2734 SMTP session. For example, it could not catch a HELO sent before the client
2735 waited for the greeting. There is now a check for outstanding input at the
2736 point when the greeting is written. Because of the duplex, asynchronous
2737 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2738 way, but not yet received, when the check is performed.
2740 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2741 on TCP/IP sockets, because this apparently causes some broken clients to
2744 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2745 unchanged) from the Cygwin maintainer.
2747 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2748 was supported instead of testing for mbx. Effectively a typo.
2750 15. The spa authenticator server code was not checking that the input it
2751 received was valid base64.
2753 16. The debug output line for the "set" modifier in ACLs was not showing the
2754 name of the variable that was being set.
2756 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2758 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2759 lookup on its own. Something else was needed to trigger the lookup. For
2760 example, a match in host_lookup or the need for a host name in a host list.
2761 Now, if $sender_host_name is referenced and the host name has not yet been
2762 looked up, a lookup is performed. If the lookup fails, the variable remains
2763 empty, and $host_lookup_failed is set to "1".
2765 19. Added "eqi" as a case-independent comparison operator.
2767 20. The saslauthd authentication condition could segfault if neither service
2768 nor realm was specified.
2770 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2771 error message that was logged was misleading, and incoming SMTP
2772 connections were dropped. The message is now more accurate, and temporary
2773 errors are given to SMTP connections.
2775 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2776 (except RSET) with a 421 error, until QUIT is received. However, it was
2777 failing to send a response to QUIT.
2779 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2780 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2781 ACL is now run after the helo_try_verify_hosts code.
2783 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2784 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2785 case-independent, so other case variants are also recognized). Apparently
2786 some systems use these upper case variants.
2788 25. If more than two messages were waiting for the same host, and a transport
2789 filter was specified for the transport, Exim sent two messages over the
2790 same TCP/IP connection, and then failed with "socket operation on non-
2791 socket" when it tried to send the third.
2793 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2795 27. The extern definition of crypt16() in expand.c was not being excluded when
2796 the OS had its own crypt16() function.
2798 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2799 as a preferred synonym for return_size_limit, both as an option and as an
2802 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2804 30. Changed the default configuration ACL to relax the local part checking rule
2805 for addresses that are not in any local domains. For these addresses,
2806 slashes and pipe symbols are allowed within local parts, but the sequence
2807 /../ is explicitly forbidden.
2809 31. SPA server authentication was not clearing the challenge buffer before
2812 32. log_message in a "warn" ACL statement was writing to the reject log as
2813 well as to the main log, which contradicts the documentation and doesn't
2814 seem right (because no rejection is happening). So I have stopped it.
2816 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2817 However, I am unable to do any testing of this.
2819 34. Fixed an infelicity in the appendfile transport. When checking directories
2820 for a mailbox, to see if any needed to be created, it was accidentally
2821 using path names with one or more superfluous leading slashes; tracing
2822 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2824 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2825 contents are added to the log line that is written for every discarded
2826 recipient. (Previously a log_message setting was ignored.)
2828 36. The ${quote: operator now quotes the string if it is empty.
2830 37. The install script runs exim in order to find its version number. If for
2831 some reason other than non-existence or emptiness, which it checks, it
2832 could not run './exim', it was installing it with an empty version number,
2833 i.e. as "exim-". This error state is now caught, and the installation is
2836 38. An argument was missing from the function that creates an error message
2837 when Exim fails to connect to the socket for saslauthd authentication.
2838 This could cause Exim to crash, or give a corrupted message.
2840 39. Added isip, isip4, and isip6 to ${if conditions.
2842 40. The ACL variables $acl_xx are now saved with the message, and can be
2843 accessed later in routers, transports, and filters.
2845 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2846 strings in the file are not string-expanded.
2848 42. If a MAIL command specified a SIZE value that was too large to fit into an
2849 int variable, the check against message_size_limit failed. Such values are
2850 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2851 day this will have to be increased, but I don't think I want to be around
2852 when emails are that large.
2859 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2860 iconv() is not standard in FreeBSD.
2862 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2863 IPv6 enabled. The observed symptom was a segmentation fault on return from
2864 the function os_common_find_running_interfaces() in src/os.c.
2866 3. In the check_special_case() function in daemon.c I had used "errno" as an
2867 argument name, which causes warnings on some systems. This was basically a
2868 typo, since it was named "eno" in the comments!
2870 4. The code that waits for the clock to tick (at a resolution of some fraction
2871 of a second) so as to ensure message-id uniqueness was always waiting for
2872 at least one whole tick, when it could have waited for less. [This is
2873 almost certainly not relevant at current processor speeds, where it is
2874 unlikely to ever wait at all. But we try to future-proof.]
2876 5. The function that sleeps for a time interval that includes fractions of a
2877 second contained a race. It did not block SIGALRM between setting the
2878 timer, and suspending (a couple of lines later). If the interval was short
2879 and the sigsuspend() was delayed until after it had expired, the suspension
2880 never ended. On busy systems this could lead to processes getting stuck for
2883 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2884 process, before it forks any delivery processes. The open lookup caching
2885 mechanism meant that the open file or database connection was passed into
2886 the delivery process. The problem was that delivery processes always tidy
2887 up cached lookup data. This could cause a problem for the next delivery
2888 process started by the queue runner, because the external queue runner
2889 process does not know about the closure. So the next delivery process
2890 still has data in the lookup cache. In the case of a file lookup, there was
2891 no problem because closing a file descriptor in a subprocess doesn't affect
2892 the parent. However, if the lookup was caching a connection to a database,
2893 the connection was closed, and the second delivery process was likely to
2894 see errors such as "PGSQL: query failed: server closed the connection
2895 unexpectedly". The problem has been fixed by closing all cached lookups
2896 in a queue runner before running a delivery process.
2898 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2899 seem to have the "const" qualifier which it has on other OS. I've
2902 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2903 *of the same type* (client or server) with the same public name that an
2904 error should be diagnosed.
2906 9. When Exim looked up a host name for an IP address, but failed to find the
2907 original IP address when looking up the host name (a safety check), it
2908 output the message "<ip address> does not match any IP for NULL", which was
2909 confusing, to say the least. The bug was that the host name should have
2910 appeared instead of "NULL".
2912 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2913 user that is built into the binary, and the -C or -D options is used, root
2914 privilege is dropped before the configuration file is read. In addition,
2915 logging is switched to stderr instead of the normal log files. If the
2916 configuration then re-defines the Exim user, the unprivileged environment
2917 is probably not what is expected, so Exim logs a panic warning message (but
2920 However, if deliver_drop_privilege is set, the unprivileged state may well
2921 be exactly what is intended, so the warning has been cut out in that case,
2922 and Exim is allowed to try to write to its normal log files.
2928 1. smtp_return_error_details was not giving details for temporary sender
2929 or receiver verification errors.
2931 2. Diagnose a configuration error if two authenticators have the same public
2934 3. Exim used not to create the message log file for a message until the first
2935 delivery attempt. This could be confusing when incoming messages were held
2936 for policy or load reasons. The message log file is now created at the time
2937 the message is received, and an initial "Received" line is written to it.
2939 4. The automatically generated man page for command line options had a minor
2940 bug that caused no ill effects; however, a more serious problem was that
2941 the procedure for building the man page automatically didn't always
2942 operate. Consequently, release 4.20 contains an out-of-date version. This
2943 shouldn't happen again.
2945 5. When building Exim with embedded Perl support, the script that builds the
2946 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2947 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2949 6. The freeze_tell option was not being used for messages that were frozen on
2950 arrival, either by an ACL or by local_scan().
2952 7. Added the smtp_incomplete_transaction log selector.
2954 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2955 was accepting AUTH without a new EHLO.
2957 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2958 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2961 10. Logging of TCP/IP connections (when configured) now happens in the main
2962 daemon process instead of the child process, so that the TCP/IP connection
2963 count is more accurate (but it can never be perfect).
2965 11. The use of "drop" in a nested ACL was not being handled correctly in the
2966 outer ACL. Now, if condition failure induced by the nested "drop" causes
2967 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2968 or "require"), the connection is dropped.
2970 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2971 that yield "discard" can now be used with an "accept" or a "discard" verb,
2972 but an error is generated for any others (because I can't see a useful way
2973 to define what should happen).
2975 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2976 are now processed in the same way as lines in the Exim configuration file.
2977 In particular, continuation lines are supported.
2979 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2981 15. Added -ti meaning -t -i.
2983 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2984 domains, and warn by logging if others are found.
2986 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2987 returned from GIFCONF when Exim is trying to find the list of interfaces on
2988 a host. The code in os.c has been modified to copy each ifreq to an aligned
2989 structure in all cases.
2991 Also, in some cases, the returned ifreq's were being copied to a 'struct
2992 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2993 means the last couple of bytes of an IPv6 address could be chopped if the
2994 ifreq contained only a normal sockaddr (14 bytes storage).
2996 18. Named domain lists were not supported in the hosts_treat_as_local option.
2997 An entry such as +xxxx was not recognized, and was treated as a literal
3000 19. Ensure that header lines added by a DATA ACL are included in the reject log
3001 if the ACL subsequently rejects the message.
3003 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3004 MD5 (which is deprecated).
3006 21. When testing a filter file using -bf, Exim was writing a message when it
3007 took the sender from a "From " line in the message, but it was not doing so
3008 when it took $return_path from a Return-Path: header line. It now does.
3010 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3011 with a valid header line field name (a series of printing characters
3012 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3014 23. Changed "disc" in the source to "disk" to conform to the documentation and
3015 the book and for uniformity.
3017 24. Ignore Sendmail's -Ooption=value command line item.
3019 25. When execve() failed while trying to run a command in a pipe transport,
3020 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3021 could be confused with a return value of 69 from the command itself. This
3022 has been changed to 127, the value the shell returns if it is asked to run
3023 a non-existent command. The wording for the related log line suggests a
3024 non-existent command as the problem.
3026 26. If received_header_text expands to an empty string, do not add a Received:
3027 header line to the message. (Well, it adds a token one on the spool, but
3028 marks it "old" so that it doesn't get used or transmitted.)
3030 27. Installed eximstats 1.28 (addition of -nt option).
3032 28. There was no check for failure on the call to getsockname() in the daemon
3033 code. This can fail if there is a shortage of resources on the system, with
3034 ENOMEM, for example. A temporary error is now given on failure.
3036 29. Contrary to the C standard, it seems that in some environments, the
3037 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3038 program. Exim now does this explicitly; it affects the formatting of
3039 timestamps using strftime().
3041 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3042 complaints. I've now initialized all the variables, to avoid this.
3044 32. Header lines added by a system filter were not being "seen" during
3045 transport-time rewrites.
3047 33. The info_callback() function passed to OpenSSL is set up with type void
3048 (*)(SSL *, int, int), as described somewhere. However, when calling the
3049 function (actually a macro) that sets it up, the type void(*)() is
3050 expected. I've put in a cast to prevent warnings from picky compilers.
3052 34. If a DNS black list lookup found a CNAME record, but there were no A
3053 records associated with the domain it pointed at, Exim crashed.
3055 35. If a DNS black list lookup returned more than one A record, Exim ignored
3056 all but the first. It now scans all returned addresses if a particular IP
3057 value is being sought. In this situation, the contents of the
3058 $dnslist_value variable are a list of all the addresses, separated by a
3061 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3062 to accept a host name and all its aliases if the forward lookup for any of
3063 them yielded the IP address of the incoming connection. Now it accepts only
3064 those names whose forward lookup yields the correct IP address. Any other
3065 names are discarded. This closes a loophole whereby a rogue DNS
3066 administrator could create reverse DNS records to break through a
3067 wildcarded host restriction in an ACL.
3069 37. If a user filter or a system filter that ran in a subprocess used any of
3070 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3071 the wrong values were passed to the pipe command ($thisaddress had the
3072 value of $0, $0 had the value of $1, etc). This bug was introduced by
3073 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3075 38. Improved the line breaking for long SMTP error messages from ACLs.
3076 Previously, if there was no break point between 40 and 75 characters, Exim
3077 left the rest of the message alone. Two changes have been made: (a) I've
3078 reduced the minimum length to 35 characters; (b) if it can't find a break
3079 point between 35 and 75 characters, it looks ahead and uses the first one
3080 that it finds. This may give the occasional overlong line, but at least the
3081 remaining text gets split now.
3083 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3084 file descriptors might be low, and that setting 1000 would always raise it.
3085 It turns out that in some environments, the limit is already over 1000 and
3086 that lowering it causes trouble. So now Exim takes care not to decrease it.
3088 40. When delivering a message, the value of $return_path is set to $sender_
3089 address at the start of routing (routers may change the value). By an
3090 oversight, this default was not being set up when an address was tested by
3091 -bt or -bv, which affected the outcome if any router or filter referred to
3094 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3095 log only when "message" or "log_message" are set. However, if one of the
3096 conditions was an address verification, or a call to a nested ACL, the
3097 messages generated by the underlying test were being passed through. This
3098 no longer happens. The underlying message is available in $acl_verify_
3099 message for both "message" and "log_message" expansions, so it can be
3100 passed through if needed.
3102 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3103 new expansion $bh_ to give the encoded byte string without charset
3104 translation. Translation happens only if iconv() is available; HAVE_ICONV
3105 indicates this at build time. HEADERS_CHARSET gives the charset to
3106 translate to; headers_charset can change it in the configuration, and
3107 "headers charset" can change it in an individual filter file.
3109 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3110 that creates RFC 2047 encoded "words" labels them as that charset instead
3111 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3112 expansion operator; (ii) when Exim creates a From: line for a local
3113 message; (iii) when a header line is rewritten to include a "phrase" part.
3115 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3116 buggy, causing it to skip the first lines of messages whose message ID
3117 ended in 'D'. This would not have bitten before Exim release 4.14, because
3118 message IDs were unlikely to end in 'D' before then. The effect was to have
3119 incorrect size information for certain domains.
3121 45. #include "config.h" was missing at the start of the crypt16.c module. This
3122 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3125 46. If there was a timeout during a "random" callout check, Exim treated it as
3126 a failure of the random address, and carried on sending RSET and the real
3127 address. If the delay was just some slowness somewhere, the response to the
3128 original RCPT would be taken as a response to RSET and so on, causing
3129 mayhem of various kinds.
3131 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3132 when I implemented it. It didn't allow for the fact that some option values
3133 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3134 the right test for positive values.
3136 48. Domain names in DNS records are case-independent. Exim always looks them up
3137 in lower case. Some resolvers return domain names in exactly the case they
3138 appear in the zone file, that is, they may contain uppercase letters. Not
3139 all resolvers do this - some return always lower case. Exim was treating a
3140 change of case by a resolver as a change of domain, similar to a widening
3141 of a domain abbreviation. This triggered its re-routing code and so it was
3142 trying to route what was effectively the same domain again. This normally
3143 caused routing to fail (because the router wouldn't handle the domain
3144 twice). Now Exim checks for this case specially, and just changes the
3145 casing of the domain that it ultimately uses when it transmits the message
3148 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3151 50. If a filter generated a file delivery with a non-absolute name (possible if
3152 no home directory exists for the router), the forbid_file option was not
3155 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3156 the existing equality matching.
3158 52. Exim was using ints instead of ino_t variables in some places where it was
3159 dealing with inode numbers.
3161 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3162 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3163 called TMPDIR, and if it finds it is different, it changes its value.
3165 54. The smtp_printf() function is now made available to local_scan() so
3166 additional output lines can be written before returning. There is also an
3167 smtp_fflush() function to enable the detection of a dropped connection.
3168 The variables smtp_input and smtp_batched_input are exported to
3171 55. Changed the default runtime configuration: the message "Unknown user"
3172 has been removed from the ACL, and instead placed on the localuser router,
3173 using the cannot_route_message feature. This means that any verification
3174 failures that generate their own messages won't get overridden. Similarly,
3175 the "Unrouteable address" message that was in the ACL for unverifiable
3176 relay addresses has also been removed.
3178 56. Added hosts_avoid_esmtp to the smtp transport.
3180 57. The exicyclog script was not checking for the esoteric option
3181 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3182 will work only if exicyclog is run under the appropriate euid.
3184 58. Following a discussion on the list, the rules by which Exim recognises line
3185 endings on incoming messages have been changed. The -dropcr and drop_cr
3186 options are now no-ops, retained only for backwards compatibility. The
3187 following line terminators are recognized: LF CRLF CR. However, special
3188 processing applies to CR:
3190 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3191 nor a local message in the state where . is a terminator.
3193 (ii) If a bare CR is encountered in a header line, an extra space is added
3194 after the line terminator so as not to end the header. The reasoning
3195 behind this is that bare CRs in header lines are most likely either
3196 to be mistakes, or people trying to play silly games.
3198 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3199 was being incorrectly given as 18 bytes larger than it should have been.
3200 This is a VOB (very old bug).
3202 60. This may never have affected anything current, but just in case it has:
3203 When the local host is found other than at the start of a list of hosts,
3204 the local host, those with the same MX, and any that follow, are discarded.
3205 When the list in question was part of a longer list of hosts, the following
3206 hosts (not currently being processed) were also being discarded. This no
3207 longer happens. I'm not sure if this situation could ever has previously
3210 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3213 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3214 header. This is something that is recommended in a new Internet Draft, and
3215 is something that is documented as being done by Sendmail. There are two
3216 possible values. For messages generated by the autoreply transport, Exim
3219 Auto-Submitted: auto-replied
3221 whereas for all other generated messages (e.g. bounces) it adds
3223 Auto-Submitted: auto-generated
3225 63. The "personal" condition in filters now includes a test for the
3226 Auto-Submitted: header. If it contains the string "auto-" the message it
3227 not considered personal.
3229 64. Added rcpt_include_affixes as a generic transport option.
3231 65. Added queue_only_override (default true).
3233 66. Added the syslog_duplication option.
3235 67. If what should have been the first header line of a message consisted of
3236 a space followed by a colon, Exim was mis-interpreting it as a header line.
3237 It isn't of course - it is syntactically invalid and should therefore be
3238 treated as the start of the message body. The misbehaviour could have
3239 caused a number of strange effects, including loss of data in subsequent
3240 header lines, and spool format errors.
3242 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3243 client host had authenticated. This control can now be exercised by an ACL
3244 for more flexibility.
3246 69. By default, callouts do not happen when testing with -bh. There is now a
3247 variant, -bhc, which does actually run the callout code, including
3248 consulting and updating the callout cache.
3250 70. Added support for saslauthd authentication, courtesy of Alexander
3253 71. If statvfs() failed on the spool or log directories while checking their
3254 size for availability, Exim confusingly gave the error "space shortage".
3255 Furthermore, in debugging mode it crashed with a floating point exception.
3256 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3257 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3258 serious problem, Exim now writes to the main and panic logs when this
3259 happens, with details of the failure. It then refuses to accept the
3260 incoming message, giving the message "spool directory problem" or "log
3261 directory problem" with a 421 code for SMTP messages.
3263 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3264 0, 1, and 2 exist, because some OS complain for execs without them (see
3265 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3266 descriptors. However, the code omitted to check that the open succeeded,
3267 causing mysterious errors if for some reason the permissions on /dev/null
3268 got screwed. Now Exim writes a message to the main and panic logs, and
3269 bombs out if it can't open /dev/null.
3271 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3272 interact so that it is all more flexible. It is supposed to remain
3273 backwards compatible. Also added extra_local_interfaces.
3275 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3276 to bomb out with an assertion failure - to the client this appears as a
3277 connection drop. This problem occurs in the part of the code that was taken
3278 from the Samba project. Fortunately, the assertion is in a very simple
3279 function, so I have fixed this by reproducing the function inline in the
3280 one place where it is called, and arranging for authentication to fail
3281 instead of killing the process with assert().
3283 75. The SPA client code was not working when the server requested OEM rather
3284 than Unicode encoding.
3286 76. Added code to make require_files with a specific uid setting more usable in
3287 the case where statting the file as root fails - usually a non-root-mounted
3288 NFS file system. When this happens and the failure is EACCES, Exim now
3289 forks a subprocess and does the per-uid checking as the relevant uid.
3291 77. Added process_log_path.
3293 78. If log_file_path was not explicitly set, a setting of check_log_space or
3294 check_log_inodes was ignored.
3296 79. If a space check for the spool or log partitions fails, the incident is now
3297 logged. Of course, in the latter case the data may get lost...
3299 80. Added the %p formatting code to string_format() so that it can be used to
3300 print addresses in debug_print(). Adjusted all the address printing in the
3301 debugging in store.c to use %p rather than %d.
3303 81. There was a concern that a line of code in smtp_in.c could overflow a
3304 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3305 initially expressed, the concern was not well-founded, because trailing
3306 spaces are removed early. However, if the trailing spaces were followed by
3307 a NULL, they did not get removed, so the overflow was possible. Two fixes
3310 (a) I re-wrote the offending code in a cleaner fashion.
3311 (b) If an incoming SMTP command contains a NULL character, it is rejected
3314 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3315 runs initgroups(), so that if the Exim user is in any additional groups,
3316 they will be used during message reception.
3322 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob