Dragonfly config files + several changes in ChangeLog.
[exim.git] / doc / doc-txt / ChangeLog
1509d3a8 1$Cambridge: exim/doc/doc-txt/ChangeLog,v 1.216 2005/09/06 13:21:06 ph10 Exp $
3Change log file for Exim from version 4.21
6Exim version 4.53
9TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14PH/02 Patch to exigrep to allow it to work on syslog lines.
16PH/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.
20PH/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.
24PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
26 operating systems.
28PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34PH/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
36 file for comments.
38PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40PH/10 Named domain lists were not working if used in a queue_smtp_domains
41 setting.
43PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48PH/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.
53PH/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.
59JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61TK/03 Fix log output including CR from clamd.
63PH/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.
66PH/15 Installed PCRE 6.2
68PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70PH/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
72 HP-UX compiler.
74PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76PH/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).
80PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86PH/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.
97PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100TK/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.
103PH/24 Include config.h inside local_scan.h so that configuration settings are
104 available.
106PH/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
108 received.
110PH/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.
115PH/27 Added the % operator to ${eval:}.
117PH/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
120 exists".
122PH/29 Added "crypteq" to the list of supported features that Exim outputs when
123 -bV or -d is used.
125PH/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.
aa2b5c79 129
130PH/31 Missing "BOOL" in function definition in filtertest.c.
132PH/32 Applied Sieve patches from the maintainer.
134TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
136PH/33 Added "verify = not_blind".
138PH/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
143 settings.
145PH/35 Installed configuration files for Dragonfly.
bbe902f0 147
148Exim version 4.52
151TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
153PH/01 When a transport filter timed out in a pipe delivery, and the pipe
154 command itself ended in error, the underlying message about the transport
155 filter timeout was being overwritten with the pipe command error. Now the
156 underlying error message should be appended to the second error message.
158TK/01 Fix poll() being unavailable on Mac OSX 10.2.
160PH/02 Reduce the amount of output that "make" produces by default. Full output
161 can still be requested.
163PH/03 The warning log line about a condition test deferring for a "warn" verb
164 was being output only once per connection, rather than after each
165 occurrence (because it was using the same function as for successful
166 "warn" verbs). This seems wrong, so I have changed it.
168TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
169 it should not have, which might have caused a crash in the right
170 circumstances, but probably never did.
172PH/04 Installed a modified version of Tony Finch's patch to make submission
173 mode fix the return path as well as the Sender: header line, and to
174 add a /name= option so that you can make the user's friendly name appear
175 in the header line.
177TF/03 Added the control = fakedefer ACL modifier.
179TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
180 Mark Lowes for thorough testing.
870f6ba8 181
182TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
184TK/03 Merged latest SRS patch from Miles Wilton.
186PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
187 with the definition in sysexits.h (which is #included earlier).
188 Fortunately, Exim does not actually use EX_OK. The code used to try to
189 preserve the sysexits.h value, by assumimg that macro definitions were
190 scanned for macro replacements. I have been disabused of this notion,
191 so now the code just undefines EX_OK before #including unistd.h.
11d337a4 192
193PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
194 in the smtp transport. When a block could not be written in a single
195 write() function, the timeout was being re-applied to each part-write.
196 This seems wrong - if the receiver was accepting one byte at a time it
197 would take for ever. The timeout is now adjusted when this happens. It
198 doesn't have to be particularly precise.
200TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
201 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
203PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
204 without a domain if the check to <postmaster@domain> fails.
206SC/01 Eximstats: added -xls and the ability to specify output files
207 (patch written by Frank Heydlauf).
209SC/02 Eximstats: use FileHandles for outputing results.
211SC/03 Eximstats: allow any combination of xls, txt, and html output.
213SC/04 Eximstats: fixed display of large numbers with -nvr option
215SC/05 Eximstats: fixed merging of reports with empty tables.
217SC/06 Eximstats: added the -include_original_destination flag
219SC/07 Eximstats: removed tabs and trailing whitespace.
221TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
223TK/06 MBOX spool code: Add real "From " MBOX separator line
224 so the .eml file is really in mbox format (even though
225 most programs do not really care). Patch from Alex Miller.
227TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
228 The latter is generated from $received_to and is only set if the
229 message has one envelope recipient. SA can use these headers,
230 obviously out-of-the-box. Patch from Alex Miller.
232PH/08 The ${def test on a variable was returning false if the variable's
233 value was "0", contrary to what the specification has always said!
234 The result should be true unless the variable is empty.
236PH/09 The syntax error of a character other than { following "${if
237 def:variable_name" (after optional whitespace) was not being diagnosed.
238 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
239 accidental colon was present, for example, could give incorrect results.
241PH/10 Tidied the code in a number of places where the st_size field of a stat()
242 result is used (not including appendfile, where other changes are about
243 to be made).
245PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
246 This involved changing a lot of size variables from int to off_t. It
247 should work with maildirs and everything.
249TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
250 spamd dying while we are connected to it.
252TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
253 <jgh@wizmail.org>
255PH/12 Applied Alex Kiernan's patch for the API change for the error callback
256 function for BDB 4.3.
258PH/13 Changed auto_thaw such that it does not apply to bounce messages.
260PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
261 the sources for PCRE have been re-arranged and more files are now
262 involved.
264PH/15 The code I had for printing potentially long long variables in PH/11
265 above was not the best (it lost precision). The length of off_t variables
266 is now inspected at build time, and an appropriate printing format (%ld
267 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
268 to be "long long int" or "long int". This is needed for the internal
269 formatting function string_vformat().
b1c749bb 270
271PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
272 the configuration file to be ":syslog", then the script "guesses" where
273 the logs files are, rather than using the compiled in default. In our
274 case the guess is not the same as the compiled default, so the script
275 suddenly stopped working when I started to use syslog. The patch checks
276 to see if log_file_path is "". If so, it attempts to read it from exim
277 with no configuration file to get the compiled in version, before it
278 falls back to the previous guessing code."
280TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
281 implementing BATV in an Exim configuration. See NewStuff for the gory
282 details.
284PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
285 Makefile that are specific to HP-UX.
287PH/18 If the "use_postmaster" option was set for a recipient callout together
288 with the "random" option, the postmaster address was used as the MAIL
289 FROM address for the random test, but not for the subsequent recipient
290 test. It is now used for both.
292PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
293 patch removes a few documentation additions to RFC 3028, because the
294 latest draft now contains them. It adds the new en;ascii-case comparator
295 and a new error check for 8bit text in MIME parts. Comparator and
296 require names are now matched exactly. I enabled the subaddress
297 extension, but it is not well tested yet (read: it works for me)."
299PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
300 rework some of the code of TK/09 above to avoid the hardwired use of
301 "%lld" and "long long". Replaced the call to snprintf() with a call to
302 string_vformat().
304PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
305 records point to non-existent hosts", "retry timeout exceeded", and
306 "retry time not reached for any host after a long failure period".
ca02eafb 307
308PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
309 experimental DomainKeys support:
311 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
312 (2) On an error such as an illegally used "control", the wrong name for
313 the control was given.
315 These problems did NOT occur unless DomainKeys support was compiled.
317PH/23 Added daemon_startup_retries and daemon_startup_sleep.
319PH/24 Added ${if match_ip condition.
321PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
322 databases so that it will be absolutely obvious if a crash occurs in the
323 DB library. This is a regular occurrence (often caused by mis-matched
324 db.h files).
ff790e47 326PH/26 Insert a lot of missing (void) casts for functions such as chown(),
327 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
328 picked up on a user's system that detects such things. There doesn't seem
329 to be a gcc warning option for this - only an attribute that has to be
330 put on the function's prototype. It seems that in Fedora Core 4 they have
331 set this on a number of new functions. No doubt there will be more in due
332 course.
ff790e47 333
334PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
335 specify a transport. However, if an address that was verified by such a
336 router was the subject of a callout, Exim crashed because it tried to
337 read the rcpt_include_affixes from the non-existent transport. Now it
338 just assumes that the setting of that option is false. This bug was
339 introduced by 4.51/PH/31.
341PH/28 Changed -d+all to exclude +memory, because that information is very
342 rarely of interest, but it makes the output a lot bigger. People tend to
343 do -d+all out of habit.
345PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
346 code in os-type was giving problems when libc.so lives in lib64, like on
347 x86_64 Fedora Core.
349PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
350 aren't the modern standard, and it seems that some systems' include files
351 don't always have them. Exim was already checking for some of the newer
352 ones like T_AAAA, and defining it itself. I've added checks for all the
353 record types that Exim uses.
355PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
356 not automatically generating a new one, as it is supposed to. This
357 prevented TLS from working. If the file did exist, but contained invalid
358 data, a new version was generated, as expected. It was only the case of a
359 non-existent file that was broken.
361TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
362 with a change in libdomainkeys > 0.64.
364TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
365 from DNS. If the selector record carries the flag, it now has
366 precedence over the domain-wide flag.
368TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
370PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
371 the use of a port name that isn't defined in /etc/services) occurred, the
372 message was deferred as in a normal delivery, and thus remained on the
373 spool, instead of being failed because of the mua_wrapper setting. This
374 is now fixed, and I tidied up some of the mua_wrapper messages at the
375 same time.
377SC/08 Eximstats: whilst parsing the mainlog(s), store information about
378 the messages in a hash of arrays rather than using individual hashes.
379 This is a bit cleaner and results in dramatic memory savings, albeit
380 at a slight CPU cost.
382SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
383 as requested by Marc Sherman.
385SC/10 Eximstats: added histograms for user specified patterns as requested
386 by Marc Sherman.
388SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
390PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
391 fopen() in the content-scanning modules that did not already have it.
e7ad8a65 393
394Exim version 4.51
397TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
398 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2f079f46 400TK/02 Fix ACL "control" statement not being available in MIME ACL.
402TK/03 Fix ACL "regex" condition not being available in MIME ACL.
404PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
405 to test Sieve filters that use "vacation".
407PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
408 that changes the way the GnuTLS parameters are stored in the cache file.
409 The new format can be generated externally. For backward compatibility,
410 if the data in the cache doesn't make sense, Exim assumes it has read an
411 old-format file, and it generates new data and writes a new file. This
412 means that you can't go back to an older release without removing the
413 file.
415PH/03 A redirect router that has both "unseen" and "one_time" set does not
416 work if there are any delivery delays because "one_time" forces the
417 parent to be marked "delivered", so its unseen clone is never tried
418 again. For this reason, Exim now forbids the simultaneous setting of
419 these two options.
421PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
422 redirected to themselves ("homonym" addresses). Read the long ChangeLog
423 entry if you want to know the details. The fix, however, neglected to
424 consider the case when local delivery batching is involved. The test for
425 "previously delivered" was not happening when checking to see if an
426 address could be batched with a previous (undelivered) one; under
427 certain circumstances this could lead to multiple deliveries to the same
c2c19e9d 428 address.
430PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
431 in its include files, and this causes problems building Exim.
433PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
434 header_syntax) but Exim was just ignoring anything given after a slash.
435 In particular, this caused confusion with an attempt to use "verify =
436 reverse_host_lookup/defer_ok". An error is now given when options are
437 supplied for verify items that do not have them. (Maybe reverse_host_
438 lookup should have a defer_ok option, but that's a different point.)
440PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
441 defined by RFC 821) to 2048, because there were problems with some AUTH
442 commands, and RFC 1869 says the size should be increased for extended
443 SMTP commands that take arguments.
445PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
446 Finch).
448PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
449 "unknown" error; now it says that the functionality isn't in the binary.
8d67ada3 450
451PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
452 an address' error message when a string expansion fails (syntax or
453 whatever). Otherwise the password may appear in the log. Following change
454 PH/42 below, there is no longer a chance of it appearing in a bounce
455 message.
49c2d5ea 456
457PH/11 Installed exipick version 20050225.0 from John Jetmore.
459PH/12 If the last host in a fallback_hosts list was multihomed, only the first
460 of its addresses was ever tried. (Bugzilla bug #2.)
462PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
463 the result incorrectly in the debug output. (It correctly added a newline
464 to what was transported.)
466TF/01 Added $received_time.
468PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
469 commented out examples of how to interface to a virus scanner and to
470 SpamAssassin. Also added commented examples of av_scanner and
471 spamd_address settings.
473PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
474 and controls are allowed in which ACLs. There were a couple of minor
475 errors. Some of the entries in the conditions table (which is a table of
476 where they are NOT allowed) were getting very unwieldy; rewrote them as a
477 negation of where the condition IS allowed.
479PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
481PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
482 header file does not have a version number, so I've had to invent a new
483 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
484 API. The code is untested by me (my Linux distribution still has 0.3.2 of
485 radiusclient), but it was contributed by a Radius user.
487PH/18 Installed Lars Mainka's patch for the support of CRL collections in
488 files or directories, for OpenSSL.
490PH/19 When an Exim process that is running as root has to create an Exim log
491 file, it does so in a subprocess that runs as exim:exim so as to get the
492 ownership right at creation (otherwise, other Exim processes might see
493 the file with the wrong ownership). There was no test for failure of this
494 fork() call, which would lead to the process getting stuck as it waited
495 for a non-existent subprocess. Forks do occasionally fail when resources
496 run out. I reviewed all the other calls to fork(); they all seem to check
497 for failure.
499PH/20 When checking for unexpected SMTP input at connect time (before writing
500 the banner), Exim was not dealing correctly with a non-positive return
501 from the read() function. If the client had disconnected by this time,
502 the result was a log entry for a synchronization error with an empty
503 string after "input=" when read() returned zero. If read() returned -1
504 (an event I could not check), uninitialized data bytes were printed.
505 There were reports of junk text (parts of files, etc) appearing after
506 "input=".
508PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
510PH/22 Added support for macro redefinition, and (re)definition in between
511 driver and ACL definitions.
513PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
514 forgetting to use the resulting value; it was using the unexpanded value.
516PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
517 hadn't been configured. The fix is from Juergen Kreileder, who
518 understands it better than I do:
520 "Here's what I see happening with three configured cyrus_sasl
521 authenticators configured (plain, login, cram-md5):
523 On startup auth_cyrus_sasl_init() gets called for each of these.
524 This means three calls to sasl_listmech() without a specified mech_list.
525 => SASL tests which mechs of all available mechs actually work
526 => three warnings about OTP not working
527 => the returned list contains: plain, login, cram-md5, digest-md5, ...
529 With the patch, sasl_listmech() also gets called three times. But now
530 SASL's mech_list option is set to the server_mech specified in the the
531 authenticator. Or in other words, the answer from sasl_listmech()
532 gets limited to just the mech you're testing for (which is different
533 for each call.)
534 => the return list contains just 'plain' or 'login', 'cram-md5' or
535 nothing depending on the value of ob->server_mech.
537 I've just tested the patch: Authentication still works fine,
538 unavailable mechs specified in the exim configuration are still
539 caught, and the auth.log warnings about OTP are gone."
541PH/25 When debugging is enabled, the contents of the command line are added
542 to the debugging output, even when log_selector=+arguments is not
543 specified.
545PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
546 answer is "GNU", and only if the return is "GNU/something" is the answer
547 "Linux".
549PH/27 $acl_verify_message is now set immediately after the failure of a
550 verification in an ACL, and so is available in subsequent modifiers. In
551 particular, the message can be preserved by coding like this:
553 warn !verify = sender
554 set acl_m0 = $acl_verify_message
556 Previously, $acl_verify_message was set only while expanding "message"
557 and "log_message" when a very denied access.
559PH/28 Modified OS/os.c-Linux with
561 -#ifndef OS_LOAD_AVERAGE
562 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
564 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
565 nomenclature these days.)
567PH/29 Installed patch from the Sieve maintainer that adds the options
568 sieve_useraddress and sieve_subaddress to the redirect router.
570PH/30 In these circumstances:
571 . Two addresses routed to the same list of hosts;
572 . First host does not offer TLS;
573 . First host accepts first address;
574 . First host gives temporary error to second address;
575 . Second host offers TLS and a TLS session is established;
576 . Second host accepts second address.
577 Exim incorrectly logged both deliveries with the TLS parameters (cipher
578 and peerdn, if requested) that were in fact used only for the second
579 address.
7e8bec7a 580
581PH/31 When doing a callout as part of verifying an address, Exim was not paying
582 attention to any local part prefix or suffix that was matched by the
583 router that accepted the address. It now behaves in the same way as it
584 does for delivery: the affixes are removed from the local part unless
585 rcpt_include_affixes is set on the transport.
587PH/32 Add the sender address, as F=<...>, to the log line when logging a
588 timeout during the DATA phase of an incoming message.
590PH/33 Sieve envelope tests were broken for match types other than :is. I have
591 applied a patch sanctioned by the Sieve maintainer.
c688b954 592
593PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
594 the uid or gid is negative. A case of a negative gid caused this to be
595 noticed. The fix allows for either to be negative.
597PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
598 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
599 overlooked.
601PH/36 The change PH/12 above was broken. Fixed it.
603PH/37 Exim used to check for duplicate addresses in the middle of routing, on
604 the grounds that routing the same address twice would always produce the
605 same answer. This might have been true once, but it is certainly no
606 longer true now. Routing a child address may depend on the previous
607 routing that produced that child. Some complicated redirection strategies
608 went wrong when messages had multiple recipients, and made Exim's
609 behaviour dependent on the order in which the addresses were given.
611 I have moved the duplicate checking until after the routing is complete.
612 Exim scans the addresses that are assigned to local and remote
613 transports, and removes any duplicates. This means that more work will be
614 done, as duplicates will always all be routed, but duplicates are
615 presumably rare, so I don't expect this is of any significance.
617 For deliveries to pipes, files, and autoreplies, the duplicate checking
618 still happens during the routing process, since they are not going to be
619 routed further.
621PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
622 It corrects a timeout issue with spamd. This is Ian's comment: "The
623 background is that sometimes spamd either never reads data from a
624 connection it has accepted, or it never writes response data. The exiscan
625 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
626 blindly assumes that writes won't block so it may never time out."
628PH/39 Allow G after quota size as well as K and M.
630PH/40 The value set for $authenticated_id in an authenticator may not contain
631 binary zeroes or newlines because the value is written to log lines and
632 to spool files. There was no check on this. Now the value is run through
633 the string_printing() function so that such characters are converted to
634 printable escape sequences.
636PH/41 $message_linecount is a new variable that contains the total number of
637 lines in the message. Compare $body_linecount, which is the count for the
638 body only.
640PH/42 Exim no longer gives details of delivery errors for specific addresses in
641 bounce and delay warning messages, except in certain special cases, which
642 are as follows:
644 (a) An SMTP error message from a remote host;
645 (b) A message specified in a :fail: redirection;
646 (c) A message specified in a "fail" command in a system filter;
647 (d) A message specified in a FAIL return from the queryprogram router;
648 (e) A message specified by the cannot_route_message router option.
650 In these cases only, Exim does include the error details in bounce and
651 warning messages. There are also a few cases where bland messages such
652 as "unrouteable address" or "local delivery error" are given.
654PH/43 $value is now also set for the "else" part of a ${run expansion.
656PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
657 being worked on, but at least Exim now implements the latest version to
658 play with."
660PH/45 In a pipe transport, although a timeout while waiting for the pipe
661 process to complete was treated as a delivery failure, a timeout while
662 writing the message to the pipe was logged, but erroneously treated as a
663 successful delivery. Such timeouts include transport filter timeouts. For
664 consistency with the overall process timeout, these timeouts are now
665 treated as errors, giving rise to delivery failures by default. However,
666 there is now a new Boolean option for the pipe transport called
667 timeout_defer, which, if set TRUE, converts the failures into defers for
668 both kinds of timeout. A transport filter timeout is now identified in
669 the log output.
671PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
672 systems where "make" and "gmake" are different, calling "gmake" at top
673 level broke things. I've arranged for the value of $(MAKE) to be passed
674 from the Makefile to this script so that it can call the same version of
675 "make".
7982096b 677
678A note about Exim versions 4.44 and 4.50
681Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
682changes of various kinds. As a consequence, a big documentation update was
683needed. This delayed the release for rather longer than seemed good, especially
684in the light of a couple of (minor) security issues. Therefore, the changes
685that fixed bugs were backported into 4.43, to create a 4.44 maintenance
686release. So 4.44 and 4.50 are in effect two different branches that both start
687from 4.43.
689I have left the 4.50 change log unchanged; it contains all the changes since
6904.43. The change log for 4.44 is below; many of its items are identical to
691those for 4.50. This seems to be the most sensible way to preserve the
692historical information.
f7b63901 695Exim version 4.50
698 1. Minor wording change to the doc/README.SIEVE file.
139059f6 700 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5fe762f6 701 computation of the current number of files was incorrect.
495ae4b0 702
703 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
704 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
705 place.
707 4. Give more explanation in the error message when the command for a transport
708 filter fails to execute.
710 5. There are several places where Exim runs a non-Exim command in a
711 subprocess. The SIGUSR1 signal should be disabled for these processes. This
712 was being done only for the command run by the queryprogram router. It is
713 now done for all such subprocesses. The other cases are: ${run, transport
714 filters, and the commands run by the lmtp and pipe transports.
716 6. Added CONFIGURE_GROUP build-time option.
718 7. Some older OS have a limit of 256 on the maximum number of file
719 descriptors. Exim was using setrlimit() to set 1000 as a large value
720 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
721 systems. I've change it so that if it can't get 1000, it tries for 256.
35edf2ff 722
723 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
724 was an oversight, and furthermore, ever since the addition of extra
725 controls (e.g. 4.43/32), the checks on when to allow different forms of
726 "control" were broken. There should now be diagnostics for all cases when a
727 control that does not make sense is encountered.
729 9. Added the /retain_sender option to "control=submission".
73110. $recipients is now available in the predata ACL (oversight).
73311. Tidy the search cache before the fork to do a delivery from a message
734 received from the command line. Otherwise the child will trigger a lookup
735 failure and thereby defer the delivery if it tries to use (for example) a
736 cached ldap connection that the parent has called unbind on.
73812. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
739 of $address_data from the recipient verification was clobbered by the
740 sender verification.
74213. The value of address_data from a sender verification is now available in
743 $sender_address_data in subsequent conditions in the ACL statement.
74514. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
74715. Added a new option "connect=<time>" to callout options, to set a different
748 connection timeout.
75016. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
751 was its contents. (It was OK if the option was not defined at all.)
75317. A "Completed" log line is now written for messages that are removed from
754 the spool by the -Mrm option.
75618. New variables $sender_verify_failure and $recipient_verify_failure contain
757 information about exactly what failed.
75919. Added -dd to debug only the daemon process.
76120. Incorporated Michael Haardt's patch to ldap.c for improving the way it
762 handles timeouts, both on the server side and network timeouts. Renamed the
763 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
76521. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
766 It is now set to "smtps".
76822. $host_address is now set to the target address during the checking of
769 ignore_target_hosts.
77123. When checking ignore_target_hosts for an ipliteral router, no host name was
772 being passed; this would have caused $sender_host_name to have been used if
773 matching the list had actually called for a host name (not very likely,
774 since this list is usually IP addresses). A host name is now passed as
775 "[x.x.x.x]".
77724. Changed the calls that set up the SIGCHLD handler in the daemon to use the
778 code that specifies a non-restarting handler (typically sigaction() in
779 modern systems) in an attempt to fix a rare and obscure crash bug.
78125. Narrowed the window for a race in the daemon that could cause it to ignore
782 SIGCHLD signals. This is not a major problem, because they are used only to
783 wake it up if nothing else does.
78526. A malformed maildirsize file could cause Exim to calculate negative values
786 for the mailbox size or file count. Odd effects could occur as a result.
787 The maildirsize information is now recalculated if the size or filecount
788 end up negative.
79027. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
791 support for a long time. Removed HAVE_SYS_VFS_H.
79328. Installed the latest version of exipick from John Jetmore.
79529. In an address list, if the pattern was not a regular expression, an empty
796 subject address (from a bounce message) matched only if the pattern was an
797 empty string. Non-empty patterns were not even tested. This was the wrong
798 because it is perfectly reasonable to use an empty address as part of a
799 database query. An empty address is now tested by patterns that are
800 lookups. However, all the other forms of pattern expect the subject to
801 contain a local part and a domain, and therefore, for them, an empty
802 address still always fails if the pattern is not itself empty.
80430. Exim went into a mad DNS loop when attempting to do a callout where the
805 host was specified on an smtp transport, and looking it up yielded more
806 than one IP address.
80831. Re-factored the code for checking spool and log partition space into a
809 function that finds that data and another that does the check. The former
810 is then used to implement four new variables: $spool_space, $log_space,
811 $spool_inodes, and $log_inodes.
81332. The RFC2047 encoding function was originally intended for short strings
814 such as real names; it was not keeping to the 75-character limit for
815 encoded words that the RFC imposes. It now respects the limit, and
816 generates multiple encoded words if necessary. To be on the safe side, I
817 have increased the buffer size for the ${rfc2047: expansion operator from
818 1024 to 2048 bytes.
82033. It is now permitted to omit both strings after an "if" condition; if the
821 condition is true, the result is "true". As before, when the second string
822 is omitted, a false condition yields an empty string. This makes it less
823 cumbersome to write custom ACL and router conditions.
82534. Failure to deliver a bounce message always caused it to be frozen, even if
826 there was an errors_to setting on the router. The errors_to setting is now
827 respected.
82935. If an IPv6 address is given for -bh or -bhc, it is now converted to the
830 canonical form (fully expanded) before being placed in
831 $sender_host_address.
83336. The table in the code that translates DNS record types into text (T_A to
834 "A" for instance) was missing entries for NS and CNAME. It is just possible
835 that this could have caused confusion if both these types were looked up
836 for the same domain, because the text type is used as part of Exim's
837 per-process caching. But the chance of anyone hitting this buglet seems
838 very small.
84037. The dnsdb lookup has been extended in a number of ways.
842 (1) There is a new type, "zns", which walks up the domain tree until it
843 finds some nameserver records. It should be used with care.
845 (2) There is a new type, "mxh", which is like "mx" except that it returns
846 just the host names, not the priorities.
848 (3) It is now possible to give a list of domains (or IP addresses) to be
849 looked up. The behaviour when one of the lookups defers can be
850 controlled by a keyword.
7bb56e1f 851
ea3bc19b 852 (4) It is now possible to specify the separator character for use when
7bb56e1f 853 multiple records are returned.
33397d19 854
85538. The dnslists ACL condition has been extended: it is now possible to supply
856 a list of IP addresses and/or domains to be looked up in a particular DNS
857 domain.
85939. Added log_selector=+queue_time_overall.
86140. When running the queue in the test harness, wait just a tad after forking a
862 delivery process, to get repeatability of debugging output.
86441. Include certificate and key file names in error message when GnuTLS fails
865 to set them up, because the GnuTLS error message doesn't include the name
866 of the failing file when there is a problem reading it.
86842. Allow both -bf and -bF in the same test run.
87043. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
87244. The "Exiscan patch" is now merged into the mainline Exim source.
87445. Sometimes the final signoff response after QUIT could fail to get
875 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
876 < 0 before doing a fflush(). This bug looks as though it goes back to the
877 introduction of TLS in release 3.20, but "sometimes" must have been rare
878 because the tests only now provoked it.
88046. Reset the locale to "C" after calling embedded Perl, in case it was changed
881 (this can affect the format of dates).
88347. exim_tidydb, when checking for the continued existence of a message for
884 which it has found a message-specific retry record, was not finding
885 messages that were in split spool directories. Consequently, it was
886 deleting retry records that should have stayed in existence.
88848. Steve fixed some bugs in eximstats.
89049. The SPA authentication driver was not abandoning authentication and moving
891 on to the next authenticator when an expansion was forced to fail,
892 contradicting the general specification for all authenticators. Instead it
893 was generating a temporary error. It now behaves as specified.
89550. The default ordering of permitted cipher suites for GnuTLS was pessimal
896 (the order specifies the preference for clients). The order is now AES256,
897 AES128, 3DES, ARCFOUR128.
89951. Small patch to Sieve code - explicitly set From: when generating an
900 autoreply.
90252. Exim crashed if a remote delivery caused a very long error message to be
903 recorded - for instance if somebody sent an entire SpamAssassin report back
904 as a large number of 550 error lines. This bug was coincidentally fixed by
905 increasing the size of one of Exim's internal buffers (big_buffer) that
906 happened as part of the Exiscan merge. However, to be on the safe side, I
907 have made the code more robust (and fixed the comments that describe what
908 is going on).
91053. Now that there can be additional text after "Completed" in log lines (if
911 the queue_time_overall log selector is set), a one-byte patch to exigrep
912 was needed to allow it to recognize "Completed" as not the last thing in
913 the line.
91554. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
916 patch that reportedly fixes this has been added. I am not expert enough to
917 create a test for it. This is what the patch creator wrote:
919 "I found a little strange behaviour of ldap code when working with
920 Windows 2003 AD Domain, where users was placed in more than one
921 Organization Units. When I tried to give exim partial DN, the exit code
922 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
923 But simultaneously result of request was absolutely normal ldap result,
924 so I produce this patch..."
926 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
927 REFERENCE, so I have modified the code to exclude the patch when that macro
928 is not defined.
93055. Some experimental protocols are using DNS PTR records for new purposes. The
931 keys for these records are domain names, not reversed IP addresses. The
932 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
933 leaves it alone. Component reversal etc. now happens only for IP addresses.
ea3a6f44 934 CAN-2005-0021
7102e136 935
93656. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
93857. Double the size of the debug message buffer (to 2048) so that more of very
939 long debug lines gets shown.
94158. The exicyclog utility now does better if the number of log files to keep
942 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
94459. Two changes related to the smtp_active_hostname option:
946 (1) $smtp_active_hostname is now available as a variable.
947 (2) The default for smtp_banner uses $smtp_active_hostname instead
948 of $primary_hostname.
95060. The host_aton() function is supposed to be passed a string that is known
951 to be a valid IP address. However, in the case of IPv6 addresses, it was
952 not checking this. This is a hostage to fortune. Exim now panics and dies
953 if the condition is not met. A case was found where this could be provoked
954 from a dnsdb PTR lookup with an IPv6 address that had more than 8
955 components; fortuitously, this particular loophole had already been fixed
956 by change 4.50/55 above.
958 If there are any other similar loopholes, the new check in host_aton()
959 itself should stop them being exploited. The report I received stated that
960 data on the command line could provoke the exploit when Exim was running as
961 exim, but did not say which command line option was involved. All I could
962 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
963 running as the user.
ea3a6f44 964 CAN-2005-0021
96661. There was a buffer overflow vulnerability in the SPA authentication code
967 (which came originally from the Samba project). I have added a test to the
968 spa_base64_to_bits() function which I hope fixes it.
ea3a6f44 969 CAN-2005-0022
b975ba52 970
97162. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
972 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
97463. The daemon start-up calls getloadavg() while still root for those OS that
975 need the first call to be done as root, but it missed one case: when
976 deliver_queue_load_max is set with deliver_drop_privilege. This is
977 necessary for the benefit of the queue runner, because there is no re-exec
978 when deliver_drop_privilege is set.
98064. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
981 This has been fixed.
98365. Caching of lookup data for "hosts =" ACL conditions, when a named host list
984 was in use, was not putting the data itself into the right store pool;
985 consequently, it could be overwritten for a subsequent message in the same
986 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
987 the caching.)
98966. Added hosts_max_try_hardlimit to the smtp transport, default 50.
99167. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
992 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
993 the function were treating the return as a boolean value, which happened to
994 work because 0=false and not-0=true, but is not correct code.
99668. The host_aton() function was not handling scoped IPv6 addresses (those
997 with, for example, "%eth0" on the end) correctly.
99969. Fixed some compiler warnings in acl.c for the bitmaps specified with
1000 negated items (that is, ~something) in unsigned ints. Some compilers
1001 apparently mutter when there is no cast.
100370. If an address verification called from an ACL failed, and did not produce a
1004 user-specific message (i.e. there was only a "system" message), nothing was
1005 put in $acl_verify_message. In this situation, it now puts the system
1006 message there.
100871. Change 4.23/11 added synchronization checking at the start of an SMTP
1009 session; change 4.31/43 added the unwanted input to the log line - except
1010 that it did not do this in the start of session case. It now does.
101272. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1013 This is wrong and can cause the other end to generate a synchronization
1014 error if it is another Exim or anything else that does the synchronization
1015 check. A QUIT command is no longer sent after a timeout.
101773. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1018 during host lookups.
102074. The defer_ok option of callout verification was not working if it was used
1021 when verifying addresses in header lines, that is, for this case:
1023 verify = header_sender/callout=defer_ok
102575. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1026 those file descriptors could be used for SMTP connections. If anything
1027 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1028 could be sent to the SMTP client, causing chaos. The daemon now opens
1029 stdin, stdout, and stderr to /dev/null when it puts itself into the
1030 background.
103276. Arrange for output from Perl's "warn" command to be written to Exim's main
1033 log by default. The user can override this with suitable Perl magic.
103577. The use of log_message on a "discard" ACL verb, which is supposed to add to
1036 the log message when discard triggers, was not working for the DATA ACL or
1037 for the non-SMTP ACL.
103978. Error message wording change in sieve.c.
104179. If smtp_accept_max_per_host was set, the number of connections could be
1042 restricted to fewer than expected, because the daemon was trying to set up
1043 a new connection before checking whether the processes handling previous
1044 connections had finished. The check for completed processes is now done
1045 earlier. On busy systems, this bug wouldn't be noticed because something
1046 else would have woken the daemon, and it would have reaped the completed
1047 process earlier.
104980. If a message was submitted locally by a user whose login name contained one
1050 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1051 It caused a spool format error. I have fixed the spool reading code. A
1052 related problem was that the "from" clause in the Received: line became
1053 illegal because of the space(s). It is now covered by ${quote_local_part.
105581. Included the latest eximstats from Steve (adds average sizes to HTML Top
1056 tables).
105882. Updated OS/Makefile-AIX as per message from Mike Meredith.
106083. Patch from Sieve maintainer to fix unterminated string problem in
1061 "vacation" handling.
106384. Some minor changes to the Linux configuration files to help with other
1064 OS variants using glibc.
106685. One more patch for Sieve to update vacation handling to latest spec.
495ae4b0 1068
1070See the note above about the 4.44 and 4.50 releases.
1074Exim version 4.44
1077 1. Change 4.43/35 introduced a bug that caused file counts to be
1078 incorrectly computed when quota_filecount was set in an appendfile
1079 transport
1081 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1082 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1083 place.
1085 3. Give more explanation in the error message when the command for a transport
1086 filter fails to execute.
1088 4. There are several places where Exim runs a non-Exim command in a
1089 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1090 was being done only for the command run by the queryprogram router. It is
1091 now done for all such subprocesses. The other cases are: ${run, transport
1092 filters, and the commands run by the lmtp and pipe transports.
1094 5. Some older OS have a limit of 256 on the maximum number of file
1095 descriptors. Exim was using setrlimit() to set 1000 as a large value
1096 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1097 systems. I've change it so that if it can't get 1000, it tries for 256.
1099 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1100 was an oversight, and furthermore, ever since the addition of extra
1101 controls (e.g. 4.43/32), the checks on when to allow different forms of
1102 "control" were broken. There should now be diagnostics for all cases when a
1103 control that does not make sense is encountered.
1105 7. $recipients is now available in the predata ACL (oversight).
1107 8. Tidy the search cache before the fork to do a delivery from a message
1108 received from the command line. Otherwise the child will trigger a lookup
1109 failure and thereby defer the delivery if it tries to use (for example) a
1110 cached ldap connection that the parent has called unbind on.
1112 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1113 of $address_data from the recipient verification was clobbered by the
1114 sender verification.
111610. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1117 was its contents. (It was OK if the option was not defined at all.)
111911. A "Completed" log line is now written for messages that are removed from
1120 the spool by the -Mrm option.
112212. $host_address is now set to the target address during the checking of
1123 ignore_target_hosts.
112513. When checking ignore_target_hosts for an ipliteral router, no host name was
1126 being passed; this would have caused $sender_host_name to have been used if
1127 matching the list had actually called for a host name (not very likely,
1128 since this list is usually IP addresses). A host name is now passed as
1129 "[x.x.x.x]".
113114. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1132 code that specifies a non-restarting handler (typically sigaction() in
1133 modern systems) in an attempt to fix a rare and obscure crash bug.
113515. Narrowed the window for a race in the daemon that could cause it to ignore
1136 SIGCHLD signals. This is not a major problem, because they are used only to
1137 wake it up if nothing else does.
113916. A malformed maildirsize file could cause Exim to calculate negative values
1140 for the mailbox size or file count. Odd effects could occur as a result.
1141 The maildirsize information is now recalculated if the size or filecount
1142 end up negative.
114417. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1145 support for a long time. Removed HAVE_SYS_VFS_H.
ea3a6f44 114718. Updated exipick to current release from John Jetmore.
114919. Allow an empty sender to be matched against a lookup in an address list.
1150 Previously the only cases considered were a regular expression, or an
1151 empty pattern.
115320. Exim went into a mad DNS lookup loop when doing a callout where the
1154 host was specified on the transport, if the DNS lookup yielded more than
1155 one IP address.
115721. The RFC2047 encoding function was originally intended for short strings
1158 such as real names; it was not keeping to the 75-character limit for
1159 encoded words that the RFC imposes. It now respects the limit, and
1160 generates multiple encoded words if necessary. To be on the safe side, I
1161 have increased the buffer size for the ${rfc2047: expansion operator from
1162 1024 to 2048 bytes.
bbe902f0 1163
116422. Failure to deliver a bounce message always caused it to be frozen, even if
1165 there was an errors_to setting on the router. The errors_to setting is now
1166 respected.
116823. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1169 canonical form (fully expanded) before being placed in
1170 $sender_host_address.
117224. Updated eximstats to version 1.33
117425. Include certificate and key file names in error message when GnuTLS fails
1175 to set them up, because the GnuTLS error message doesn't include the name
1176 of the failing file when there is a problem reading it.
117826. Expand error message when OpenSSL has problems setting up cert/key files.
ea3a6f44 1179 As per change 25.
bbe902f0 1180
118127. Reset the locale to "C" after calling embedded Perl, in case it was changed
1182 (this can affect the format of dates).
bbe902f0 1183
118428. exim_tidydb, when checking for the continued existence of a message for
1185 which it has found a message-specific retry record, was not finding
1186 messages that were in split spool directories. Consequently, it was
1187 deleting retry records that should have stayed in existence.
118929. eximstats updated to version 1.35
1190 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1191 1.35 - bugfix such that pie charts by volume are generated correctly
119330. The SPA authentication driver was not abandoning authentication and moving
1194 on to the next authenticator when an expansion was forced to fail,
1195 contradicting the general specification for all authenticators. Instead it
1196 was generating a temporary error. It now behaves as specified.
bbe902f0 1197
119831. The default ordering of permitted cipher suites for GnuTLS was pessimal
1199 (the order specifies the preference for clients). The order is now AES256,
1200 AES128, 3DES, ARCFOUR128.
bbe902f0 1201
120231. Small patch to Sieve code - explicitly set From: when generating an
1203 autoreply.
bbe902f0 1204
120532. Exim crashed if a remote delivery caused a very long error message to be
1206 recorded - for instance if somebody sent an entire SpamAssassin report back
1207 as a large number of 550 error lines. This bug was coincidentally fixed by
1208 increasing the size of one of Exim's internal buffers (big_buffer) that
1209 happened as part of the Exiscan merge. However, to be on the safe side, I
1210 have made the code more robust (and fixed the comments that describe what
1211 is going on).
bbe902f0 1212
121333. Some experimental protocols are using DNS PTR records for new purposes. The
1214 keys for these records are domain names, not reversed IP addresses. The
1215 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1216 leaves it alone. Component reversal etc. now happens only for IP addresses.
1217 CAN-2005-0021
121934. The host_aton() function is supposed to be passed a string that is known
1220 to be a valid IP address. However, in the case of IPv6 addresses, it was
1221 not checking this. This is a hostage to fortune. Exim now panics and dies
1222 if the condition is not met. A case was found where this could be provoked
1223 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1224 components; fortuitously, this particular loophole had already been fixed
1225 by change 4.50/55 or 4.44/33 above.
1227 If there are any other similar loopholes, the new check in host_aton()
1228 itself should stop them being exploited. The report I received stated that
1229 data on the command line could provoke the exploit when Exim was running as
1230 exim, but did not say which command line option was involved. All I could
1231 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1232 running as the user.
1233 CAN-2005-0021
123535. There was a buffer overflow vulnerability in the SPA authentication code
1236 (which came originally from the Samba project). I have added a test to the
1237 spa_base64_to_bits() function which I hope fixes it.
1238 CAN-2005-0022
124036. The daemon start-up calls getloadavg() while still root for those OS that
1241 need the first call to be done as root, but it missed one case: when
1242 deliver_queue_load_max is set with deliver_drop_privilege. This is
1243 necessary for the benefit of the queue runner, because there is no re-exec
1244 when deliver_drop_privilege is set.
bbe902f0 1245
124637. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1247 was in use, was not putting the data itself into the right store pool;
1248 consequently, it could be overwritten for a subsequent message in the same
1249 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1250 the caching.)
bbe902f0 1251
125238. Sometimes the final signoff response after QUIT could fail to get
1253 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1254 < 0 before doing a fflush(). This bug looks as though it goes back to the
1255 introduction of TLS in release 3.20, but "sometimes" must have been rare
1256 because the tests only now provoked it.
1259Exim version 4.43
1262 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1263 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1264 Now it returns an integer. A number of calls were still expecting a T/F
1265 return. Fortuitously, in all cases, the tests worked in OK situations,
1266 which is the norm. However, things would have gone wrong on any write
1267 failures on the smtp file descriptor. This function is used when sending
1268 messages over SMTP and also when doing verify callouts.
1270 2. When Exim is called to do synchronous delivery of a locally submitted
1271 message (the -odf or -odi options), it no longer closes stderr before doing
1272 the delivery.
1274 3. Implemented the mua_wrapper option.
1276 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1278 5. Implemented the functions header_remove(), header_testname(),
1279 header_add_at_position(), and receive_remove_recipient(), and exported them
1280 to local_scan().
1282 6. If an ACL "warn" statement specified the addition of headers, Exim already
1283 inserted X-ACL-Warn: at the start if there was no header name. However, it
1284 was not making this test for the second and subsequent header lines if
1285 there were newlines in the string. This meant that an invalid header could
1286 be inserted if Exim was badly configured.
1288 7. Allow an ACL "warn" statement to add header lines at the start or after all
1289 the Received: headers, as well as at the end.
1291 8. Added the rcpt_4xx retry error code.
1293 9. Added postmaster_mailfrom=xxx to callout verification option.
129510. Added mailfrom=xxxx to the callout verification option, for verify=
1296 header_sender only.
129811. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1299 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
130112. Inserted some casts to stop certain compilers warning when using pointer
1302 differences as field lengths or precisions in printf-type calls (mostly
1303 affecting debugging statements).
130513. Added optional readline() support for -be (dynamically loaded).
130714. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1308 same clock tick as a message's arrival, so that its received time was the
1309 same as the "first fail" time on the retry record, and that message
1310 remained on the queue past the ultimate address timeout, every queue runner
1311 would try a delivery (because it was past the ultimate address timeout) but
1312 after another failure, the ultimate address timeout, which should have then
1313 bounced the address, did not kick in. This was a "< instead of <=" error;
1314 in most cases the first failure would have been in the next clock tick
1315 after the received time, and all would be well.
131715. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1318 being recognized when the domain list was tested by the match_domain
1319 condition in an expansion string.
132116. Added the ${str2b64: operator.
132317. Exim was always calling setrlimit() to set a large limit for the number of
1324 processes, without checking whether the existing limit was already
1325 adequate. (It did check for the limit on file descriptors.) Furthermore,
1326 errors from getrlimit() and setrlimit() were being ignored. Now they are
1327 logged to the main and panic logs, but Exim does carry on, to try to do its
1328 job under whatever limits there are.
133018. Imported PCRE 5.0.
133219. Trivial typo in log message " temporarily refused connection" (the leading
1333 space).
133520. If the log selector return_path_on_delivery was set and an address was
1336 redirected to /dev/null, the delivery process crashed because it assumed
1337 that a return path would always be set for a "successful" delivery. In this
1338 case, the whole delivery is bypassed as an optimization, and therefore no
1339 return path is set.
134121. Internal re-arrangement: the function for sending a challenge and reading
1342 a response while authentication was assuming a zero-terminated challenge
1343 string. It's now changed to take a pointer and a length, to allow for
1344 binary data in such strings.
134622. Added the cyrus_sasl authenticator (code supplied by MBM).
134823. Exim was not respecting finduser_retries when seeking the login of the
1349 uid under which it was called; it was always trying 10 times. (The default
1350 setting of finduser_retries is zero.) Also, it was sleeping after the final
1351 failure, which is pointless.
135324. Implemented tls_on_connect_ports.
135525. Implemented acl_smtp_predata.
135726. If the domain in control=submission is set empty, Exim assumes that the
1358 authenticated id is a complete email address when it generates From: or
1359 Sender: header lines.
136127. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1362 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1363 chown and chgrp in /bin and hostname in /usr/bin.
136528. Exim was keeping the "process log" file open after each use, just as it
1366 does for the main log. This opens the possibility of it remaining open for
1367 long periods when the USR1 signal hits a daemon. Occasional processlog
1368 errors were reported, that could have been caused by this. Anyway, it seems
1369 much more sensible not to leave this file open at all, so that is what now
1370 happens.
137229. The long-running daemon process does not normally write to the log once it
1373 has entered its main loop, and it closes the log before doing so. This is
1374 so that log files can straightforwardly be renamed and moved. However,
1375 there are a couple of unusual error situations where the daemon does write
1376 log entries, and I had neglected to close the log afterwards.
137830. The text of an SMTP error response that was received during a remote
1379 delivery was being truncated at 512 bytes. This is too short for some of
1380 the long messages that one sometimes sees. I've increased the limit to
1381 1024.
138331. It is now possible to make retry rules that apply only when a message has a
1384 specific sender, in particular, an empty sender.
138632. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1387 it possible to be selective about when SMTP synchronization is enforced.
138933. Added "control = caseful_local_part" and "control = "caselower_local_part".
139132. Implemented hosts_connection_nolog.
139333. Added an ACL for QUIT.
139534. Setting "delay_warning=" to disable warnings was not working; it gave a
1396 syntax error.
139835. Added mailbox_size and mailbox_filecount to appendfile.
140036. Added control = no_multiline_responses to ACLs.
140237. There was a bug in the logic of the code that waits for the clock to tick
1403 in the case where the clock went backwards by a substantial amount such
1404 that the microsecond fraction of "now" was more than the microsecond
1405 fraction of "then" (but the whole seconds number was less).
140738. Added support for the libradius Radius client library this is found on
1408 FreeBSD (previously only the radiusclient library was supported).
1411Exim version 4.42
1414 1. When certain lookups returned multiple values in the form name=value, the
1415 quoting of the values was not always being done properly. Specifically:
1416 (a) If the value started with a double quote, but contained no whitespace,
1417 it was not quoted.
1418 (b) If the value contained whitespace other than a space character (i.e.
1419 tabs or newlines or carriage returns) it was not quoted.
1420 This fix has been applied to the mysql and pgsql lookups by writing a
1421 separate quoting function and calling it from the lookup code. The fix
1422 should probably also be applied to nisplus, ibase and oracle lookups, but
1423 since I cannot test any of those, I have not disturbed their existing code.
1425 2. A hit in the callout cache for a specific address caused a log line with no
1426 reason for rejecting RCPT. Now it says "Previous (cached) callout
1427 verification failure".
1429 3. There was an off-by-one bug in the queryprogram router. An over-long
1430 return line was truncated at 256 instead of 255 characters, thereby
1431 overflowing its buffer with the terminating zero. As well as fixing this, I
1432 have increased the buffer size to 1024 (and made a note to document this).
1434 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1435 when Exim is waiting for an SMTP response from a remote server, Exim
1436 restarts its select() call on the socket, thereby resetting its timeout.
1437 This is not a problem when such interrupts are rare. Somebody set up a cron
1438 job to run exiwhat every 2 minutes, which is less than the normal select()
1439 timeout (5 or 10 minutes). This meant that the select() timeout never
1440 kicked in because it was always reset. I have fixed this by comparing the
1441 time when an interrupt arrives with the time at the start of the first call
1442 to select(). If more time than the timeout has elapsed, the interrupt is
1443 treated as a timeout.
1445 5. Some internal re-factoring in preparation for the addition of Sieve
1446 extensions (by MH). In particular, the "personal" test is moved to a
1447 separate function, and given an option for scanning Cc: and Bcc: (which is
1448 not set for Exim filters).
1450 6. When Exim created an email address using the login of the caller as the
1451 local part (e.g. when creating a From: or Sender: header line), it was not
1452 quoting the local part when it contained special characters such as @.
1454 7. Installed new OpenBSD configuration files.
1456 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1457 try to make them clearer.
1459 9. Callout options, other than the timeout value, were being ignored when
1460 verifying sender addresses in header lines. For example, when using
1462 verify = header_sender/callout=no_cache
1464 the cache was (incorrectly) being used.
146610. Added a missing instance of ${EXE} to the exim_install script; this affects
1467 only the Cygwin environment.
146911. When return_path_on_delivery was set as a log selector, if different remote
1470 addresses in the same message used different return paths and parallel
1471 remote delivery occurred, the wrong values would sometimes be logged.
1472 (Whenever a remote delivery process finished, the return path value from
1473 the most recently started remote delivery process was logged.)
147512. RFC 3848 specifies standard names for the "with" phrase in Received: header
1476 lines when AUTH and/or TLS are in use. This is the "received protocol"
1477 field. Exim used to use "asmtp" for authenticated SMTP, without any
1478 indication (in the protocol name) for TLS use. Now it follows the RFC and
1479 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1480 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1481 names appear in log lines as well as in Received: header lines.
148313. Installed MH's patches for Sieve to add the "copy" and "vacation"
1484 extensions, and comparison tests, and to fix some bugs.
148614. Changes to the "personal" filter test:
1488 (1) The test was buggy in that it was just doing the equivalent of
1489 "contains" tests on header lines. For example, if a user's address was
1490 anne@some.where, the "personal" test would incorrectly be true for
1492 To: susanne@some.where
1494 This test is now done by extracting each address from the header in turn,
1495 and checking the entire address. Other tests that are part of "personal"
1496 are now done using regular expressions (for example, to check local parts
1497 of addresses in From: header lines).
1499 (2) The list of non-personal local parts in From: addresses has been
1500 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1501 taken from the Sieve specification recommendations.
1503 (3) If the message contains any header line starting with "List-" it is
1504 treated as non-personal.
1506 (4) The test for "circular" in the Subject: header line has been removed
1507 because it now seems ill-conceived.
150915. Minor typos in src/EDITME comments corrected.
151116. Installed latest exipick from John Jetmore.
151317. If headers_add on a router specified a text string that was too long for
1514 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1515 of string_sprintf() is now avoided.
151718. $message_body_size was not set (it was always zero) when running the DATA
1518 ACL and the local_scan() function.
152019. For the "mail" command in an Exim filter, no default was being set for
1521 the once_repeat time, causing a random time value to be used if "once" was
1522 specified. (If the value happened to be <= 0, no repeat happened.) The
1523 default is now 0s, meaning "never repeat". The "vacation" command was OK
1524 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1525 (I found it when inspecting the code).
152720. There is now an overall timeout for performing a callout verification. It
1528 defaults to 4 times the callout timeout, which applies to individual SMTP
1529 commands during the callout. The overall timeout applies when there is more
1530 than one host that can be tried. The timeout is checked before trying the
1531 next host. This prevents very long delays if there are a large number of
1532 hosts and all are timing out (e.g. when the network connections are timing
1533 out). The value of the overall timeout can be changed by specifying an
1534 additional sub-option for "callout", called "maxwait". For example:
1536 verify = sender/callout=5s,maxwait=20s
153821. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1539 to the end before writing, but this should make it even safer).
154122. Exim was forgetting that it had advertised PIPELINING for the second and
1542 subsequent messages on an SMTP connection. It was also not resetting its
1543 memory on STARTTLS and an internal HELO.
154523. When Exim logs an SMTP synchronization error within a session, it now
1546 records whether PIPELINING has been advertised or not.
154824. Added 3 instances of "(long int)" casts to time_t variables that were being
1549 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1550 rather than long int.
155225. Installed the latest Cygwin configuration files from the Cygwin maintainer.
155426. Added the never_mail option to autoreply.
1557Exim version 4.41
1560 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1561 crash if the getsockname() call failed; this can happen if a connection is
1562 closed very soon after it is established. The problem was simply in the
1563 order in which certain operations were done, causing Exim to try to write
1564 to the SMTP stream before it had set up the file descriptor. The bug has
1565 been fixed by making things happen in the correct order.
1568Exim version 4.40
1571 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1572 before the connection was closed, thus losing the rejection response.
1574 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1575 some early Solaris releases, but causes trouble in current releases where
1576 socklen_t is defined.
1578 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1579 always exist.
1581 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1582 configured.
1584 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1585 that releasing the top of it at the end releases what was used for sub-
1586 expansions (unless the block got too big). However, discard this block if
1587 the first thing is a variable or header, so that we can use its block when
1588 it is dynamic (useful for very large $message_headers, for example).
1590 6. Lookups now cache *every* query, not just the most recent. A new, separate
1591 store pool is used for this. It can be recovered when all lookup caches are
1592 flushed. Lookups now release memory at the end of their result strings.
1593 This has involved some general refactoring of the lookup sources.
1595 7. Some code has been added to the store_xxx() functions to reduce the amount
1596 of flapping under certain conditions.
1598 8. log_incoming_interface used to affect only the <= reception log lines. Now
1599 it causes the local interface and port to be added to several more SMTP log
1600 lines, for example "SMTP connection from", and rejection lines.
1602 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
160410. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
160611. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1607 could be overwritten at the end of the current message (or the start of a
1608 new message if it was set in a HELO ACL). The value is now preserved for
1609 the duration of the SMTP connection.
161112. If a transport had a headers_rewrite setting, and a matching header line
1612 contained an unqualified address, that address was qualified, even if it
1613 did not match any rewriting rules. The underlying bug was that the values
1614 of the flags that permit the existence of unqualified sender and recipient
1615 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1616 non-local messages, and by -bnq for local messages) were not being
1617 preserved with the message after it was received.
161913. When Exim was logging an SMTP synchronization error, it could sometimes log
1620 "next input=" as part of the text comprising the host identity instead of
1621 the correct text. The code was using the same buffer for two different
1622 strings. However, depending on which order the printing function evaluated
1623 its arguments, the bug did not always show up. Under Linux, for example, my
1624 test suite worked just fine.
162614. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1627 doesn't work with some older versions of Perl. It has been changed to "my",
1628 which in any case is probably the better facility to use.
163015. A really picky compiler found some instances of statements for creating
1631 error messages that either had too many or two few arguments for the format
1632 string.
163416. The size of the buffer for calls to the DNS resolver has been increased
1635 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1636 for addresses that have a lot of PTR records. This alleviates a problem; it
1637 does not fully solve it.
163917. A dnsdb lookup for PTR records that receives more data than will fit in the
1640 buffer now truncates the list and logs the incident, which is the same
1641 action as happens when Exim is looking up a host name and its aliases.
1642 Previously in this situation something unpredictable would happen;
1643 sometimes it was "internal error: store_reset failed".
164518. If a server dropped the connection unexpectedly when an Exim client was
1646 using GnuTLS and trying to read a response, the client delivery process
1647 crashed while trying to generate an error log message.
164919. If a "warn" verb in an ACL added multiple headers to a message in a single
1650 string, for example:
1652 warn message = H1: something\nH2: something
1654 the text was added as a single header line from Exim's point of view
1655 though it ended up OK in the delivered message. However, searching for the
1656 second and subsequent header lines using $h_h2: did not work. This has been
1657 fixed. Similarly, if a system filter added multiple headers in this way,
1658 the routers could not see them.
166020. Expanded the error message when iplsearch is called with an invalid key to
1661 suggest using net-iplsearch in a host list.
166321. When running tests using -bh, any delays imposed by "delay" modifiers in
1664 ACLs are no longer actually imposed (and a message to that effect is
1665 output).
166722. If a "gecos" field in a passwd entry contained escaped characters, in
1668 particular, if it contained a \" sequence, Exim got it wrong when building
1669 a From: or a Sender: header from that name. A second bug also caused
1670 incorrect handling when an unquoted " was present following a character
1671 that needed quoting.
167323. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1674 was not being matched caselessly.
167624. Arranged for all hyphens in the exim.8 source to be escaped with
1677 backslashes.
167925. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1680 properly. Recipient callout cache records were still being keyed to include
1681 the sender, even when use_sender was set false. This led to far more
1682 callouts that were necessary. The sender is no longer included in the key
1683 when use_sender is false.
168526. Added "control = submission" modifier to ACLs.
168727. Added the ${base62d: operator to decode base 62 numbers.
168928. dnsdb lookups can now access SRV records.
169129. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1692 the configuration file.
169430. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1695 (-v) mode. This makes the output for a verbose queue run more intelligible.
169731. Added a use_postmaster feature to recipient callouts.
169932. Added the $body_zerocount variable, containing the number of binary zero
1700 bytes in the message body.
170233. The time of last modification of the "new" subdirectory is now used as the
1703 "mailbox time last read" when there is a quota error for a maildir
1704 delivery.
170634. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
170835. Added +ignore_unknown as a special item in host lists.
171036. Code for decoding IPv6 addresses in host lists is now included, even if
1711 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1712 address was recognized as an IP address, but was then not correctly decoded
1713 into binary, causing unexpected and incorrect effects when compared with
1714 another IP address.
1717Exim version 4.34
1720 1. Very minor rewording of debugging text in manualroute to say "list of
1721 hosts" instead of "hostlist".
1723 2. If verify=header_syntax was set, and a header line with an unqualified
1724 address (no domain) and a large number of spaces between the end of the
1725 name and the colon was received, the reception process suffered a buffer
1726 overflow, and (when I tested it) crashed. This was caused by some obsolete
1727 code that should have been removed. The fix is to remove it!
1729 3. When running in the test harness, delay a bit after writing a bounce
1730 message to get a bit more predictability in the log output.
1732 4. Added a call to search_tidyup() just before forking a reception process. In
1733 theory, someone could use a lookup in the expansion of smtp_accept_max_
1734 per_host which, without the tidyup, could leave open a database connection.
1736 5. Added the variables $recipient_data and $sender_data which get set from a
1737 lookup success in an ACL "recipients" or "senders" condition, or a router
1738 "senders" option, similar to $domain_data and $local_part_data.
1740 6. Moved the writing of debug_print from before to after the "senders" test
1741 for routers.
1743 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1744 problems for message scanning, either using a data ACL, or using
1745 local_scan() because the Received: header was not generated till after they
1746 were called (in order to set the time as the time of reception completion).
1747 I have revised the way this works. The header is now generated after the
1748 body is received, but before the ACL or local_scan() are called. After they
1749 are run, the timestamp in the header is updated.
1752Exim version 4.33
1755 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1756 before starting a queue runner without re-exec. This happened only when
1757 deliver_drop_privilege was set or when the Exim user was set to root. The
1758 effect of the bug was that timeouts during subsequent deliveries caused
1759 crashes instead of being properly handled. The handler is now left at its
1760 default (and expected) setting.
1762 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1763 message, again when deliver_drop_privilege is set or Exim is run as root.
1764 The bug described in (1) was not present in this case, but the tidying up
1765 of the other signals was missing. I have made the two cases consistent.
1767 3. The ignore_target_hosts setting on a manualroute router was being ignored
1768 for hosts that were looked up using the /MX notation.
1770 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1771 in domain lists.
1773 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1774 operated on the sender address. After changing the $sender_address to <>
1775 for the sender address verify, Exim was re-instated it as the original
1776 (before rewriting) address, but remembering that it had rewritten it, so it
1777 wasn't rewriting it again. This bug also had the effect of breaking the
1778 sender address verification caching when the sender address was rewritten.
1780 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1781 This has been changed so that if the ip literal address matches
1782 ignore_target_hosts, the router declines.
1784 7. Added expansion conditions match_domain, match_address, and match_local_
1785 part (NOT match_host).
1787 8. The placeholder for the Received: header didn't have a length field set.
1789 9. Added code to Exim itself and to exim_lock to test for a specific race
1790 condition that could lead to file corruption when using MBX delivery. The
1791 issue is with the lockfile that is created in /tmp. If this file is removed
1792 after a process has opened it but before that process has acquired a lock,
1793 there is the potential for a second process to recreate the file and also
1794 acquire a lock. This could lead to two Exim processes writing to the file
1795 at the same time. The added code performs the same test as UW imapd; it
1796 checks after acquiring the lock that its file descriptor still refers to
1797 the same named file.
179910. The buffer for building added header lines was of fixed size, 8192 bytes.
1800 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1801 when Exim is built.
180311. Added the smtp_active_hostname option. If used, this will typically be made
1804 to depend on the incoming interface address. Because $interface_address is
1805 not set up until the daemon has forked a reception process, error responses
1806 that can happen earlier (such as "too many connections") no longer contain
1807 a host name.
180912. If an expansion in a condition on a "warn" statement fails because a lookup
1810 defers, the "warn" statement is abandoned, and the next ACL statement is
1811 processed. Previously this caused the whole ACL to be aborted.
181313. Added the iplsearch lookup type.
181514. Added ident_timeout as a log selector.
181715. Added tls_certificate_verified as a log selector.
181916. Added a global option tls_require_ciphers (compare the smtp transport
1820 option of the same name). This controls incoming TLS connections.
182217. I finally figured out how to make tls_require_ciphers do a similar thing
1823 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1824 before starting the TLS session.
182618. Tabs are now shown as \t in -bP output.
182819. If the log selector return_path_on_delivery was set, Exim crashed when
1829 bouncing a message because it had too many Received: header lines.
183120. If two routers both had headers_remove settings, and the first one included
1832 a superfluous trailing colon, the final name in the first list and the
1833 first name in the second list were incorrectly joined into one item (with a
1834 colon in the middle).
1837Exim version 4.32
1840 1. Added -C and -D options to the exinext utility, mainly to make it easier
1841 to include in the automated testing, but these could be helpful when
1842 multiple configurations are in use.
1844 2. The exinext utility was not formatting the output nicely when there was
1845 an alternate port involved in the retry record key, nor when there was a
1846 message id as well (for retries that were specific to a specific message
1847 and a specific host). It was also confused by IPv6 addresses, because of
1848 the additional colons they contain. I have fixed the IPv4 problem, and
1849 patched it up to do a reasonable job for IPv6.
1851 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1852 delivery, the log line now contains "pipelined" if PIPELINING was used.
1854 4. An SMTP transport process used to panic and die if the bind() call to set
1855 an explicit outgoing interface failed. This has been changed; it is now
1856 treated in the same way as a connect() failure.
1858 5. A reference to $sender_host_name in the part of a conditional expansion
1859 that was being skipped was still causing a DNS lookup. This no longer
1860 occurs.
1862 6. The def: expansion condition was not recognizing references to header lines
1863 that used bh_ and bheader_.
1865 7. Added the _cache feature to named lists.
1867 8. The code for checking quota_filecount in the appendfile transport was
1868 allowing one more file than it should have been.
1870 9. For compatibility with Sendmail, the command line option
1872 -prval:sval
1874 is equivalent to
1876 -oMr rval -oMs sval
1878 and sets the incoming protocol and host name (for trusted callers). The
1879 host name and its colon can be omitted when only the protocol is to be set.
1880 Note the Exim already has two private options, -pd and -ps, that refer to
1881 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1882 "s", but I don't think that's a major issue.
188410. A number of refactoring changes to the code, none of which should affect
1885 Exim's behaviour:
1887 (a) The number of logging options was getting close to filling up the
1888 32-bit word that was used as a bit map. I have split them into two classes:
1889 those that are passed in the argument to log_write(), and those that are
1890 only ever tested independently outside of that function. These are now in
1891 separate 32-bit words, so there is plenty of room for expansion again.
1892 There is no change in the user interface or the logging behaviour.
1894 (b) When building, for example, log lines, the code previously used a
1895 macro that called string_cat() twice, in order to add two strings. This is
1896 not really sufficiently general. Furthermore, there was one instance where
1897 it was actually wrong because one of the argument was used twice, and in
1898 one call a function was used. (As it happened, calling the function twice
1899 did not affect the overall behaviour.) The macro has been replaced by a
1900 function that can join an arbitrary number of extra strings onto a growing
1901 string.
1903 (c) The code for expansion conditions now uses a table and a binary chop
1904 instead of a serial search (which was left over from when there were very
1905 few conditions). Also, it now recognizes conditions like "pam" even when
1906 the relevant support is not compiled in: a suitably worded error message is
1907 given if an attempt is made to use such a condition.
190911. Added ${time_interval:xxxxx}.
191112. A bug was causing one of the ddress fields not to be passed back correctly
1912 from remote delivery subprocesses. The field in question was not being
1913 subsequently used, so this caused to problems in practice.
191513. Added new log selectors queue_time and deliver_time.
191714. Might have fixed a bug in maildirsizefile handling that threw up
1918 "unexpected character" debug warnings, and recalculated the data
1919 unnecessarily. In any case, I expanded the warning message to give more
1920 information.
192215. Added the message "Restricted characters in address" to the statements in
1923 the default ACL that block characters like @ and % in local parts.
192516. Change 71 for release 4.31 proved to be much less benign that I imagined.
1926 Three changes have been made:
1928 (a) There was a serious bug; a negative response to MAIL caused the whole
1929 recipient domain to be cached as invalid, thereby blocking all messages
1930 to all local parts at the same domain, from all senders. This bug has
1931 been fixed. The domain is no longer cached after a negative response to
1932 MAIL if the sender used is not empty.
1934 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1935 been restored.
1937 (c) A new callout option, "use_sender" has been added for people who want
1938 the modified behaviour.
1941Exim version 4.31
1944 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1945 Larry Rosenman.
1947 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1948 indeed breaks things for older releases.
1950 3. Added additional logging to the case where there is a problem reading data
1951 from a filter that is running in a subprocess using a pipe, in order to
1952 try to track down a specific problem.
1954 4. Testing facility fudge: when running in the test harness and attempting
1955 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1956 getting "No route to host". Convert this to a timeout.
1958 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1959 warning.
1961 6. Some OS don't have socklen_t but use size_t instead. This affects the
1962 fifth argument of getsockopt() amongst other things. This is now
1963 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1964 can be set for individual OS. I have set it for SunOS5, OSF1, and
1965 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1966 some earlier ones do not.
1968 7. Change 4.30/15 was not doing the test caselessly.
1970 8. The standard form for an IPv6 address literal was being rejected by address
1971 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1972 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1973 this, as well as the form without the "IPv6" on the front (but only when
1974 address literals are enabled, of course).
1976 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
197810. Exim crashed if a message with an empty sender address specified by -f
1979 encountered a router with an errors_to setting. This could be provoked only
1980 by a command such as
1982 exim -f "" ...
1984 where an empty string was supplied; "<>" did not hit this bug.
198611. Installed PCRE release 4.5.
198812. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1989 remained set. It is now erased.
199113. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1992 times from message ids (which are base 36 rather than the normal 62).
199414. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1995 were being counted as actual protocol errors, and logged if the log
1996 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1997 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1998 and DATA following a set of rejected RCPTs do not count as protocol errors.
1999 In other words, Exim assumes they were pipelined, though this may not
2000 actually be the case. Of course, in all cases the client gets an
2001 appropriate error code.
200315. If a lookup fails in an ACL condition, a message about the failure may
2004 be available; it is used if testing the ACL cannot continue, because most
2005 such messages specify what the cause of the deferral is. However, some
2006 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2007 that caused an old message to be retained and used if a later statement
2008 caused a defer, replacing the real cause of the deferral.
201016. If an IP address had so many PTR records that the DNS lookup buffer
2011 was not large enough to hold them, Exim could crash while trying to process
2012 the truncated data. It now detects and logs this case.
201417. Further to 4.21/58, another change has been made: if (and only if) the
2015 first line of a message (the first header line) ends with CRLF, a bare LF
2016 in a subsequent header line has a space inserted after it, so as not to
2017 terminate the header.
201918. Refactoring: tidied an ugly bit of code in appendfile that copied data
2020 unnecessarily, used atoi() instead of strtol(), and didn't check the
2021 termination when getting file sizes from file names by regex.
202319. Completely re-implemented the support for maildirsize files, in the light
2024 of a number of problems with the previous contributed implementation
2025 (4.30/29). In particular:
2027 . If the quota is zero, the maildirsize file is maintained, but no quota is
2028 imposed.
2030 . If the maildir directory does not exist, it is created before any attempt
2031 to write a maildirsize file.
2033 . The quota value in the file is just a cache; if the quota is changed in
2034 the transport, the new value overrides.
2036 . A regular expression is available for excluding directories from the
2037 count.
203920. The autoreply transport checks the characters in options that define the
2040 message's headers; it allows continued headers, but it was checking with
2041 isspace() after an embedded newline instead of explicitly looking for a
2042 space or a tab.
204421. If all the "regular" hosts to which an address was routed had passed their
2045 expiry times, and had not reached their retry times, the address was
2046 bounced, even if fallback hosts were defined. Now Exim should go on to try
2047 the fallback hosts.
204922. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2050 equivalent code in the SMTP transport. Some hosts send humungous responses
2051 to HELO/EHLO, more than 1024 it seems.
205323. Refactoring: code in filter.c used (void *) for "any old type" but this
2054 gives compiler warnings in some environments. I've now done it "properly",
2055 using a union.
205724. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2058 (because of problems with the built-in one) was declared to return uschar *
2059 instead of char *, causing compiler failure.
206125. Fixed a file descriptor leak when processing alias/forward files.
206326. Fixed a minor format string issue in dbfn.c.
206527. Typo in exim.c: ("dmbnz" for "dbmnz").
206728. If a filter file refered to $h_xxx or $message_headers, and the headers
2068 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2069 become corrupted.
207129. When a sender address is verified, it is cached, to save repeating the test
2072 when there is more than one recipient in a message. However, when the
2073 verification involves a callout, it is possible for different callout
2074 options to be set for different recipients. It is too complicated to keep
2075 track of this in the cache, so now Exim always runs a verification when a
2076 callout is required, relying on the callout cache for the optimization.
2077 The overhead is duplication of the address routing, but this should not be
2078 too great.
208030. Fixed a bug in callout caching. If a RCPT command caused the sender address
2081 to be verified with callout=postmaster, and the main callout worked but the
2082 postmaster check failed, the verification correctly failed. However, if a
2083 subsequent RCPT command asked for sender verification *without* the
2084 postmaster check, incorrect caching caused this verification also to fail,
2085 incorrectly.
208731. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2088 it was not caching the DNS options (qualify_single, search_parents) that
2089 were used when the lookup failed. A subsequent lookup with different
2090 options therefore always gave the same answer, though there were cases
2091 where it should not have. (Example: a "domains = !$mx_any" option on a
2092 dnslookup router: the "domains" option is always processed without any
2093 widening, but the router might have qualify_single set.) Now Exim uses the
2094 cached value only when the same options are set.
209632. Added John Jetmore's "exipick" utility to the distribution.
209833. GnuTLS: When an attempt to start a TLS session fails for any reason other
2099 than a timeout (e.g. a certificate is required, and is not provided), an
2100 Exim server now closes the connection immediately. Previously it waited for
2101 the client to close - but if the client is SSL, it seems that they each
2102 wait for each other, leading to a delay before one of them times out.
210434: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2105 maintained 0.8.x compatibility because I don't think many are using it, and
2106 it is clearly obsolete.
210835. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2109 transport.
211136. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2112 client certificate was expired. A simple patch fixes this, though I don't
2113 understand the full logic of why the verify callback is called multiple
2114 times.
211637. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2117 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2118 which causes problems with some clients (such as the Certicom SSL Plus
2119 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2120 disables the coutermeasure allowing Eudora to connect."
212238. Exim was not checking that a write() to a log file succeeded. This could
2123 lead to Bad Things if a log got too big, in particular if it hit a file
2124 size limit. Exim now panics and dies if it cannot write to a log file, just
2125 as it does if it cannot open a log file.
212739. Modified OS/Makefile-Linux so that it now contains
2131 The two -D definitions ensure that Exim is compiled with large file
2132 support, which makes it possible to handle log files that are bigger than
2133 2^31.
213540. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2136 instance) a domain was checked against a named list that involved a lookup,
2137 causing $domain_data to be set, then another domain was checked against the
2138 same list, then the first domain was re-checked, the value of $domain_data
2139 after the final check could be wrong. In particular, if the second check
2140 failed, it could be set empty. This bug probably also applied to
2141 $localpart_data.
214341. The strip_trailing_dot option was not being applied to the address given
2144 with the -f command-line option.
214642. The code for reading a message's header from the spool was incrementing
2147 $received_count, but never initializing it. This meant that the value was
2148 incorrect (doubled) while delivering a message in the same process in which
2149 it was received. In the most common configuration of Exim, this never
2150 happens - a fresh exec is done - but it can happen when
2151 deliver_drop_privilege is set.
215343. When Exim logs an SMTP synchronization error - client data sent too soon -
2154 it now includes up to 150 characters of the unexpected data in the log
2155 line.
215744. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2158 and building data strings. The size of both of these buffers was 10 000
2159 bytes - far larger than anybody would *ever* want, thought I. Needless to
2160 say, somebody hit the limit. I have increased the maximum line length to
2161 20 000 and the maximum data length of concatenated lines to 100 000. I have
2162 also fixed two bugs, because there was no checking on these buffers. Tsk,
2163 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2164 buffer is too small.
216645. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2167 lsearch lookups. Now it does.
216946. When parsing a route_list item in a manualroute router, a fixed-length
2170 buffer was used for the list of hosts. I made this 1024 bytes long,
2171 thinking that nobody would ever have a list of hosts that long. Wrong.
2172 Somebody had a whole pile of complicated expansion conditions, and the
2173 string was silently truncated, leading to an expansion error. It turns out
2174 that it is easier to change to an unlimited length (owing to other changes
2175 that have happened since this code was originally written) than to build
2176 structure for giving a limitation error. The length of the item that
2177 expands into the list of hosts is now unlimited.
217947. The lsearch lookup could not handle data where the length of text line was
2180 more than 4095 characters. Such lines were truncated, leading to shortened
2181 data being returned. It should now handle lines of any length.
218348. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2184 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2185 ACL").
218749. Cosmetic tidy to scripts like exicyclog that are generated by globally
2188 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2189 no longer happens in comment lines. A list of replacements is now placed
2190 at the head of all of the source files, except those whose only change is
2191 to replace PERL_COMMAND in the very first #! line.
219350. Replaced the slow insertion sort in queue.c, for sorting the list of
2194 messages on the queue, with a bottom-up merge sort, using code contributed
2195 by Michael Haardt. This should make operations like -bp somewhat faster on
2196 large queues. It won't affect queue runners, except when queue_run_in_order
2197 is set.
219951. Installed eximstats 1.31 in the distribution.
220152. Added support for SRV lookups to the dnslookup router.
220353. If an ACL referred to $message_body or $message_body_end, the value was not
2204 reset for any messages that followed in the same SMTP session.
220654. The store-handling optimization for building very long strings was not
2207 differentiating between the different store pools. I don't think this
2208 actually made any difference in practice, but I've tidied it.
221055. While running the routers to verify a sender address, $sender_address
2211 was still set to the sender address. This is wrong, because when routing to
2212 send a bounce to the sender, it would be empty. Therefore, I have changed
2213 it so that, while verifying a sender address, $sender_address is set to <>.
2214 (There is no change to what happens when verifying a recipient address.)
221656. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2217 target A or AAAA records (if not already returned) without resetting the
2218 qualify_single or search_parents options of the DNS resolver. These are
2219 inappropriate in this case because the targets of MX and SRV records must
2220 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2221 target that, when qualified, matched something in the local domain. These
2222 two options are now turned off when doing these lookups.
222457. It seems that at least some releases of Reiserfs (which does not have the
2225 concept of a fixed number of inodes) returns zero and not -1 for the
2226 number of available inodes. This interacted badly with check_spool_inodes,
2227 which assumed that -1 was the "no such thing" setting. What I have done is
2228 to check that the total number of inodes is greater than zero before doing
2229 the test of how many are available.
223158. When a "warn" ACL statement has a log_message modifier, the message is
2232 remembered, and not repeated. This is to avoid a lot of repetition when a
2233 message has many recipients that cause the same warning to be written.
2234 Howewer, Exim was preserving the list of already written lines for an
2235 entire SMTP session, which doesn't seem right. The memory is now reset if a
2236 new message is started.
223859. The "rewrite" debugging flag was not showing the result of rewriting in the
2239 debugging output unless log_rewrite was also set.
224160. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2242 of (int)(handle) when we know that handle contains (void *)(-1).
224461. The Exim daemon panic-logs an error return when it closes the incoming
2245 connection. However "connection reset by peer" seems to be common, and
2246 isn't really an error worthy of noting specially, so that particular error
2247 is no long logged.
224962. When Exim is trying to find all the local interfaces, it used to panic and
2250 die if the ioctl to get the interface flags failed. However, it seems that
2251 on at least one OS (Solaris 9) it is possible to have an interface that is
2252 included in the list of interfaces, but for which you get a failure error
2253 for this call. This happens when the interface is not "plumbed" into a
2254 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2255 failure of the "get flags" call assumes that the interface is down.
225763. Added a ${eval10: operator, which assumes all numbers are decimal. This
2258 makes life easier for people who are doing arithmetic on fields extracted
2259 from dates, where you often get leading zeros that should not be
2260 interpreted as octal.
226264. Added qualify_domain to the redirect router, to override the global
2263 setting.
226565. If a pathologically long header line contained very many addresses (the
2266 report of this problem mentioned 10 000) and each of them was rewritten,
2267 Exim could use up a very large amount of memory. (It kept on making new
2268 copies of the header line as it rewrote, and never released the old ones.)
2269 At the expense of a bit more processing, the header rewriting function has
2270 been changed so that it no longer eats memory in this way.
227266. The generation of the Received: header has been moved from the time that a
2273 message starts to be received, to the time that it finishes. The timestamp
2274 in the Received: header should now be very close to that of the <= log
2275 line. There are two side-effects of this change:
2277 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2278 logged header lines no longer include the local Received: line, because
2279 it has not yet been created. The same applies to a copy of the message
2280 that is returned to a non-SMTP sender when a message is rejected.
2282 (b) When a filter file is tested using -bf, no additional Received: header
2283 is added to the test message. After some thought, I decided that this
2284 is a bug fix.
2286 This change does not affect the value of $received_for. It is still set
2287 after address rewriting, but before local_scan() is called.
228967. Installed the latest Cygwin-specific files from the Cygwin maintainer.
229168. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2292 gave an unhelpful panic error message, and a defer error. I have managed to
2293 change this behaviour so that it now rejects any supplied certificate,
2294 which seems right, as the list of acceptable certificates is empty.
229669. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2297 gave an unhelpful defer error. I have not managed to make this reject any
2298 supplied certificates, but the error message it gives is "no certificate
2299 supplied", which is not helpful.
230170. exigrep's output now also includes lines that are not associated with any
2302 message, but which match the given pattern. Implemented by a patch from
2303 Martin Sluka, which also tidied up the Perl a bit.
230571. Recipient callout verification, like sender verification, was using <> in
2306 the MAIL FROM command. This isn't really the right thing, since the actual
2307 sender may affect whether the remote host accepts the recipient or not. I
2308 have changed it to use the actual sender in the callout; this means that
2309 the cache record is now keyed on a recipient/sender pair, not just the
2310 recipient address. There doesn't seem to be a real danger of callout loops,
2311 since a callout by the remote host to check the sender would use <>.
2312 [SEE ABOVE: changed after hitting problems.]
231472. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2315 temporary errors. However, in the case of such a code being given after
2316 the end of a data transmission (i.e. after ".") Exim was failing to write
2317 a retry record for the message. (Yes, there was some broken host that was
2318 actually sending 8xx at this point.)
232073. An unknown lookup type in a host list could cause Exim to panic-die when
2321 the list was checked. (An example that provoked this was putting <; in the
2322 middle of a list instead of at the start.) If this happened during a DATA
2323 ACL check, a -D file could be left lying around. This kind of configuration
2324 error no longer causes Exim to die; instead it causes a defer errror. The
2325 incident is still logged to the main and panic logs.
232774. Buglet left over from Exim 3 conversion. The message "too many messages
2328 in one connection" was written to the rejectlog but not the mainlog, except
2329 when address rewriting (yes!) was being logged.
233175. Added write_rejectlog option.
233376. When a system filter was run not as root (that is, when system_filter_user
2334 was set), the values of the $n variables were not being returned to the
2335 main process; thus, they were not subsequently available in the $sn
2336 variables.
233877. Added +return_path_on_delivery log selector.
234078. A connection timeout was being treated differently from recipients deferred
2341 when testing hosts_max_try with a message that was older than the host's
2342 retry timeout. (The host should not be counted, thus allowing all hosts to
2343 be tried at least once before bouncing.) This may have been the cause of an
2344 occasionally reported bug whereby a message would remain on the queue
2345 longer than the retry timeout, but would be bounced if a delivery was
2346 forced. I say "may" because I never totally pinned down the problem;
2347 setting up timeout/retry tests is difficult. See also the next item.
234979. The ultimate address timeout was not being applied to errors that involved
2350 a combination of host plus message (for example, a timeout on a MAIL
2351 command). When an address resolved to a number of possible hosts, and they
2352 were not all tried for each delivery (e.g. because of hosts_max_try), a
2353 message could remain on the queue longer than the retry timeout.
235580. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2356 Haardt.
235881. Fixed an obscure SMTP outgoing bug which required at least the following
2359 conditions: (a) there was another message waiting for the same server;
2360 (b) the server returned 5xx to all RCPT commands in the first message so
2361 that the message was not completed; (c) the server dropped the connection
2362 or gave a negative response to the RSET that Exim sends to abort the
2363 transaction. The observed case was a dropped connection after DATA that had
2364 been sent in pipelining mode. That is, the server had advertised PIPELINING
2365 but was not implementing it correctly. The effect of the bug was incorrect
2366 behaviour, such as trying another host, and this could lead to a crash.
2369Exim version 4.30
2372 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2373 and daemon.c were passed as pointers to ints; they should have been
2374 pointers to socklen_t variables (which are typically unsigned ints).
2376 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2377 fixed.
2379 3. Fixed a really odd bug that affected only the testing scheme; patching a
2380 certain fixed string in the binary changed the value of another string that
2381 happened to be identical to the end of the original first string.
2383 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2384 name", it returns that address as the IP address. On some operating
2385 systems (e.g. Solaris), it also passes back the IP address string as the
2386 "host name". However, on others (e.g. Linux), it passes back an empty
2387 string. Exim wasn't checking for this, and was changing the host name to an
2388 empty string, assuming it had been canonicized.
2390 5. Although rare, it is permitted to have more than one PTR record for a given
2391 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2392 all the names associated with an address, because they do in Solaris.
2393 However, it seems that they do not in Linux for data that comes from the
2394 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2395 I found this out when I moved to a new Linux workstation and tried to run
2396 the Exim test suite.
2398 To get round this problem I have changed the code so that it now does its
2399 own call to the DNS to look up PTR records when searching for a host name.
2400 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2401 addresses that are only in /etc/hosts are still found.
2403 This behaviour is, however, controlled by an option called host_lookup_
2404 order, which defaults to "bydns:byaddr". If people want to use the other
2405 order, or indeed, just use one or the other means of lookup, they can
2406 specify it in this variable.
2408 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2409 some operating systems, this comes back from gethostbyaddr() as an empty
2410 string, and this is what Exim used to test for. However, it seems that in
2411 other systems, "." is yielded. Exim now tests for this case too.
2413 7. The values of check_spool_space and check_log_space are now held internally
2414 as a number of kilobytes instead of an absolute number of bytes. If a
2415 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2416 kilobyte. This means that much larger values can be stored.
2418 8. Exim monitor: an attempt to get the action menu when not actually pointing
2419 at a message produces an empty menu entitled "No message selected". This
2420 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2421 no entries in it ("Shell widget menu has zero width and/or height"). So I
2422 have added a single, blank menu entry in this case.
2424 9. Added ${quote_local_part.
242610. MIME decoding is now applied to the contents of Subject: header lines when
2427 they are logged.
242911. Now that a reference to $sender_host_address automatically causes a reverse
2430 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2431 host lookup before query-style lookups in lists that might use this
2432 variable. This has therefore been abolished, and the "net-" prefix is no
2433 longer necessary for query-style lookups.
243512. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2436 have been a typo for LFLAGS, so it has been changed.
243813. The install script calls Exim with "-C /dev/null" in order to find the
2439 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2440 to be output. Howeve, since Exim outputs its version number before the
2441 error, it didn't break the script. It just looked ugly. I fixed this by
2442 always allowing "-C /dev/null" if the caller is root.
244414. Ignore overlarge ACL variable number when reading spool file - insurance
2445 against a later release with more variables having written the file.
244715. The standard form for an IPv6 address literal was being rejected by EHLO.
2448 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2449 this, as well as the form without the "IPv6" on the front.
245116. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2452 OS/Makefile-Darwin file.
245417. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2455 only to LDAP libraries that do not have LDAP_OPT_DEREF.
245718. After change 4.21/52, "%ld" was used to format the contents of the $inode
2458 variable. However, some OS use ints for inodes. I've added cast to long int
2459 to get rid of the compiler warning.
246119. I had forgotten to lock out "/../" in configuration file names when
2462 ALT_CONFIG_PREFIX was set.
246420. Routers used for verification do not need to specify transports. However,
2465 if such a router generated a host list, and callout was configured, Exim
2466 crashed, because it could not find a port number from the (non-existent)
2467 transport. It now assumes port 25 in this circumstance.
246921. Added the -t option to exigrep.
247122. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2472 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2473 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2474 them is set, LOOKUP_LSEARCH is forced.
247623. "exim -bV" now outputs a list of lookups that are included in the binary.
247824. Added sender and host information to the "rejected by local_scan()" log
2479 line; previously there was no indication of these.
248125. Added .include_if_exists.
248326. Change 3.952/11 added an explicit directory sync on top of a file sync for
2484 Linux. It turns out that not all file systems support this. Apparently some
2485 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2486 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2487 not supported on the file descriptor, is now ignored when Exim is trying to
2488 sync a directory. This applies only to Linux.
249027. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
249228. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2493 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2494 this.
249629. Added support for maildirsize files from supplied patch (modified a bit).
249830. The use of :fail: followed by an empty string could lead Exim to respond to
2499 sender verification failures with (e.g.):
2501 550 Verification failed for <xxx>
2502 550 Sender verify failed
2504 where the first response line was missing the '-' that indicates it is not
2505 the final line of the response.
250731. The loop for finding the name of the user that called Exim had a hardwired
2508 limit of 10; it now uses the value of finduser_retries, which is used for
2509 all other user lookups.
251132. Added $received_count variable, available in data and not_smtp ACLs, and at
2512 delivery time.
251433. Exim was neglecting to zero errno before one call of strtol() when
2515 expanding a string and expecting an integer value. On some systems this
2516 resulted in spurious "integer overflow" errors. Also, it was casting the
2517 result into an int without checking.
251934. Testing for a connection timeout using "timeout_connect" in the retry rules
2520 did not work. The code looks as if it has *never* worked, though it appears
2521 to have been documented since at least releast 1.62. I have made it work.
252335. The "timeout_DNS" error in retry rules, also documented since at least
2524 1.62, also never worked. As it isn't clear exactly what this means, and
2525 clearly it isn't a major issue, I have abolished the feature by treating it
2526 as "timeout", and writing a warning to the main and panic logs.
252836. The display of retry rules for -brt wasn't always showing the error code
2529 correctly.
253137. Added new error conditions to retry rules: timeout_A, timeout_MX,
2532 timeout_connect_A, timeout_connect_MX.
253438. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2535 to the empty sender.
253739. The daemon was not analysing the content of -oX till after it had closed
2538 stderr and disconnected from the controlling terminal. This meant that any
2539 syntax errors were only noted on the panic log, and the return code from
2540 the command was 0. By re-arranging the code a little, I've made the
2541 decoding happen first, so such errors now appear on stderr, and the return
2542 code is 1. However, the actual setting up of the sockets still happens in
2543 the disconnected process, so errors there are still only recorded on the
2544 panic log.
254640. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2547 connections (as happens on some IP stacks) was logged at start up time as
2548 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2549 it from "IPv6 and IPv4", which means that two separate sockets are being
2550 used.
255241. The debug output for gethostbyname2() or getipnodebyname() failures now
2553 says whether AF_INET or AF_INET6 was passed as an argument.
255542. Exiwhat output was messed up when time zones were included in log
2556 timestamps.
255843. Exiwhat now gives more information about the daemon's listening ports,
2559 and whether -tls-on-connect was used.
256144. The "port" option of the smtp transport is now expanded.
256345. A "message" modifier in a "warn" statement in a non-message ACL was being
2564 silently ignored. Now an error message is written to the main and panic
2565 logs.
256746. There's a new ACL modifier called "logwrite" which writes to a log file
2568 as soon as it is encountered.
257047. Added $local_user_uid and $local_user_gid at routing time.
257248. Exim crashed when trying to verify a sender address that was being
2573 rewritten to "<>".
257549. Exim was recognizing only a space character after ".include". It now also
2576 recognizes a tab character.
257850. Fixed several bugs in the Perl script that creates the exim.8 man page by
2579 extracting the relevant information from the specification. The man page no
2580 longer contains scrambled data for the -d option, and I've added a section
2581 at the front about calling Exim under different names.
258351. Added "extra_headers" argument to the "mail" command in filter files.
258552. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2586 crash.
258853. Installed eximstats 1.29.
259054. Added transport_filter_timeout as a generic transport option.
259255. Exim no longer adds an empty Bcc: header to messages that have no To: or
2593 Cc: header lines. This was required by RFC 822, but it not required by RFC
2594 2822.
259656. Exim used to add From:, Date:, and Message-Id: header lines to any
2597 incoming messages that did not have them. Now it does so only if the
2598 message originates locally, that is, if there is no associated remote host
2599 address. When Resent- header lines are present, this applies to the Resent-
2600 lines rather than the non-Resent- lines.
260257. Drop incoming SMTP connection after too many syntax or protocol errors. The
2603 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
260558. Messages for configuration errors now include the name of the main
2606 configuration file - useful now that there may be more than one file in a
2607 list (.included file names were always shown).
260959. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2610 for those rare installations that do not start the daemon as root or run it
2611 setuid root. I've cut out the call to initgroups() if the daemon is not
2612 root at that time.
261460. The Exim user and group can now be bound into the binary as text strings
2615 that are looked up at the start of Exim's processing.
261761. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
261962. Added $mailstore_basename variable.
262163. Installed patch to sieve.c from Michael Haardt.
262364. When Exim failed to open the panic log after failing to open the main log,
2624 the original message it was trying to log was written to stderr and debug
2625 output, but if they were not available (the usual case in production), it
2626 was lost. Now it is written to syslog before the two lines that record the
2627 failures to open the logs.
262965. Users' Exim filters run in subprocesses under the user's uid. It is
2630 possible for a "deliver" command or an alias in a "personal" command to
2631 provoke an address rewrite. If logging of address rewriting is configured,
2632 this fails because the process is not running as root or exim. There may be
2633 a better way of dealing with this, but for the moment (because 4.30 needs
2634 to be released), I have disabled address rewrite logging when running a
2635 filter in a non-root, non-exim process.
2638Exim version 4.24
2641 1. The buildconfig auxiliary program wasn't quoting the value set for
2642 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2643 not defined. This bug was masked in 4.22 by the effect that was fixed in
2644 change 4.23/1.
2646 2. Some messages that were rejected after a message id was allocated were
2647 shown as "incomplete" by exigrep. It no longer does this for messages that
2648 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2650 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2651 have allow_domain_literals set, the ID did not get logged in the <= line.
2652 Domain literals are now always recognized in Message-ID: header lines.
2654 4. The first argument for a ${extract expansion item is the key name or field
2655 number. Leading and trailing spaces in this item were not being ignored,
2656 causing some misleading effects.
2658 5. When deliver_drop_privilege was set, single queue runner processes started
2659 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2660 same command in the process it spins off) were not dropping privilege.
2662 6. When the daemon running as "exim" started a queue runner, it always
2663 re-executed Exim in the spun-off process. This is a waste of effort when
2664 deliver_drop_privilege is set. The new process now just calls the
2665 queue-runner function directly.
2668Exim version 4.23
2671 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2674 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2675 ignored. Though the use of -oP was forcing the writing of a pid file, it
2676 was always written to the default place.
2678 3. If the message "no IP address found for host xxxx" is generated during
2679 incoming verification, it is now followed by identification of the incoming
2680 connection (so you can more easily find what provoked it).
2682 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2684 5. Added some features to "harden" Exim a bit more against certain attacks:
2686 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2687 be put in Local/Makefile. This is like the never_users runtime option,
2688 but it cannot be overridden. The default setting is "root".
2690 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2691 prefix string with which any file named in a -C command line option
2692 must start.
2694 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2695 is retained for -C and -D only if the caller of Exim is root. Without
2696 it, the exim user may also use -C and -D and retain privilege.
2698 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2699 command line option is disabled.
2701 6. Macro names set by the -D option must start with an upper case letter, just
2702 like macro names defined in the configuration file.
2704 7. Added "dereference=" facility to LDAP.
2706 8. Two instances of the typo "uknown" in the source files are fixed.
2708 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2709 the Configure-Makefile script screwed up while processing it.
271110. Incorporated PCRE 4.4.
271311. The SMTP synchronization check was not operating right at the start of an
2714 SMTP session. For example, it could not catch a HELO sent before the client
2715 waited for the greeting. There is now a check for outstanding input at the
2716 point when the greeting is written. Because of the duplex, asynchronous
2717 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2718 way, but not yet received, when the check is performed.
272012. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2721 on TCP/IP sockets, because this apparently causes some broken clients to
2722 timeout.
272413. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2725 unchanged) from the Cygwin maintainer.
272714. The code for -bV that shows what is in the binary showed "mbx" when maildir
2728 was supported instead of testing for mbx. Effectively a typo.
273015. The spa authenticator server code was not checking that the input it
2731 received was valid base64.
273316. The debug output line for the "set" modifier in ACLs was not showing the
2734 name of the variable that was being set.
273617. Code tidy: the variable type "vtype_string" was never used. Removed it.
273818. Previously, a reference to $sender_host_name did not cause a DNS reverse
2739 lookup on its own. Something else was needed to trigger the lookup. For
2740 example, a match in host_lookup or the need for a host name in a host list.
2741 Now, if $sender_host_name is referenced and the host name has not yet been
2742 looked up, a lookup is performed. If the lookup fails, the variable remains
2743 empty, and $host_lookup_failed is set to "1".
274519. Added "eqi" as a case-independent comparison operator.
274720. The saslauthd authentication condition could segfault if neither service
2748 nor realm was specified.
275021. If an overflowing value such as "2048M" was set for message_size_limit, the
2751 error message that was logged was misleading, and incoming SMTP
2752 connections were dropped. The message is now more accurate, and temporary
2753 errors are given to SMTP connections.
275522. In some error situations (such as 21 above) Exim rejects all SMTP commands
2756 (except RSET) with a 421 error, until QUIT is received. However, it was
2757 failing to send a response to QUIT.
275923. The HELO ACL was being run before the code for helo_try_verify_hosts,
2760 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2761 ACL is now run after the helo_try_verify_hosts code.
276324. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2764 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2765 case-independent, so other case variants are also recognized). Apparently
2766 some systems use these upper case variants.
276825. If more than two messages were waiting for the same host, and a transport
2769 filter was specified for the transport, Exim sent two messages over the
2770 same TCP/IP connection, and then failed with "socket operation on non-
2771 socket" when it tried to send the third.
277326. Added Exim::debug_write and Exim::log_write for embedded Perl use.
277527. The extern definition of crypt16() in expand.c was not being excluded when
2776 the OS had its own crypt16() function.
277828. Added bounce_return_body as a new option, and bounce_return_size_limit
2779 as a preferred synonym for return_size_limit, both as an option and as an
2780 expansion variable.
278229. Added LIBS=-liconv to OS/Makefile-OSF1.
278430. Changed the default configuration ACL to relax the local part checking rule
2785 for addresses that are not in any local domains. For these addresses,
2786 slashes and pipe symbols are allowed within local parts, but the sequence
2787 /../ is explicitly forbidden.
278931. SPA server authentication was not clearing the challenge buffer before
2790 using it.
279232. log_message in a "warn" ACL statement was writing to the reject log as
2793 well as to the main log, which contradicts the documentation and doesn't
2794 seem right (because no rejection is happening). So I have stopped it.
279633. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2797 However, I am unable to do any testing of this.
279934. Fixed an infelicity in the appendfile transport. When checking directories
2800 for a mailbox, to see if any needed to be created, it was accidentally
2801 using path names with one or more superfluous leading slashes; tracing
2802 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
280435. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2805 contents are added to the log line that is written for every discarded
2806 recipient. (Previously a log_message setting was ignored.)
280836. The ${quote: operator now quotes the string if it is empty.
281037. The install script runs exim in order to find its version number. If for
2811 some reason other than non-existence or emptiness, which it checks, it
2812 could not run './exim', it was installing it with an empty version number,
2813 i.e. as "exim-". This error state is now caught, and the installation is
2814 aborted.
281638. An argument was missing from the function that creates an error message
2817 when Exim fails to connect to the socket for saslauthd authentication.
2818 This could cause Exim to crash, or give a corrupted message.
282039. Added isip, isip4, and isip6 to ${if conditions.
282240. The ACL variables $acl_xx are now saved with the message, and can be
2823 accessed later in routers, transports, and filters.
282541. The new lookup type nwildlsearch is like wildlsearch, except that the key
2826 strings in the file are not string-expanded.
282842. If a MAIL command specified a SIZE value that was too large to fit into an
2829 int variable, the check against message_size_limit failed. Such values are
2830 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2831 day this will have to be increased, but I don't think I want to be around
2832 when emails are that large.
2836Exim version 4.22
2839 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2840 iconv() is not standard in FreeBSD.
2842 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2843 IPv6 enabled. The observed symptom was a segmentation fault on return from
2844 the function os_common_find_running_interfaces() in src/os.c.
2846 3. In the check_special_case() function in daemon.c I had used "errno" as an
2847 argument name, which causes warnings on some systems. This was basically a
2848 typo, since it was named "eno" in the comments!
2850 4. The code that waits for the clock to tick (at a resolution of some fraction
2851 of a second) so as to ensure message-id uniqueness was always waiting for
2852 at least one whole tick, when it could have waited for less. [This is
2853 almost certainly not relevant at current processor speeds, where it is
2854 unlikely to ever wait at all. But we try to future-proof.]
2856 5. The function that sleeps for a time interval that includes fractions of a
2857 second contained a race. It did not block SIGALRM between setting the
2858 timer, and suspending (a couple of lines later). If the interval was short
2859 and the sigsuspend() was delayed until after it had expired, the suspension
2860 never ended. On busy systems this could lead to processes getting stuck for
2861 ever.
2863 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2864 process, before it forks any delivery processes. The open lookup caching
2865 mechanism meant that the open file or database connection was passed into
2866 the delivery process. The problem was that delivery processes always tidy
2867 up cached lookup data. This could cause a problem for the next delivery
2868 process started by the queue runner, because the external queue runner
2869 process does not know about the closure. So the next delivery process
2870 still has data in the lookup cache. In the case of a file lookup, there was
2871 no problem because closing a file descriptor in a subprocess doesn't affect
2872 the parent. However, if the lookup was caching a connection to a database,
2873 the connection was closed, and the second delivery process was likely to
2874 see errors such as "PGSQL: query failed: server closed the connection
2875 unexpectedly". The problem has been fixed by closing all cached lookups
2876 in a queue runner before running a delivery process.
2878 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2879 seem to have the "const" qualifier which it has on other OS. I've
2880 parameterised it.
2882 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2883 *of the same type* (client or server) with the same public name that an
2884 error should be diagnosed.
2886 9. When Exim looked up a host name for an IP address, but failed to find the
2887 original IP address when looking up the host name (a safety check), it
2888 output the message "<ip address> does not match any IP for NULL", which was
2889 confusing, to say the least. The bug was that the host name should have
2890 appeared instead of "NULL".
289210. Since release 3.03, if Exim is called by a uid other than root or the Exim
2893 user that is built into the binary, and the -C or -D options is used, root
2894 privilege is dropped before the configuration file is read. In addition,
2895 logging is switched to stderr instead of the normal log files. If the
2896 configuration then re-defines the Exim user, the unprivileged environment
2897 is probably not what is expected, so Exim logs a panic warning message (but
2898 proceeds).
2900 However, if deliver_drop_privilege is set, the unprivileged state may well
2901 be exactly what is intended, so the warning has been cut out in that case,
2902 and Exim is allowed to try to write to its normal log files.
2905Exim version 4.21
2908 1. smtp_return_error_details was not giving details for temporary sender
2909 or receiver verification errors.
2911 2. Diagnose a configuration error if two authenticators have the same public
2912 name.
2914 3. Exim used not to create the message log file for a message until the first
2915 delivery attempt. This could be confusing when incoming messages were held
2916 for policy or load reasons. The message log file is now created at the time
2917 the message is received, and an initial "Received" line is written to it.
2919 4. The automatically generated man page for command line options had a minor
2920 bug that caused no ill effects; however, a more serious problem was that
2921 the procedure for building the man page automatically didn't always
2922 operate. Consequently, release 4.20 contains an out-of-date version. This
2923 shouldn't happen again.
2925 5. When building Exim with embedded Perl support, the script that builds the
2926 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2927 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2929 6. The freeze_tell option was not being used for messages that were frozen on
2930 arrival, either by an ACL or by local_scan().
2932 7. Added the smtp_incomplete_transaction log selector.
2934 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2935 was accepting AUTH without a new EHLO.
2937 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2938 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2939 RFC.
294110. Logging of TCP/IP connections (when configured) now happens in the main
2942 daemon process instead of the child process, so that the TCP/IP connection
2943 count is more accurate (but it can never be perfect).
294511. The use of "drop" in a nested ACL was not being handled correctly in the
2946 outer ACL. Now, if condition failure induced by the nested "drop" causes
2947 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2948 or "require"), the connection is dropped.
295012. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2951 that yield "discard" can now be used with an "accept" or a "discard" verb,
2952 but an error is generated for any others (because I can't see a useful way
2953 to define what should happen).
295513. When an ACL is read dynamically from a file (or anywhere else), the lines
2956 are now processed in the same way as lines in the Exim configuration file.
2957 In particular, continuation lines are supported.
295914. Added the "dnslists = a.b.c!=n.n.n.n" feature.
296115. Added -ti meaning -t -i.
296316. Check for letters, digits, hyphens, and dots in the names of dnslist
2964 domains, and warn by logging if others are found.
296617. At least on BSD, alignment is not guarenteed for the array of ifreq's
2967 returned from GIFCONF when Exim is trying to find the list of interfaces on
2968 a host. The code in os.c has been modified to copy each ifreq to an aligned
2969 structure in all cases.
2971 Also, in some cases, the returned ifreq's were being copied to a 'struct
2972 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2973 means the last couple of bytes of an IPv6 address could be chopped if the
2974 ifreq contained only a normal sockaddr (14 bytes storage).
297618. Named domain lists were not supported in the hosts_treat_as_local option.
2977 An entry such as +xxxx was not recognized, and was treated as a literal
2978 domain name.
298019. Ensure that header lines added by a DATA ACL are included in the reject log
2981 if the ACL subsequently rejects the message.
298320. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2984 MD5 (which is deprecated).
298621. When testing a filter file using -bf, Exim was writing a message when it
2987 took the sender from a "From " line in the message, but it was not doing so
2988 when it took $return_path from a Return-Path: header line. It now does.
299022. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2991 with a valid header line field name (a series of printing characters
2992 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
299423. Changed "disc" in the source to "disk" to conform to the documentation and
2995 the book and for uniformity.
299724. Ignore Sendmail's -Ooption=value command line item.
299925. When execve() failed while trying to run a command in a pipe transport,
3000 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3001 could be confused with a return value of 69 from the command itself. This
3002 has been changed to 127, the value the shell returns if it is asked to run
3003 a non-existent command. The wording for the related log line suggests a
3004 non-existent command as the problem.
300626. If received_header_text expands to an empty string, do not add a Received:
3007 header line to the message. (Well, it adds a token one on the spool, but
3008 marks it "old" so that it doesn't get used or transmitted.)
301027. Installed eximstats 1.28 (addition of -nt option).
301228. There was no check for failure on the call to getsockname() in the daemon
3013 code. This can fail if there is a shortage of resources on the system, with
3014 ENOMEM, for example. A temporary error is now given on failure.
301629. Contrary to the C standard, it seems that in some environments, the
3017 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3018 program. Exim now does this explicitly; it affects the formatting of
3019 timestamps using strftime().
302130. If exiqsumm was given junk data, it threw up some uninitialized variable
3022 complaints. I've now initialized all the variables, to avoid this.
302432. Header lines added by a system filter were not being "seen" during
3025 transport-time rewrites.
302733. The info_callback() function passed to OpenSSL is set up with type void
3028 (*)(SSL *, int, int), as described somewhere. However, when calling the
3029 function (actually a macro) that sets it up, the type void(*)() is
3030 expected. I've put in a cast to prevent warnings from picky compilers.
303234. If a DNS black list lookup found a CNAME record, but there were no A
3033 records associated with the domain it pointed at, Exim crashed.
303535. If a DNS black list lookup returned more than one A record, Exim ignored
3036 all but the first. It now scans all returned addresses if a particular IP
3037 value is being sought. In this situation, the contents of the
3038 $dnslist_value variable are a list of all the addresses, separated by a
3039 comma and a space.
304136. Tightened up the rules for host name lookups using reverse DNS. Exim used
3042 to accept a host name and all its aliases if the forward lookup for any of
3043 them yielded the IP address of the incoming connection. Now it accepts only
3044 those names whose forward lookup yields the correct IP address. Any other
3045 names are discarded. This closes a loophole whereby a rogue DNS
3046 administrator could create reverse DNS records to break through a
3047 wildcarded host restriction in an ACL.
304937. If a user filter or a system filter that ran in a subprocess used any of
3050 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3051 the wrong values were passed to the pipe command ($thisaddress had the
3052 value of $0, $0 had the value of $1, etc). This bug was introduced by
3053 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
305538. Improved the line breaking for long SMTP error messages from ACLs.
3056 Previously, if there was no break point between 40 and 75 characters, Exim
3057 left the rest of the message alone. Two changes have been made: (a) I've
3058 reduced the minimum length to 35 characters; (b) if it can't find a break
3059 point between 35 and 75 characters, it looks ahead and uses the first one
3060 that it finds. This may give the occasional overlong line, but at least the
3061 remaining text gets split now.
306339. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3064 file descriptors might be low, and that setting 1000 would always raise it.
3065 It turns out that in some environments, the limit is already over 1000 and
3066 that lowering it causes trouble. So now Exim takes care not to decrease it.
306840. When delivering a message, the value of $return_path is set to $sender_
3069 address at the start of routing (routers may change the value). By an
3070 oversight, this default was not being set up when an address was tested by
3071 -bt or -bv, which affected the outcome if any router or filter referred to
3072 $return_path.
307441. The idea of the "warn" ACL verb is that it adds a header or writes to the
3075 log only when "message" or "log_message" are set. However, if one of the
3076 conditions was an address verification, or a call to a nested ACL, the
3077 messages generated by the underlying test were being passed through. This
3078 no longer happens. The underlying message is available in $acl_verify_
3079 message for both "message" and "log_message" expansions, so it can be
3080 passed through if needed.
308242. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3083 new expansion $bh_ to give the encoded byte string without charset
3084 translation. Translation happens only if iconv() is available; HAVE_ICONV
3085 indicates this at build time. HEADERS_CHARSET gives the charset to
3086 translate to; headers_charset can change it in the configuration, and
3087 "headers charset" can change it in an individual filter file.
308943. Now that we have a default RFC 2047 charset (see above), the code in Exim
3090 that creates RFC 2047 encoded "words" labels them as that charset instead
3091 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3092 expansion operator; (ii) when Exim creates a From: line for a local
3093 message; (iii) when a header line is rewritten to include a "phrase" part.
309544. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3096 buggy, causing it to skip the first lines of messages whose message ID
3097 ended in 'D'. This would not have bitten before Exim release 4.14, because
3098 message IDs were unlikely to end in 'D' before then. The effect was to have
3099 incorrect size information for certain domains.
310145. #include "config.h" was missing at the start of the crypt16.c module. This
3102 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3103 noticed.
310546. If there was a timeout during a "random" callout check, Exim treated it as
3106 a failure of the random address, and carried on sending RSET and the real
3107 address. If the delay was just some slowness somewhere, the response to the
3108 original RCPT would be taken as a response to RSET and so on, causing
3109 mayhem of various kinds.
311147. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3112 when I implemented it. It didn't allow for the fact that some option values
3113 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3114 the right test for positive values.
311648. Domain names in DNS records are case-independent. Exim always looks them up
3117 in lower case. Some resolvers return domain names in exactly the case they
3118 appear in the zone file, that is, they may contain uppercase letters. Not
3119 all resolvers do this - some return always lower case. Exim was treating a
3120 change of case by a resolver as a change of domain, similar to a widening
3121 of a domain abbreviation. This triggered its re-routing code and so it was
3122 trying to route what was effectively the same domain again. This normally
3123 caused routing to fail (because the router wouldn't handle the domain
3124 twice). Now Exim checks for this case specially, and just changes the
3125 casing of the domain that it ultimately uses when it transmits the message
3126 envelope.
312849. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3129 module.
313150. If a filter generated a file delivery with a non-absolute name (possible if
3132 no home directory exists for the router), the forbid_file option was not
3133 forbidding it.
313551. Added '&' feature to dnslists, to provide bit mask matching in addition to
3136 the existing equality matching.
313852. Exim was using ints instead of ino_t variables in some places where it was
3139 dealing with inode numbers.
314153. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3142 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3143 called TMPDIR, and if it finds it is different, it changes its value.
314554. The smtp_printf() function is now made available to local_scan() so
3146 additional output lines can be written before returning. There is also an
3147 smtp_fflush() function to enable the detection of a dropped connection.
3148 The variables smtp_input and smtp_batched_input are exported to
3149 local_scan().
315155. Changed the default runtime configuration: the message "Unknown user"
3152 has been removed from the ACL, and instead placed on the localuser router,
3153 using the cannot_route_message feature. This means that any verification
3154 failures that generate their own messages won't get overridden. Similarly,
3155 the "Unrouteable address" message that was in the ACL for unverifiable
3156 relay addresses has also been removed.
315856. Added hosts_avoid_esmtp to the smtp transport.
316057. The exicyclog script was not checking for the esoteric option
3161 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3162 will work only if exicyclog is run under the appropriate euid.
316458. Following a discussion on the list, the rules by which Exim recognises line
3165 endings on incoming messages have been changed. The -dropcr and drop_cr
3166 options are now no-ops, retained only for backwards compatibility. The
3167 following line terminators are recognized: LF CRLF CR. However, special
3168 processing applies to CR:
3170 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3171 nor a local message in the state where . is a terminator.
3173 (ii) If a bare CR is encountered in a header line, an extra space is added
3174 after the line terminator so as not to end the header. The reasoning
3175 behind this is that bare CRs in header lines are most likely either
3176 to be mistakes, or people trying to play silly games.
317859. The size of a message, as listed by "-bp" or in the Exim monitor window,
3179 was being incorrectly given as 18 bytes larger than it should have been.
3180 This is a VOB (very old bug).
318260. This may never have affected anything current, but just in case it has:
3183 When the local host is found other than at the start of a list of hosts,
3184 the local host, those with the same MX, and any that follow, are discarded.
3185 When the list in question was part of a longer list of hosts, the following
3186 hosts (not currently being processed) were also being discarded. This no
3187 longer happens. I'm not sure if this situation could ever has previously
3188 arisen.
319061. Added the "/MX" feature to lists of hosts in the manualroute and query
3191 program routers.
319362. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3194 header. This is something that is recommended in a new Internet Draft, and
3195 is something that is documented as being done by Sendmail. There are two
3196 possible values. For messages generated by the autoreply transport, Exim
3197 adds:
3199 Auto-Submitted: auto-replied
3201 whereas for all other generated messages (e.g. bounces) it adds
3203 Auto-Submitted: auto-generated
320563. The "personal" condition in filters now includes a test for the
3206 Auto-Submitted: header. If it contains the string "auto-" the message it
3207 not considered personal.
320964. Added rcpt_include_affixes as a generic transport option.
321165. Added queue_only_override (default true).
321366. Added the syslog_duplication option.
321567. If what should have been the first header line of a message consisted of
3216 a space followed by a colon, Exim was mis-interpreting it as a header line.
3217 It isn't of course - it is syntactically invalid and should therefore be
3218 treated as the start of the message body. The misbehaviour could have
3219 caused a number of strange effects, including loss of data in subsequent
3220 header lines, and spool format errors.
322268. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3223 client host had authenticated. This control can now be exercised by an ACL
3224 for more flexibility.
322669. By default, callouts do not happen when testing with -bh. There is now a
3227 variant, -bhc, which does actually run the callout code, including
3228 consulting and updating the callout cache.
323070. Added support for saslauthd authentication, courtesy of Alexander
3231 Sabourenkov.
323371. If statvfs() failed on the spool or log directories while checking their
3234 size for availability, Exim confusingly gave the error "space shortage".
3235 Furthermore, in debugging mode it crashed with a floating point exception.
3236 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3237 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3238 serious problem, Exim now writes to the main and panic logs when this
3239 happens, with details of the failure. It then refuses to accept the
3240 incoming message, giving the message "spool directory problem" or "log
3241 directory problem" with a 421 code for SMTP messages.
324372. When Exim is about to re-exec itself, it ensures that the file descriptors
3244 0, 1, and 2 exist, because some OS complain for execs without them (see
3245 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3246 descriptors. However, the code omitted to check that the open succeeded,
3247 causing mysterious errors if for some reason the permissions on /dev/null
3248 got screwed. Now Exim writes a message to the main and panic logs, and
3249 bombs out if it can't open /dev/null.
325173. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3252 interact so that it is all more flexible. It is supposed to remain
3253 backwards compatible. Also added extra_local_interfaces.
325574. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3256 to bomb out with an assertion failure - to the client this appears as a
3257 connection drop. This problem occurs in the part of the code that was taken
3258 from the Samba project. Fortunately, the assertion is in a very simple
3259 function, so I have fixed this by reproducing the function inline in the
3260 one place where it is called, and arranging for authentication to fail
3261 instead of killing the process with assert().
326375. The SPA client code was not working when the server requested OEM rather
3264 than Unicode encoding.
326676. Added code to make require_files with a specific uid setting more usable in
3267 the case where statting the file as root fails - usually a non-root-mounted
3268 NFS file system. When this happens and the failure is EACCES, Exim now
3269 forks a subprocess and does the per-uid checking as the relevant uid.
327177. Added process_log_path.
327378. If log_file_path was not explicitly set, a setting of check_log_space or
3274 check_log_inodes was ignored.
327679. If a space check for the spool or log partitions fails, the incident is now
3277 logged. Of course, in the latter case the data may get lost...
327980. Added the %p formatting code to string_format() so that it can be used to
3280 print addresses in debug_print(). Adjusted all the address printing in the
3281 debugging in store.c to use %p rather than %d.
328381. There was a concern that a line of code in smtp_in.c could overflow a
3284 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3285 initially expressed, the concern was not well-founded, because trailing
3286 spaces are removed early. However, if the trailing spaces were followed by
3287 a NULL, they did not get removed, so the overflow was possible. Two fixes
3288 were applied:
3290 (a) I re-wrote the offending code in a cleaner fashion.
3291 (b) If an incoming SMTP command contains a NULL character, it is rejected
3292 as invalid.
329482. When Exim changes uid/gid to the Exim user at daemon start time, it now
3295 runs initgroups(), so that if the Exim user is in any additional groups,
3296 they will be used during message reception.
3299Exim version 4.20
3302The change log for 4.20 and earlier releases has been archived.