1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.212 2005/08/30 09:19:33 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
66 PH/15 Installed PCRE 6.2
68 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
70 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
71 of the problem. Specifically, suggested +O2 rather than +O1 for the
74 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
76 PH/20 If a delivery was routed to a non-standard port by means of an SRV
77 record, the port was not correctly logged when the outgoing_port log
78 selector was set (it logged the transort's default port).
80 PH/21 Added support for host-specific ports to manualroute, queryprogram,
81 fallback_hosts, and "hosts" in the smtp transport.
83 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
84 host errors such as "Connection refused".
86 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
87 authentication with radiusclient 0.4.9:
89 - Error returned from rc_read_config was caught wrongly
90 - Username/password not passed on to radius server due to wrong length.
92 The presumption is that some radiusclient API changes for 4.51/PH/17
93 were not taken care of correctly. The code is still untested by me (my
94 Linux distribution still has 0.3.2 of radiusclient), but it was
95 contributed by a Radius user.
97 PH/24 When doing a callout, the value of $domain wasn't set correctly when
98 expanding the "port" option of the smtp transport.
100 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
101 while reading a MIME header. Thanks to Tom Hughes for a patch.
103 PH/24 Include config.h inside local_scan.h so that configuration settings are
106 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
107 that in an ACL for RCPT (for example), you can examine exactly what was
110 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
111 commands, but it was not correctly comparing the address with the actual
112 client host address. Thus, it would show the EHLO address in Received:
113 header lines when this was not necessary.
115 PH/27 Added the % operator to ${eval:}.
117 PH/28 Exim tries to create and chdir to its spool directory when it starts;
118 it should be ignoring failures (because with -C, for example, it has lost
119 privilege). It wasn't ignoring creation failures other than "already
122 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
125 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
126 because an input line was too long, either on its own, or by virtue of
127 too many continuations, the temporary file was not being removed.
133 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
135 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
136 command itself ended in error, the underlying message about the transport
137 filter timeout was being overwritten with the pipe command error. Now the
138 underlying error message should be appended to the second error message.
140 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
142 PH/02 Reduce the amount of output that "make" produces by default. Full output
143 can still be requested.
145 PH/03 The warning log line about a condition test deferring for a "warn" verb
146 was being output only once per connection, rather than after each
147 occurrence (because it was using the same function as for successful
148 "warn" verbs). This seems wrong, so I have changed it.
150 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
151 it should not have, which might have caused a crash in the right
152 circumstances, but probably never did.
154 PH/04 Installed a modified version of Tony Finch's patch to make submission
155 mode fix the return path as well as the Sender: header line, and to
156 add a /name= option so that you can make the user's friendly name appear
159 TF/03 Added the control = fakedefer ACL modifier.
161 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
162 Mark Lowes for thorough testing.
164 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
166 TK/03 Merged latest SRS patch from Miles Wilton.
168 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
169 with the definition in sysexits.h (which is #included earlier).
170 Fortunately, Exim does not actually use EX_OK. The code used to try to
171 preserve the sysexits.h value, by assumimg that macro definitions were
172 scanned for macro replacements. I have been disabused of this notion,
173 so now the code just undefines EX_OK before #including unistd.h.
175 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
176 in the smtp transport. When a block could not be written in a single
177 write() function, the timeout was being re-applied to each part-write.
178 This seems wrong - if the receiver was accepting one byte at a time it
179 would take for ever. The timeout is now adjusted when this happens. It
180 doesn't have to be particularly precise.
182 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
183 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
185 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
186 without a domain if the check to <postmaster@domain> fails.
188 SC/01 Eximstats: added -xls and the ability to specify output files
189 (patch written by Frank Heydlauf).
191 SC/02 Eximstats: use FileHandles for outputing results.
193 SC/03 Eximstats: allow any combination of xls, txt, and html output.
195 SC/04 Eximstats: fixed display of large numbers with -nvr option
197 SC/05 Eximstats: fixed merging of reports with empty tables.
199 SC/06 Eximstats: added the -include_original_destination flag
201 SC/07 Eximstats: removed tabs and trailing whitespace.
203 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
205 TK/06 MBOX spool code: Add real "From " MBOX separator line
206 so the .eml file is really in mbox format (even though
207 most programs do not really care). Patch from Alex Miller.
209 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
210 The latter is generated from $received_to and is only set if the
211 message has one envelope recipient. SA can use these headers,
212 obviously out-of-the-box. Patch from Alex Miller.
214 PH/08 The ${def test on a variable was returning false if the variable's
215 value was "0", contrary to what the specification has always said!
216 The result should be true unless the variable is empty.
218 PH/09 The syntax error of a character other than { following "${if
219 def:variable_name" (after optional whitespace) was not being diagnosed.
220 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
221 accidental colon was present, for example, could give incorrect results.
223 PH/10 Tidied the code in a number of places where the st_size field of a stat()
224 result is used (not including appendfile, where other changes are about
227 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
228 This involved changing a lot of size variables from int to off_t. It
229 should work with maildirs and everything.
231 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
232 spamd dying while we are connected to it.
234 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
237 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
238 function for BDB 4.3.
240 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
242 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
243 the sources for PCRE have been re-arranged and more files are now
246 PH/15 The code I had for printing potentially long long variables in PH/11
247 above was not the best (it lost precision). The length of off_t variables
248 is now inspected at build time, and an appropriate printing format (%ld
249 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
250 to be "long long int" or "long int". This is needed for the internal
251 formatting function string_vformat().
253 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
254 the configuration file to be ":syslog", then the script "guesses" where
255 the logs files are, rather than using the compiled in default. In our
256 case the guess is not the same as the compiled default, so the script
257 suddenly stopped working when I started to use syslog. The patch checks
258 to see if log_file_path is "". If so, it attempts to read it from exim
259 with no configuration file to get the compiled in version, before it
260 falls back to the previous guessing code."
262 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
263 implementing BATV in an Exim configuration. See NewStuff for the gory
266 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
267 Makefile that are specific to HP-UX.
269 PH/18 If the "use_postmaster" option was set for a recipient callout together
270 with the "random" option, the postmaster address was used as the MAIL
271 FROM address for the random test, but not for the subsequent recipient
272 test. It is now used for both.
274 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
275 patch removes a few documentation additions to RFC 3028, because the
276 latest draft now contains them. It adds the new en;ascii-case comparator
277 and a new error check for 8bit text in MIME parts. Comparator and
278 require names are now matched exactly. I enabled the subaddress
279 extension, but it is not well tested yet (read: it works for me)."
281 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
282 rework some of the code of TK/09 above to avoid the hardwired use of
283 "%lld" and "long long". Replaced the call to snprintf() with a call to
286 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
287 records point to non-existent hosts", "retry timeout exceeded", and
288 "retry time not reached for any host after a long failure period".
290 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
291 experimental DomainKeys support:
293 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
294 (2) On an error such as an illegally used "control", the wrong name for
295 the control was given.
297 These problems did NOT occur unless DomainKeys support was compiled.
299 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
301 PH/24 Added ${if match_ip condition.
303 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
304 databases so that it will be absolutely obvious if a crash occurs in the
305 DB library. This is a regular occurrence (often caused by mis-matched
308 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
309 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
310 picked up on a user's system that detects such things. There doesn't seem
311 to be a gcc warning option for this - only an attribute that has to be
312 put on the function's prototype. It seems that in Fedora Core 4 they have
313 set this on a number of new functions. No doubt there will be more in due
316 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
317 specify a transport. However, if an address that was verified by such a
318 router was the subject of a callout, Exim crashed because it tried to
319 read the rcpt_include_affixes from the non-existent transport. Now it
320 just assumes that the setting of that option is false. This bug was
321 introduced by 4.51/PH/31.
323 PH/28 Changed -d+all to exclude +memory, because that information is very
324 rarely of interest, but it makes the output a lot bigger. People tend to
325 do -d+all out of habit.
327 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
328 code in os-type was giving problems when libc.so lives in lib64, like on
331 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
332 aren't the modern standard, and it seems that some systems' include files
333 don't always have them. Exim was already checking for some of the newer
334 ones like T_AAAA, and defining it itself. I've added checks for all the
335 record types that Exim uses.
337 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
338 not automatically generating a new one, as it is supposed to. This
339 prevented TLS from working. If the file did exist, but contained invalid
340 data, a new version was generated, as expected. It was only the case of a
341 non-existent file that was broken.
343 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
344 with a change in libdomainkeys > 0.64.
346 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
347 from DNS. If the selector record carries the flag, it now has
348 precedence over the domain-wide flag.
350 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
352 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
353 the use of a port name that isn't defined in /etc/services) occurred, the
354 message was deferred as in a normal delivery, and thus remained on the
355 spool, instead of being failed because of the mua_wrapper setting. This
356 is now fixed, and I tidied up some of the mua_wrapper messages at the
359 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
360 the messages in a hash of arrays rather than using individual hashes.
361 This is a bit cleaner and results in dramatic memory savings, albeit
362 at a slight CPU cost.
364 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
365 as requested by Marc Sherman.
367 SC/10 Eximstats: added histograms for user specified patterns as requested
370 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
372 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
373 fopen() in the content-scanning modules that did not already have it.
379 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
380 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
382 TK/02 Fix ACL "control" statement not being available in MIME ACL.
384 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
386 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
387 to test Sieve filters that use "vacation".
389 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
390 that changes the way the GnuTLS parameters are stored in the cache file.
391 The new format can be generated externally. For backward compatibility,
392 if the data in the cache doesn't make sense, Exim assumes it has read an
393 old-format file, and it generates new data and writes a new file. This
394 means that you can't go back to an older release without removing the
397 PH/03 A redirect router that has both "unseen" and "one_time" set does not
398 work if there are any delivery delays because "one_time" forces the
399 parent to be marked "delivered", so its unseen clone is never tried
400 again. For this reason, Exim now forbids the simultaneous setting of
403 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
404 redirected to themselves ("homonym" addresses). Read the long ChangeLog
405 entry if you want to know the details. The fix, however, neglected to
406 consider the case when local delivery batching is involved. The test for
407 "previously delivered" was not happening when checking to see if an
408 address could be batched with a previous (undelivered) one; under
409 certain circumstances this could lead to multiple deliveries to the same
412 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
413 in its include files, and this causes problems building Exim.
415 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
416 header_syntax) but Exim was just ignoring anything given after a slash.
417 In particular, this caused confusion with an attempt to use "verify =
418 reverse_host_lookup/defer_ok". An error is now given when options are
419 supplied for verify items that do not have them. (Maybe reverse_host_
420 lookup should have a defer_ok option, but that's a different point.)
422 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
423 defined by RFC 821) to 2048, because there were problems with some AUTH
424 commands, and RFC 1869 says the size should be increased for extended
425 SMTP commands that take arguments.
427 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
430 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
431 "unknown" error; now it says that the functionality isn't in the binary.
433 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
434 an address' error message when a string expansion fails (syntax or
435 whatever). Otherwise the password may appear in the log. Following change
436 PH/42 below, there is no longer a chance of it appearing in a bounce
439 PH/11 Installed exipick version 20050225.0 from John Jetmore.
441 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
442 of its addresses was ever tried. (Bugzilla bug #2.)
444 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
445 the result incorrectly in the debug output. (It correctly added a newline
446 to what was transported.)
448 TF/01 Added $received_time.
450 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
451 commented out examples of how to interface to a virus scanner and to
452 SpamAssassin. Also added commented examples of av_scanner and
453 spamd_address settings.
455 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
456 and controls are allowed in which ACLs. There were a couple of minor
457 errors. Some of the entries in the conditions table (which is a table of
458 where they are NOT allowed) were getting very unwieldy; rewrote them as a
459 negation of where the condition IS allowed.
461 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
463 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
464 header file does not have a version number, so I've had to invent a new
465 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
466 API. The code is untested by me (my Linux distribution still has 0.3.2 of
467 radiusclient), but it was contributed by a Radius user.
469 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
470 files or directories, for OpenSSL.
472 PH/19 When an Exim process that is running as root has to create an Exim log
473 file, it does so in a subprocess that runs as exim:exim so as to get the
474 ownership right at creation (otherwise, other Exim processes might see
475 the file with the wrong ownership). There was no test for failure of this
476 fork() call, which would lead to the process getting stuck as it waited
477 for a non-existent subprocess. Forks do occasionally fail when resources
478 run out. I reviewed all the other calls to fork(); they all seem to check
481 PH/20 When checking for unexpected SMTP input at connect time (before writing
482 the banner), Exim was not dealing correctly with a non-positive return
483 from the read() function. If the client had disconnected by this time,
484 the result was a log entry for a synchronization error with an empty
485 string after "input=" when read() returned zero. If read() returned -1
486 (an event I could not check), uninitialized data bytes were printed.
487 There were reports of junk text (parts of files, etc) appearing after
490 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
492 PH/22 Added support for macro redefinition, and (re)definition in between
493 driver and ACL definitions.
495 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
496 forgetting to use the resulting value; it was using the unexpanded value.
498 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
499 hadn't been configured. The fix is from Juergen Kreileder, who
500 understands it better than I do:
502 "Here's what I see happening with three configured cyrus_sasl
503 authenticators configured (plain, login, cram-md5):
505 On startup auth_cyrus_sasl_init() gets called for each of these.
506 This means three calls to sasl_listmech() without a specified mech_list.
507 => SASL tests which mechs of all available mechs actually work
508 => three warnings about OTP not working
509 => the returned list contains: plain, login, cram-md5, digest-md5, ...
511 With the patch, sasl_listmech() also gets called three times. But now
512 SASL's mech_list option is set to the server_mech specified in the the
513 authenticator. Or in other words, the answer from sasl_listmech()
514 gets limited to just the mech you're testing for (which is different
516 => the return list contains just 'plain' or 'login', 'cram-md5' or
517 nothing depending on the value of ob->server_mech.
519 I've just tested the patch: Authentication still works fine,
520 unavailable mechs specified in the exim configuration are still
521 caught, and the auth.log warnings about OTP are gone."
523 PH/25 When debugging is enabled, the contents of the command line are added
524 to the debugging output, even when log_selector=+arguments is not
527 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
528 answer is "GNU", and only if the return is "GNU/something" is the answer
531 PH/27 $acl_verify_message is now set immediately after the failure of a
532 verification in an ACL, and so is available in subsequent modifiers. In
533 particular, the message can be preserved by coding like this:
535 warn !verify = sender
536 set acl_m0 = $acl_verify_message
538 Previously, $acl_verify_message was set only while expanding "message"
539 and "log_message" when a very denied access.
541 PH/28 Modified OS/os.c-Linux with
543 -#ifndef OS_LOAD_AVERAGE
544 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
546 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
547 nomenclature these days.)
549 PH/29 Installed patch from the Sieve maintainer that adds the options
550 sieve_useraddress and sieve_subaddress to the redirect router.
552 PH/30 In these circumstances:
553 . Two addresses routed to the same list of hosts;
554 . First host does not offer TLS;
555 . First host accepts first address;
556 . First host gives temporary error to second address;
557 . Second host offers TLS and a TLS session is established;
558 . Second host accepts second address.
559 Exim incorrectly logged both deliveries with the TLS parameters (cipher
560 and peerdn, if requested) that were in fact used only for the second
563 PH/31 When doing a callout as part of verifying an address, Exim was not paying
564 attention to any local part prefix or suffix that was matched by the
565 router that accepted the address. It now behaves in the same way as it
566 does for delivery: the affixes are removed from the local part unless
567 rcpt_include_affixes is set on the transport.
569 PH/32 Add the sender address, as F=<...>, to the log line when logging a
570 timeout during the DATA phase of an incoming message.
572 PH/33 Sieve envelope tests were broken for match types other than :is. I have
573 applied a patch sanctioned by the Sieve maintainer.
575 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
576 the uid or gid is negative. A case of a negative gid caused this to be
577 noticed. The fix allows for either to be negative.
579 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
580 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
583 PH/36 The change PH/12 above was broken. Fixed it.
585 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
586 the grounds that routing the same address twice would always produce the
587 same answer. This might have been true once, but it is certainly no
588 longer true now. Routing a child address may depend on the previous
589 routing that produced that child. Some complicated redirection strategies
590 went wrong when messages had multiple recipients, and made Exim's
591 behaviour dependent on the order in which the addresses were given.
593 I have moved the duplicate checking until after the routing is complete.
594 Exim scans the addresses that are assigned to local and remote
595 transports, and removes any duplicates. This means that more work will be
596 done, as duplicates will always all be routed, but duplicates are
597 presumably rare, so I don't expect this is of any significance.
599 For deliveries to pipes, files, and autoreplies, the duplicate checking
600 still happens during the routing process, since they are not going to be
603 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
604 It corrects a timeout issue with spamd. This is Ian's comment: "The
605 background is that sometimes spamd either never reads data from a
606 connection it has accepted, or it never writes response data. The exiscan
607 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
608 blindly assumes that writes won't block so it may never time out."
610 PH/39 Allow G after quota size as well as K and M.
612 PH/40 The value set for $authenticated_id in an authenticator may not contain
613 binary zeroes or newlines because the value is written to log lines and
614 to spool files. There was no check on this. Now the value is run through
615 the string_printing() function so that such characters are converted to
616 printable escape sequences.
618 PH/41 $message_linecount is a new variable that contains the total number of
619 lines in the message. Compare $body_linecount, which is the count for the
622 PH/42 Exim no longer gives details of delivery errors for specific addresses in
623 bounce and delay warning messages, except in certain special cases, which
626 (a) An SMTP error message from a remote host;
627 (b) A message specified in a :fail: redirection;
628 (c) A message specified in a "fail" command in a system filter;
629 (d) A message specified in a FAIL return from the queryprogram router;
630 (e) A message specified by the cannot_route_message router option.
632 In these cases only, Exim does include the error details in bounce and
633 warning messages. There are also a few cases where bland messages such
634 as "unrouteable address" or "local delivery error" are given.
636 PH/43 $value is now also set for the "else" part of a ${run expansion.
638 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
639 being worked on, but at least Exim now implements the latest version to
642 PH/45 In a pipe transport, although a timeout while waiting for the pipe
643 process to complete was treated as a delivery failure, a timeout while
644 writing the message to the pipe was logged, but erroneously treated as a
645 successful delivery. Such timeouts include transport filter timeouts. For
646 consistency with the overall process timeout, these timeouts are now
647 treated as errors, giving rise to delivery failures by default. However,
648 there is now a new Boolean option for the pipe transport called
649 timeout_defer, which, if set TRUE, converts the failures into defers for
650 both kinds of timeout. A transport filter timeout is now identified in
653 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
654 systems where "make" and "gmake" are different, calling "gmake" at top
655 level broke things. I've arranged for the value of $(MAKE) to be passed
656 from the Makefile to this script so that it can call the same version of
660 A note about Exim versions 4.44 and 4.50
661 ----------------------------------------
663 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
664 changes of various kinds. As a consequence, a big documentation update was
665 needed. This delayed the release for rather longer than seemed good, especially
666 in the light of a couple of (minor) security issues. Therefore, the changes
667 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
668 release. So 4.44 and 4.50 are in effect two different branches that both start
671 I have left the 4.50 change log unchanged; it contains all the changes since
672 4.43. The change log for 4.44 is below; many of its items are identical to
673 those for 4.50. This seems to be the most sensible way to preserve the
674 historical information.
680 1. Minor wording change to the doc/README.SIEVE file.
682 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
683 computation of the current number of files was incorrect.
685 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
686 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
689 4. Give more explanation in the error message when the command for a transport
690 filter fails to execute.
692 5. There are several places where Exim runs a non-Exim command in a
693 subprocess. The SIGUSR1 signal should be disabled for these processes. This
694 was being done only for the command run by the queryprogram router. It is
695 now done for all such subprocesses. The other cases are: ${run, transport
696 filters, and the commands run by the lmtp and pipe transports.
698 6. Added CONFIGURE_GROUP build-time option.
700 7. Some older OS have a limit of 256 on the maximum number of file
701 descriptors. Exim was using setrlimit() to set 1000 as a large value
702 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
703 systems. I've change it so that if it can't get 1000, it tries for 256.
705 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
706 was an oversight, and furthermore, ever since the addition of extra
707 controls (e.g. 4.43/32), the checks on when to allow different forms of
708 "control" were broken. There should now be diagnostics for all cases when a
709 control that does not make sense is encountered.
711 9. Added the /retain_sender option to "control=submission".
713 10. $recipients is now available in the predata ACL (oversight).
715 11. Tidy the search cache before the fork to do a delivery from a message
716 received from the command line. Otherwise the child will trigger a lookup
717 failure and thereby defer the delivery if it tries to use (for example) a
718 cached ldap connection that the parent has called unbind on.
720 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
721 of $address_data from the recipient verification was clobbered by the
724 13. The value of address_data from a sender verification is now available in
725 $sender_address_data in subsequent conditions in the ACL statement.
727 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
729 15. Added a new option "connect=<time>" to callout options, to set a different
732 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
733 was its contents. (It was OK if the option was not defined at all.)
735 17. A "Completed" log line is now written for messages that are removed from
736 the spool by the -Mrm option.
738 18. New variables $sender_verify_failure and $recipient_verify_failure contain
739 information about exactly what failed.
741 19. Added -dd to debug only the daemon process.
743 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
744 handles timeouts, both on the server side and network timeouts. Renamed the
745 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
747 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
748 It is now set to "smtps".
750 22. $host_address is now set to the target address during the checking of
753 23. When checking ignore_target_hosts for an ipliteral router, no host name was
754 being passed; this would have caused $sender_host_name to have been used if
755 matching the list had actually called for a host name (not very likely,
756 since this list is usually IP addresses). A host name is now passed as
759 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
760 code that specifies a non-restarting handler (typically sigaction() in
761 modern systems) in an attempt to fix a rare and obscure crash bug.
763 25. Narrowed the window for a race in the daemon that could cause it to ignore
764 SIGCHLD signals. This is not a major problem, because they are used only to
765 wake it up if nothing else does.
767 26. A malformed maildirsize file could cause Exim to calculate negative values
768 for the mailbox size or file count. Odd effects could occur as a result.
769 The maildirsize information is now recalculated if the size or filecount
772 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
773 support for a long time. Removed HAVE_SYS_VFS_H.
775 28. Installed the latest version of exipick from John Jetmore.
777 29. In an address list, if the pattern was not a regular expression, an empty
778 subject address (from a bounce message) matched only if the pattern was an
779 empty string. Non-empty patterns were not even tested. This was the wrong
780 because it is perfectly reasonable to use an empty address as part of a
781 database query. An empty address is now tested by patterns that are
782 lookups. However, all the other forms of pattern expect the subject to
783 contain a local part and a domain, and therefore, for them, an empty
784 address still always fails if the pattern is not itself empty.
786 30. Exim went into a mad DNS loop when attempting to do a callout where the
787 host was specified on an smtp transport, and looking it up yielded more
790 31. Re-factored the code for checking spool and log partition space into a
791 function that finds that data and another that does the check. The former
792 is then used to implement four new variables: $spool_space, $log_space,
793 $spool_inodes, and $log_inodes.
795 32. The RFC2047 encoding function was originally intended for short strings
796 such as real names; it was not keeping to the 75-character limit for
797 encoded words that the RFC imposes. It now respects the limit, and
798 generates multiple encoded words if necessary. To be on the safe side, I
799 have increased the buffer size for the ${rfc2047: expansion operator from
802 33. It is now permitted to omit both strings after an "if" condition; if the
803 condition is true, the result is "true". As before, when the second string
804 is omitted, a false condition yields an empty string. This makes it less
805 cumbersome to write custom ACL and router conditions.
807 34. Failure to deliver a bounce message always caused it to be frozen, even if
808 there was an errors_to setting on the router. The errors_to setting is now
811 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
812 canonical form (fully expanded) before being placed in
813 $sender_host_address.
815 36. The table in the code that translates DNS record types into text (T_A to
816 "A" for instance) was missing entries for NS and CNAME. It is just possible
817 that this could have caused confusion if both these types were looked up
818 for the same domain, because the text type is used as part of Exim's
819 per-process caching. But the chance of anyone hitting this buglet seems
822 37. The dnsdb lookup has been extended in a number of ways.
824 (1) There is a new type, "zns", which walks up the domain tree until it
825 finds some nameserver records. It should be used with care.
827 (2) There is a new type, "mxh", which is like "mx" except that it returns
828 just the host names, not the priorities.
830 (3) It is now possible to give a list of domains (or IP addresses) to be
831 looked up. The behaviour when one of the lookups defers can be
832 controlled by a keyword.
834 (4) It is now possible to specify the separator character for use when
835 multiple records are returned.
837 38. The dnslists ACL condition has been extended: it is now possible to supply
838 a list of IP addresses and/or domains to be looked up in a particular DNS
841 39. Added log_selector=+queue_time_overall.
843 40. When running the queue in the test harness, wait just a tad after forking a
844 delivery process, to get repeatability of debugging output.
846 41. Include certificate and key file names in error message when GnuTLS fails
847 to set them up, because the GnuTLS error message doesn't include the name
848 of the failing file when there is a problem reading it.
850 42. Allow both -bf and -bF in the same test run.
852 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
854 44. The "Exiscan patch" is now merged into the mainline Exim source.
856 45. Sometimes the final signoff response after QUIT could fail to get
857 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
858 < 0 before doing a fflush(). This bug looks as though it goes back to the
859 introduction of TLS in release 3.20, but "sometimes" must have been rare
860 because the tests only now provoked it.
862 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
863 (this can affect the format of dates).
865 47. exim_tidydb, when checking for the continued existence of a message for
866 which it has found a message-specific retry record, was not finding
867 messages that were in split spool directories. Consequently, it was
868 deleting retry records that should have stayed in existence.
870 48. Steve fixed some bugs in eximstats.
872 49. The SPA authentication driver was not abandoning authentication and moving
873 on to the next authenticator when an expansion was forced to fail,
874 contradicting the general specification for all authenticators. Instead it
875 was generating a temporary error. It now behaves as specified.
877 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
878 (the order specifies the preference for clients). The order is now AES256,
879 AES128, 3DES, ARCFOUR128.
881 51. Small patch to Sieve code - explicitly set From: when generating an
884 52. Exim crashed if a remote delivery caused a very long error message to be
885 recorded - for instance if somebody sent an entire SpamAssassin report back
886 as a large number of 550 error lines. This bug was coincidentally fixed by
887 increasing the size of one of Exim's internal buffers (big_buffer) that
888 happened as part of the Exiscan merge. However, to be on the safe side, I
889 have made the code more robust (and fixed the comments that describe what
892 53. Now that there can be additional text after "Completed" in log lines (if
893 the queue_time_overall log selector is set), a one-byte patch to exigrep
894 was needed to allow it to recognize "Completed" as not the last thing in
897 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
898 patch that reportedly fixes this has been added. I am not expert enough to
899 create a test for it. This is what the patch creator wrote:
901 "I found a little strange behaviour of ldap code when working with
902 Windows 2003 AD Domain, where users was placed in more than one
903 Organization Units. When I tried to give exim partial DN, the exit code
904 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
905 But simultaneously result of request was absolutely normal ldap result,
906 so I produce this patch..."
908 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
909 REFERENCE, so I have modified the code to exclude the patch when that macro
912 55. Some experimental protocols are using DNS PTR records for new purposes. The
913 keys for these records are domain names, not reversed IP addresses. The
914 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
915 leaves it alone. Component reversal etc. now happens only for IP addresses.
918 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
920 57. Double the size of the debug message buffer (to 2048) so that more of very
921 long debug lines gets shown.
923 58. The exicyclog utility now does better if the number of log files to keep
924 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
926 59. Two changes related to the smtp_active_hostname option:
928 (1) $smtp_active_hostname is now available as a variable.
929 (2) The default for smtp_banner uses $smtp_active_hostname instead
930 of $primary_hostname.
932 60. The host_aton() function is supposed to be passed a string that is known
933 to be a valid IP address. However, in the case of IPv6 addresses, it was
934 not checking this. This is a hostage to fortune. Exim now panics and dies
935 if the condition is not met. A case was found where this could be provoked
936 from a dnsdb PTR lookup with an IPv6 address that had more than 8
937 components; fortuitously, this particular loophole had already been fixed
938 by change 4.50/55 above.
940 If there are any other similar loopholes, the new check in host_aton()
941 itself should stop them being exploited. The report I received stated that
942 data on the command line could provoke the exploit when Exim was running as
943 exim, but did not say which command line option was involved. All I could
944 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
948 61. There was a buffer overflow vulnerability in the SPA authentication code
949 (which came originally from the Samba project). I have added a test to the
950 spa_base64_to_bits() function which I hope fixes it.
953 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
954 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
956 63. The daemon start-up calls getloadavg() while still root for those OS that
957 need the first call to be done as root, but it missed one case: when
958 deliver_queue_load_max is set with deliver_drop_privilege. This is
959 necessary for the benefit of the queue runner, because there is no re-exec
960 when deliver_drop_privilege is set.
962 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
965 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
966 was in use, was not putting the data itself into the right store pool;
967 consequently, it could be overwritten for a subsequent message in the same
968 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
971 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
973 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
974 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
975 the function were treating the return as a boolean value, which happened to
976 work because 0=false and not-0=true, but is not correct code.
978 68. The host_aton() function was not handling scoped IPv6 addresses (those
979 with, for example, "%eth0" on the end) correctly.
981 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
982 negated items (that is, ~something) in unsigned ints. Some compilers
983 apparently mutter when there is no cast.
985 70. If an address verification called from an ACL failed, and did not produce a
986 user-specific message (i.e. there was only a "system" message), nothing was
987 put in $acl_verify_message. In this situation, it now puts the system
990 71. Change 4.23/11 added synchronization checking at the start of an SMTP
991 session; change 4.31/43 added the unwanted input to the log line - except
992 that it did not do this in the start of session case. It now does.
994 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
995 This is wrong and can cause the other end to generate a synchronization
996 error if it is another Exim or anything else that does the synchronization
997 check. A QUIT command is no longer sent after a timeout.
999 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1000 during host lookups.
1002 74. The defer_ok option of callout verification was not working if it was used
1003 when verifying addresses in header lines, that is, for this case:
1005 verify = header_sender/callout=defer_ok
1007 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1008 those file descriptors could be used for SMTP connections. If anything
1009 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1010 could be sent to the SMTP client, causing chaos. The daemon now opens
1011 stdin, stdout, and stderr to /dev/null when it puts itself into the
1014 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1015 log by default. The user can override this with suitable Perl magic.
1017 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1018 the log message when discard triggers, was not working for the DATA ACL or
1019 for the non-SMTP ACL.
1021 78. Error message wording change in sieve.c.
1023 79. If smtp_accept_max_per_host was set, the number of connections could be
1024 restricted to fewer than expected, because the daemon was trying to set up
1025 a new connection before checking whether the processes handling previous
1026 connections had finished. The check for completed processes is now done
1027 earlier. On busy systems, this bug wouldn't be noticed because something
1028 else would have woken the daemon, and it would have reaped the completed
1031 80. If a message was submitted locally by a user whose login name contained one
1032 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1033 It caused a spool format error. I have fixed the spool reading code. A
1034 related problem was that the "from" clause in the Received: line became
1035 illegal because of the space(s). It is now covered by ${quote_local_part.
1037 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1040 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1042 83. Patch from Sieve maintainer to fix unterminated string problem in
1043 "vacation" handling.
1045 84. Some minor changes to the Linux configuration files to help with other
1046 OS variants using glibc.
1048 85. One more patch for Sieve to update vacation handling to latest spec.
1051 ----------------------------------------------------
1052 See the note above about the 4.44 and 4.50 releases.
1053 ----------------------------------------------------
1059 1. Change 4.43/35 introduced a bug that caused file counts to be
1060 incorrectly computed when quota_filecount was set in an appendfile
1063 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1064 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1067 3. Give more explanation in the error message when the command for a transport
1068 filter fails to execute.
1070 4. There are several places where Exim runs a non-Exim command in a
1071 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1072 was being done only for the command run by the queryprogram router. It is
1073 now done for all such subprocesses. The other cases are: ${run, transport
1074 filters, and the commands run by the lmtp and pipe transports.
1076 5. Some older OS have a limit of 256 on the maximum number of file
1077 descriptors. Exim was using setrlimit() to set 1000 as a large value
1078 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1079 systems. I've change it so that if it can't get 1000, it tries for 256.
1081 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1082 was an oversight, and furthermore, ever since the addition of extra
1083 controls (e.g. 4.43/32), the checks on when to allow different forms of
1084 "control" were broken. There should now be diagnostics for all cases when a
1085 control that does not make sense is encountered.
1087 7. $recipients is now available in the predata ACL (oversight).
1089 8. Tidy the search cache before the fork to do a delivery from a message
1090 received from the command line. Otherwise the child will trigger a lookup
1091 failure and thereby defer the delivery if it tries to use (for example) a
1092 cached ldap connection that the parent has called unbind on.
1094 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1095 of $address_data from the recipient verification was clobbered by the
1096 sender verification.
1098 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1099 was its contents. (It was OK if the option was not defined at all.)
1101 11. A "Completed" log line is now written for messages that are removed from
1102 the spool by the -Mrm option.
1104 12. $host_address is now set to the target address during the checking of
1105 ignore_target_hosts.
1107 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1108 being passed; this would have caused $sender_host_name to have been used if
1109 matching the list had actually called for a host name (not very likely,
1110 since this list is usually IP addresses). A host name is now passed as
1113 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1114 code that specifies a non-restarting handler (typically sigaction() in
1115 modern systems) in an attempt to fix a rare and obscure crash bug.
1117 15. Narrowed the window for a race in the daemon that could cause it to ignore
1118 SIGCHLD signals. This is not a major problem, because they are used only to
1119 wake it up if nothing else does.
1121 16. A malformed maildirsize file could cause Exim to calculate negative values
1122 for the mailbox size or file count. Odd effects could occur as a result.
1123 The maildirsize information is now recalculated if the size or filecount
1126 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1127 support for a long time. Removed HAVE_SYS_VFS_H.
1129 18. Updated exipick to current release from John Jetmore.
1131 19. Allow an empty sender to be matched against a lookup in an address list.
1132 Previously the only cases considered were a regular expression, or an
1135 20. Exim went into a mad DNS lookup loop when doing a callout where the
1136 host was specified on the transport, if the DNS lookup yielded more than
1139 21. The RFC2047 encoding function was originally intended for short strings
1140 such as real names; it was not keeping to the 75-character limit for
1141 encoded words that the RFC imposes. It now respects the limit, and
1142 generates multiple encoded words if necessary. To be on the safe side, I
1143 have increased the buffer size for the ${rfc2047: expansion operator from
1146 22. Failure to deliver a bounce message always caused it to be frozen, even if
1147 there was an errors_to setting on the router. The errors_to setting is now
1150 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1151 canonical form (fully expanded) before being placed in
1152 $sender_host_address.
1154 24. Updated eximstats to version 1.33
1156 25. Include certificate and key file names in error message when GnuTLS fails
1157 to set them up, because the GnuTLS error message doesn't include the name
1158 of the failing file when there is a problem reading it.
1160 26. Expand error message when OpenSSL has problems setting up cert/key files.
1163 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1164 (this can affect the format of dates).
1166 28. exim_tidydb, when checking for the continued existence of a message for
1167 which it has found a message-specific retry record, was not finding
1168 messages that were in split spool directories. Consequently, it was
1169 deleting retry records that should have stayed in existence.
1171 29. eximstats updated to version 1.35
1172 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1173 1.35 - bugfix such that pie charts by volume are generated correctly
1175 30. The SPA authentication driver was not abandoning authentication and moving
1176 on to the next authenticator when an expansion was forced to fail,
1177 contradicting the general specification for all authenticators. Instead it
1178 was generating a temporary error. It now behaves as specified.
1180 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1181 (the order specifies the preference for clients). The order is now AES256,
1182 AES128, 3DES, ARCFOUR128.
1184 31. Small patch to Sieve code - explicitly set From: when generating an
1187 32. Exim crashed if a remote delivery caused a very long error message to be
1188 recorded - for instance if somebody sent an entire SpamAssassin report back
1189 as a large number of 550 error lines. This bug was coincidentally fixed by
1190 increasing the size of one of Exim's internal buffers (big_buffer) that
1191 happened as part of the Exiscan merge. However, to be on the safe side, I
1192 have made the code more robust (and fixed the comments that describe what
1195 33. Some experimental protocols are using DNS PTR records for new purposes. The
1196 keys for these records are domain names, not reversed IP addresses. The
1197 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1198 leaves it alone. Component reversal etc. now happens only for IP addresses.
1201 34. The host_aton() function is supposed to be passed a string that is known
1202 to be a valid IP address. However, in the case of IPv6 addresses, it was
1203 not checking this. This is a hostage to fortune. Exim now panics and dies
1204 if the condition is not met. A case was found where this could be provoked
1205 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1206 components; fortuitously, this particular loophole had already been fixed
1207 by change 4.50/55 or 4.44/33 above.
1209 If there are any other similar loopholes, the new check in host_aton()
1210 itself should stop them being exploited. The report I received stated that
1211 data on the command line could provoke the exploit when Exim was running as
1212 exim, but did not say which command line option was involved. All I could
1213 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1214 running as the user.
1217 35. There was a buffer overflow vulnerability in the SPA authentication code
1218 (which came originally from the Samba project). I have added a test to the
1219 spa_base64_to_bits() function which I hope fixes it.
1222 36. The daemon start-up calls getloadavg() while still root for those OS that
1223 need the first call to be done as root, but it missed one case: when
1224 deliver_queue_load_max is set with deliver_drop_privilege. This is
1225 necessary for the benefit of the queue runner, because there is no re-exec
1226 when deliver_drop_privilege is set.
1228 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1229 was in use, was not putting the data itself into the right store pool;
1230 consequently, it could be overwritten for a subsequent message in the same
1231 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1234 38. Sometimes the final signoff response after QUIT could fail to get
1235 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1236 < 0 before doing a fflush(). This bug looks as though it goes back to the
1237 introduction of TLS in release 3.20, but "sometimes" must have been rare
1238 because the tests only now provoked it.
1244 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1245 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1246 Now it returns an integer. A number of calls were still expecting a T/F
1247 return. Fortuitously, in all cases, the tests worked in OK situations,
1248 which is the norm. However, things would have gone wrong on any write
1249 failures on the smtp file descriptor. This function is used when sending
1250 messages over SMTP and also when doing verify callouts.
1252 2. When Exim is called to do synchronous delivery of a locally submitted
1253 message (the -odf or -odi options), it no longer closes stderr before doing
1256 3. Implemented the mua_wrapper option.
1258 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1260 5. Implemented the functions header_remove(), header_testname(),
1261 header_add_at_position(), and receive_remove_recipient(), and exported them
1264 6. If an ACL "warn" statement specified the addition of headers, Exim already
1265 inserted X-ACL-Warn: at the start if there was no header name. However, it
1266 was not making this test for the second and subsequent header lines if
1267 there were newlines in the string. This meant that an invalid header could
1268 be inserted if Exim was badly configured.
1270 7. Allow an ACL "warn" statement to add header lines at the start or after all
1271 the Received: headers, as well as at the end.
1273 8. Added the rcpt_4xx retry error code.
1275 9. Added postmaster_mailfrom=xxx to callout verification option.
1277 10. Added mailfrom=xxxx to the callout verification option, for verify=
1280 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1281 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1283 12. Inserted some casts to stop certain compilers warning when using pointer
1284 differences as field lengths or precisions in printf-type calls (mostly
1285 affecting debugging statements).
1287 13. Added optional readline() support for -be (dynamically loaded).
1289 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1290 same clock tick as a message's arrival, so that its received time was the
1291 same as the "first fail" time on the retry record, and that message
1292 remained on the queue past the ultimate address timeout, every queue runner
1293 would try a delivery (because it was past the ultimate address timeout) but
1294 after another failure, the ultimate address timeout, which should have then
1295 bounced the address, did not kick in. This was a "< instead of <=" error;
1296 in most cases the first failure would have been in the next clock tick
1297 after the received time, and all would be well.
1299 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1300 being recognized when the domain list was tested by the match_domain
1301 condition in an expansion string.
1303 16. Added the ${str2b64: operator.
1305 17. Exim was always calling setrlimit() to set a large limit for the number of
1306 processes, without checking whether the existing limit was already
1307 adequate. (It did check for the limit on file descriptors.) Furthermore,
1308 errors from getrlimit() and setrlimit() were being ignored. Now they are
1309 logged to the main and panic logs, but Exim does carry on, to try to do its
1310 job under whatever limits there are.
1312 18. Imported PCRE 5.0.
1314 19. Trivial typo in log message " temporarily refused connection" (the leading
1317 20. If the log selector return_path_on_delivery was set and an address was
1318 redirected to /dev/null, the delivery process crashed because it assumed
1319 that a return path would always be set for a "successful" delivery. In this
1320 case, the whole delivery is bypassed as an optimization, and therefore no
1323 21. Internal re-arrangement: the function for sending a challenge and reading
1324 a response while authentication was assuming a zero-terminated challenge
1325 string. It's now changed to take a pointer and a length, to allow for
1326 binary data in such strings.
1328 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1330 23. Exim was not respecting finduser_retries when seeking the login of the
1331 uid under which it was called; it was always trying 10 times. (The default
1332 setting of finduser_retries is zero.) Also, it was sleeping after the final
1333 failure, which is pointless.
1335 24. Implemented tls_on_connect_ports.
1337 25. Implemented acl_smtp_predata.
1339 26. If the domain in control=submission is set empty, Exim assumes that the
1340 authenticated id is a complete email address when it generates From: or
1341 Sender: header lines.
1343 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1344 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1345 chown and chgrp in /bin and hostname in /usr/bin.
1347 28. Exim was keeping the "process log" file open after each use, just as it
1348 does for the main log. This opens the possibility of it remaining open for
1349 long periods when the USR1 signal hits a daemon. Occasional processlog
1350 errors were reported, that could have been caused by this. Anyway, it seems
1351 much more sensible not to leave this file open at all, so that is what now
1354 29. The long-running daemon process does not normally write to the log once it
1355 has entered its main loop, and it closes the log before doing so. This is
1356 so that log files can straightforwardly be renamed and moved. However,
1357 there are a couple of unusual error situations where the daemon does write
1358 log entries, and I had neglected to close the log afterwards.
1360 30. The text of an SMTP error response that was received during a remote
1361 delivery was being truncated at 512 bytes. This is too short for some of
1362 the long messages that one sometimes sees. I've increased the limit to
1365 31. It is now possible to make retry rules that apply only when a message has a
1366 specific sender, in particular, an empty sender.
1368 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1369 it possible to be selective about when SMTP synchronization is enforced.
1371 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1373 32. Implemented hosts_connection_nolog.
1375 33. Added an ACL for QUIT.
1377 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1380 35. Added mailbox_size and mailbox_filecount to appendfile.
1382 36. Added control = no_multiline_responses to ACLs.
1384 37. There was a bug in the logic of the code that waits for the clock to tick
1385 in the case where the clock went backwards by a substantial amount such
1386 that the microsecond fraction of "now" was more than the microsecond
1387 fraction of "then" (but the whole seconds number was less).
1389 38. Added support for the libradius Radius client library this is found on
1390 FreeBSD (previously only the radiusclient library was supported).
1396 1. When certain lookups returned multiple values in the form name=value, the
1397 quoting of the values was not always being done properly. Specifically:
1398 (a) If the value started with a double quote, but contained no whitespace,
1400 (b) If the value contained whitespace other than a space character (i.e.
1401 tabs or newlines or carriage returns) it was not quoted.
1402 This fix has been applied to the mysql and pgsql lookups by writing a
1403 separate quoting function and calling it from the lookup code. The fix
1404 should probably also be applied to nisplus, ibase and oracle lookups, but
1405 since I cannot test any of those, I have not disturbed their existing code.
1407 2. A hit in the callout cache for a specific address caused a log line with no
1408 reason for rejecting RCPT. Now it says "Previous (cached) callout
1409 verification failure".
1411 3. There was an off-by-one bug in the queryprogram router. An over-long
1412 return line was truncated at 256 instead of 255 characters, thereby
1413 overflowing its buffer with the terminating zero. As well as fixing this, I
1414 have increased the buffer size to 1024 (and made a note to document this).
1416 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1417 when Exim is waiting for an SMTP response from a remote server, Exim
1418 restarts its select() call on the socket, thereby resetting its timeout.
1419 This is not a problem when such interrupts are rare. Somebody set up a cron
1420 job to run exiwhat every 2 minutes, which is less than the normal select()
1421 timeout (5 or 10 minutes). This meant that the select() timeout never
1422 kicked in because it was always reset. I have fixed this by comparing the
1423 time when an interrupt arrives with the time at the start of the first call
1424 to select(). If more time than the timeout has elapsed, the interrupt is
1425 treated as a timeout.
1427 5. Some internal re-factoring in preparation for the addition of Sieve
1428 extensions (by MH). In particular, the "personal" test is moved to a
1429 separate function, and given an option for scanning Cc: and Bcc: (which is
1430 not set for Exim filters).
1432 6. When Exim created an email address using the login of the caller as the
1433 local part (e.g. when creating a From: or Sender: header line), it was not
1434 quoting the local part when it contained special characters such as @.
1436 7. Installed new OpenBSD configuration files.
1438 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1439 try to make them clearer.
1441 9. Callout options, other than the timeout value, were being ignored when
1442 verifying sender addresses in header lines. For example, when using
1444 verify = header_sender/callout=no_cache
1446 the cache was (incorrectly) being used.
1448 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1449 only the Cygwin environment.
1451 11. When return_path_on_delivery was set as a log selector, if different remote
1452 addresses in the same message used different return paths and parallel
1453 remote delivery occurred, the wrong values would sometimes be logged.
1454 (Whenever a remote delivery process finished, the return path value from
1455 the most recently started remote delivery process was logged.)
1457 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1458 lines when AUTH and/or TLS are in use. This is the "received protocol"
1459 field. Exim used to use "asmtp" for authenticated SMTP, without any
1460 indication (in the protocol name) for TLS use. Now it follows the RFC and
1461 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1462 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1463 names appear in log lines as well as in Received: header lines.
1465 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1466 extensions, and comparison tests, and to fix some bugs.
1468 14. Changes to the "personal" filter test:
1470 (1) The test was buggy in that it was just doing the equivalent of
1471 "contains" tests on header lines. For example, if a user's address was
1472 anne@some.where, the "personal" test would incorrectly be true for
1474 To: susanne@some.where
1476 This test is now done by extracting each address from the header in turn,
1477 and checking the entire address. Other tests that are part of "personal"
1478 are now done using regular expressions (for example, to check local parts
1479 of addresses in From: header lines).
1481 (2) The list of non-personal local parts in From: addresses has been
1482 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1483 taken from the Sieve specification recommendations.
1485 (3) If the message contains any header line starting with "List-" it is
1486 treated as non-personal.
1488 (4) The test for "circular" in the Subject: header line has been removed
1489 because it now seems ill-conceived.
1491 15. Minor typos in src/EDITME comments corrected.
1493 16. Installed latest exipick from John Jetmore.
1495 17. If headers_add on a router specified a text string that was too long for
1496 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1497 of string_sprintf() is now avoided.
1499 18. $message_body_size was not set (it was always zero) when running the DATA
1500 ACL and the local_scan() function.
1502 19. For the "mail" command in an Exim filter, no default was being set for
1503 the once_repeat time, causing a random time value to be used if "once" was
1504 specified. (If the value happened to be <= 0, no repeat happened.) The
1505 default is now 0s, meaning "never repeat". The "vacation" command was OK
1506 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1507 (I found it when inspecting the code).
1509 20. There is now an overall timeout for performing a callout verification. It
1510 defaults to 4 times the callout timeout, which applies to individual SMTP
1511 commands during the callout. The overall timeout applies when there is more
1512 than one host that can be tried. The timeout is checked before trying the
1513 next host. This prevents very long delays if there are a large number of
1514 hosts and all are timing out (e.g. when the network connections are timing
1515 out). The value of the overall timeout can be changed by specifying an
1516 additional sub-option for "callout", called "maxwait". For example:
1518 verify = sender/callout=5s,maxwait=20s
1520 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1521 to the end before writing, but this should make it even safer).
1523 22. Exim was forgetting that it had advertised PIPELINING for the second and
1524 subsequent messages on an SMTP connection. It was also not resetting its
1525 memory on STARTTLS and an internal HELO.
1527 23. When Exim logs an SMTP synchronization error within a session, it now
1528 records whether PIPELINING has been advertised or not.
1530 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1531 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1532 rather than long int.
1534 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1536 26. Added the never_mail option to autoreply.
1542 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1543 crash if the getsockname() call failed; this can happen if a connection is
1544 closed very soon after it is established. The problem was simply in the
1545 order in which certain operations were done, causing Exim to try to write
1546 to the SMTP stream before it had set up the file descriptor. The bug has
1547 been fixed by making things happen in the correct order.
1553 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1554 before the connection was closed, thus losing the rejection response.
1556 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1557 some early Solaris releases, but causes trouble in current releases where
1558 socklen_t is defined.
1560 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1563 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1566 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1567 that releasing the top of it at the end releases what was used for sub-
1568 expansions (unless the block got too big). However, discard this block if
1569 the first thing is a variable or header, so that we can use its block when
1570 it is dynamic (useful for very large $message_headers, for example).
1572 6. Lookups now cache *every* query, not just the most recent. A new, separate
1573 store pool is used for this. It can be recovered when all lookup caches are
1574 flushed. Lookups now release memory at the end of their result strings.
1575 This has involved some general refactoring of the lookup sources.
1577 7. Some code has been added to the store_xxx() functions to reduce the amount
1578 of flapping under certain conditions.
1580 8. log_incoming_interface used to affect only the <= reception log lines. Now
1581 it causes the local interface and port to be added to several more SMTP log
1582 lines, for example "SMTP connection from", and rejection lines.
1584 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1586 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1588 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1589 could be overwritten at the end of the current message (or the start of a
1590 new message if it was set in a HELO ACL). The value is now preserved for
1591 the duration of the SMTP connection.
1593 12. If a transport had a headers_rewrite setting, and a matching header line
1594 contained an unqualified address, that address was qualified, even if it
1595 did not match any rewriting rules. The underlying bug was that the values
1596 of the flags that permit the existence of unqualified sender and recipient
1597 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1598 non-local messages, and by -bnq for local messages) were not being
1599 preserved with the message after it was received.
1601 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1602 "next input=" as part of the text comprising the host identity instead of
1603 the correct text. The code was using the same buffer for two different
1604 strings. However, depending on which order the printing function evaluated
1605 its arguments, the bug did not always show up. Under Linux, for example, my
1606 test suite worked just fine.
1608 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1609 doesn't work with some older versions of Perl. It has been changed to "my",
1610 which in any case is probably the better facility to use.
1612 15. A really picky compiler found some instances of statements for creating
1613 error messages that either had too many or two few arguments for the format
1616 16. The size of the buffer for calls to the DNS resolver has been increased
1617 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1618 for addresses that have a lot of PTR records. This alleviates a problem; it
1619 does not fully solve it.
1621 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1622 buffer now truncates the list and logs the incident, which is the same
1623 action as happens when Exim is looking up a host name and its aliases.
1624 Previously in this situation something unpredictable would happen;
1625 sometimes it was "internal error: store_reset failed".
1627 18. If a server dropped the connection unexpectedly when an Exim client was
1628 using GnuTLS and trying to read a response, the client delivery process
1629 crashed while trying to generate an error log message.
1631 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1632 string, for example:
1634 warn message = H1: something\nH2: something
1636 the text was added as a single header line from Exim's point of view
1637 though it ended up OK in the delivered message. However, searching for the
1638 second and subsequent header lines using $h_h2: did not work. This has been
1639 fixed. Similarly, if a system filter added multiple headers in this way,
1640 the routers could not see them.
1642 20. Expanded the error message when iplsearch is called with an invalid key to
1643 suggest using net-iplsearch in a host list.
1645 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1646 ACLs are no longer actually imposed (and a message to that effect is
1649 22. If a "gecos" field in a passwd entry contained escaped characters, in
1650 particular, if it contained a \" sequence, Exim got it wrong when building
1651 a From: or a Sender: header from that name. A second bug also caused
1652 incorrect handling when an unquoted " was present following a character
1653 that needed quoting.
1655 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1656 was not being matched caselessly.
1658 24. Arranged for all hyphens in the exim.8 source to be escaped with
1661 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1662 properly. Recipient callout cache records were still being keyed to include
1663 the sender, even when use_sender was set false. This led to far more
1664 callouts that were necessary. The sender is no longer included in the key
1665 when use_sender is false.
1667 26. Added "control = submission" modifier to ACLs.
1669 27. Added the ${base62d: operator to decode base 62 numbers.
1671 28. dnsdb lookups can now access SRV records.
1673 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1674 the configuration file.
1676 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1677 (-v) mode. This makes the output for a verbose queue run more intelligible.
1679 31. Added a use_postmaster feature to recipient callouts.
1681 32. Added the $body_zerocount variable, containing the number of binary zero
1682 bytes in the message body.
1684 33. The time of last modification of the "new" subdirectory is now used as the
1685 "mailbox time last read" when there is a quota error for a maildir
1688 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1690 35. Added +ignore_unknown as a special item in host lists.
1692 36. Code for decoding IPv6 addresses in host lists is now included, even if
1693 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1694 address was recognized as an IP address, but was then not correctly decoded
1695 into binary, causing unexpected and incorrect effects when compared with
1702 1. Very minor rewording of debugging text in manualroute to say "list of
1703 hosts" instead of "hostlist".
1705 2. If verify=header_syntax was set, and a header line with an unqualified
1706 address (no domain) and a large number of spaces between the end of the
1707 name and the colon was received, the reception process suffered a buffer
1708 overflow, and (when I tested it) crashed. This was caused by some obsolete
1709 code that should have been removed. The fix is to remove it!
1711 3. When running in the test harness, delay a bit after writing a bounce
1712 message to get a bit more predictability in the log output.
1714 4. Added a call to search_tidyup() just before forking a reception process. In
1715 theory, someone could use a lookup in the expansion of smtp_accept_max_
1716 per_host which, without the tidyup, could leave open a database connection.
1718 5. Added the variables $recipient_data and $sender_data which get set from a
1719 lookup success in an ACL "recipients" or "senders" condition, or a router
1720 "senders" option, similar to $domain_data and $local_part_data.
1722 6. Moved the writing of debug_print from before to after the "senders" test
1725 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1726 problems for message scanning, either using a data ACL, or using
1727 local_scan() because the Received: header was not generated till after they
1728 were called (in order to set the time as the time of reception completion).
1729 I have revised the way this works. The header is now generated after the
1730 body is received, but before the ACL or local_scan() are called. After they
1731 are run, the timestamp in the header is updated.
1737 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1738 before starting a queue runner without re-exec. This happened only when
1739 deliver_drop_privilege was set or when the Exim user was set to root. The
1740 effect of the bug was that timeouts during subsequent deliveries caused
1741 crashes instead of being properly handled. The handler is now left at its
1742 default (and expected) setting.
1744 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1745 message, again when deliver_drop_privilege is set or Exim is run as root.
1746 The bug described in (1) was not present in this case, but the tidying up
1747 of the other signals was missing. I have made the two cases consistent.
1749 3. The ignore_target_hosts setting on a manualroute router was being ignored
1750 for hosts that were looked up using the /MX notation.
1752 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1755 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1756 operated on the sender address. After changing the $sender_address to <>
1757 for the sender address verify, Exim was re-instated it as the original
1758 (before rewriting) address, but remembering that it had rewritten it, so it
1759 wasn't rewriting it again. This bug also had the effect of breaking the
1760 sender address verification caching when the sender address was rewritten.
1762 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1763 This has been changed so that if the ip literal address matches
1764 ignore_target_hosts, the router declines.
1766 7. Added expansion conditions match_domain, match_address, and match_local_
1767 part (NOT match_host).
1769 8. The placeholder for the Received: header didn't have a length field set.
1771 9. Added code to Exim itself and to exim_lock to test for a specific race
1772 condition that could lead to file corruption when using MBX delivery. The
1773 issue is with the lockfile that is created in /tmp. If this file is removed
1774 after a process has opened it but before that process has acquired a lock,
1775 there is the potential for a second process to recreate the file and also
1776 acquire a lock. This could lead to two Exim processes writing to the file
1777 at the same time. The added code performs the same test as UW imapd; it
1778 checks after acquiring the lock that its file descriptor still refers to
1779 the same named file.
1781 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1782 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1785 11. Added the smtp_active_hostname option. If used, this will typically be made
1786 to depend on the incoming interface address. Because $interface_address is
1787 not set up until the daemon has forked a reception process, error responses
1788 that can happen earlier (such as "too many connections") no longer contain
1791 12. If an expansion in a condition on a "warn" statement fails because a lookup
1792 defers, the "warn" statement is abandoned, and the next ACL statement is
1793 processed. Previously this caused the whole ACL to be aborted.
1795 13. Added the iplsearch lookup type.
1797 14. Added ident_timeout as a log selector.
1799 15. Added tls_certificate_verified as a log selector.
1801 16. Added a global option tls_require_ciphers (compare the smtp transport
1802 option of the same name). This controls incoming TLS connections.
1804 17. I finally figured out how to make tls_require_ciphers do a similar thing
1805 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1806 before starting the TLS session.
1808 18. Tabs are now shown as \t in -bP output.
1810 19. If the log selector return_path_on_delivery was set, Exim crashed when
1811 bouncing a message because it had too many Received: header lines.
1813 20. If two routers both had headers_remove settings, and the first one included
1814 a superfluous trailing colon, the final name in the first list and the
1815 first name in the second list were incorrectly joined into one item (with a
1816 colon in the middle).
1822 1. Added -C and -D options to the exinext utility, mainly to make it easier
1823 to include in the automated testing, but these could be helpful when
1824 multiple configurations are in use.
1826 2. The exinext utility was not formatting the output nicely when there was
1827 an alternate port involved in the retry record key, nor when there was a
1828 message id as well (for retries that were specific to a specific message
1829 and a specific host). It was also confused by IPv6 addresses, because of
1830 the additional colons they contain. I have fixed the IPv4 problem, and
1831 patched it up to do a reasonable job for IPv6.
1833 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1834 delivery, the log line now contains "pipelined" if PIPELINING was used.
1836 4. An SMTP transport process used to panic and die if the bind() call to set
1837 an explicit outgoing interface failed. This has been changed; it is now
1838 treated in the same way as a connect() failure.
1840 5. A reference to $sender_host_name in the part of a conditional expansion
1841 that was being skipped was still causing a DNS lookup. This no longer
1844 6. The def: expansion condition was not recognizing references to header lines
1845 that used bh_ and bheader_.
1847 7. Added the _cache feature to named lists.
1849 8. The code for checking quota_filecount in the appendfile transport was
1850 allowing one more file than it should have been.
1852 9. For compatibility with Sendmail, the command line option
1860 and sets the incoming protocol and host name (for trusted callers). The
1861 host name and its colon can be omitted when only the protocol is to be set.
1862 Note the Exim already has two private options, -pd and -ps, that refer to
1863 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1864 "s", but I don't think that's a major issue.
1866 10. A number of refactoring changes to the code, none of which should affect
1869 (a) The number of logging options was getting close to filling up the
1870 32-bit word that was used as a bit map. I have split them into two classes:
1871 those that are passed in the argument to log_write(), and those that are
1872 only ever tested independently outside of that function. These are now in
1873 separate 32-bit words, so there is plenty of room for expansion again.
1874 There is no change in the user interface or the logging behaviour.
1876 (b) When building, for example, log lines, the code previously used a
1877 macro that called string_cat() twice, in order to add two strings. This is
1878 not really sufficiently general. Furthermore, there was one instance where
1879 it was actually wrong because one of the argument was used twice, and in
1880 one call a function was used. (As it happened, calling the function twice
1881 did not affect the overall behaviour.) The macro has been replaced by a
1882 function that can join an arbitrary number of extra strings onto a growing
1885 (c) The code for expansion conditions now uses a table and a binary chop
1886 instead of a serial search (which was left over from when there were very
1887 few conditions). Also, it now recognizes conditions like "pam" even when
1888 the relevant support is not compiled in: a suitably worded error message is
1889 given if an attempt is made to use such a condition.
1891 11. Added ${time_interval:xxxxx}.
1893 12. A bug was causing one of the ddress fields not to be passed back correctly
1894 from remote delivery subprocesses. The field in question was not being
1895 subsequently used, so this caused to problems in practice.
1897 13. Added new log selectors queue_time and deliver_time.
1899 14. Might have fixed a bug in maildirsizefile handling that threw up
1900 "unexpected character" debug warnings, and recalculated the data
1901 unnecessarily. In any case, I expanded the warning message to give more
1904 15. Added the message "Restricted characters in address" to the statements in
1905 the default ACL that block characters like @ and % in local parts.
1907 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1908 Three changes have been made:
1910 (a) There was a serious bug; a negative response to MAIL caused the whole
1911 recipient domain to be cached as invalid, thereby blocking all messages
1912 to all local parts at the same domain, from all senders. This bug has
1913 been fixed. The domain is no longer cached after a negative response to
1914 MAIL if the sender used is not empty.
1916 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1919 (c) A new callout option, "use_sender" has been added for people who want
1920 the modified behaviour.
1926 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1929 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1930 indeed breaks things for older releases.
1932 3. Added additional logging to the case where there is a problem reading data
1933 from a filter that is running in a subprocess using a pipe, in order to
1934 try to track down a specific problem.
1936 4. Testing facility fudge: when running in the test harness and attempting
1937 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1938 getting "No route to host". Convert this to a timeout.
1940 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1943 6. Some OS don't have socklen_t but use size_t instead. This affects the
1944 fifth argument of getsockopt() amongst other things. This is now
1945 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1946 can be set for individual OS. I have set it for SunOS5, OSF1, and
1947 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1948 some earlier ones do not.
1950 7. Change 4.30/15 was not doing the test caselessly.
1952 8. The standard form for an IPv6 address literal was being rejected by address
1953 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1954 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1955 this, as well as the form without the "IPv6" on the front (but only when
1956 address literals are enabled, of course).
1958 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1960 10. Exim crashed if a message with an empty sender address specified by -f
1961 encountered a router with an errors_to setting. This could be provoked only
1962 by a command such as
1966 where an empty string was supplied; "<>" did not hit this bug.
1968 11. Installed PCRE release 4.5.
1970 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1971 remained set. It is now erased.
1973 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1974 times from message ids (which are base 36 rather than the normal 62).
1976 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1977 were being counted as actual protocol errors, and logged if the log
1978 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1979 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1980 and DATA following a set of rejected RCPTs do not count as protocol errors.
1981 In other words, Exim assumes they were pipelined, though this may not
1982 actually be the case. Of course, in all cases the client gets an
1983 appropriate error code.
1985 15. If a lookup fails in an ACL condition, a message about the failure may
1986 be available; it is used if testing the ACL cannot continue, because most
1987 such messages specify what the cause of the deferral is. However, some
1988 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1989 that caused an old message to be retained and used if a later statement
1990 caused a defer, replacing the real cause of the deferral.
1992 16. If an IP address had so many PTR records that the DNS lookup buffer
1993 was not large enough to hold them, Exim could crash while trying to process
1994 the truncated data. It now detects and logs this case.
1996 17. Further to 4.21/58, another change has been made: if (and only if) the
1997 first line of a message (the first header line) ends with CRLF, a bare LF
1998 in a subsequent header line has a space inserted after it, so as not to
1999 terminate the header.
2001 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2002 unnecessarily, used atoi() instead of strtol(), and didn't check the
2003 termination when getting file sizes from file names by regex.
2005 19. Completely re-implemented the support for maildirsize files, in the light
2006 of a number of problems with the previous contributed implementation
2007 (4.30/29). In particular:
2009 . If the quota is zero, the maildirsize file is maintained, but no quota is
2012 . If the maildir directory does not exist, it is created before any attempt
2013 to write a maildirsize file.
2015 . The quota value in the file is just a cache; if the quota is changed in
2016 the transport, the new value overrides.
2018 . A regular expression is available for excluding directories from the
2021 20. The autoreply transport checks the characters in options that define the
2022 message's headers; it allows continued headers, but it was checking with
2023 isspace() after an embedded newline instead of explicitly looking for a
2026 21. If all the "regular" hosts to which an address was routed had passed their
2027 expiry times, and had not reached their retry times, the address was
2028 bounced, even if fallback hosts were defined. Now Exim should go on to try
2031 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2032 equivalent code in the SMTP transport. Some hosts send humungous responses
2033 to HELO/EHLO, more than 1024 it seems.
2035 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2036 gives compiler warnings in some environments. I've now done it "properly",
2039 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2040 (because of problems with the built-in one) was declared to return uschar *
2041 instead of char *, causing compiler failure.
2043 25. Fixed a file descriptor leak when processing alias/forward files.
2045 26. Fixed a minor format string issue in dbfn.c.
2047 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2049 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2050 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2053 29. When a sender address is verified, it is cached, to save repeating the test
2054 when there is more than one recipient in a message. However, when the
2055 verification involves a callout, it is possible for different callout
2056 options to be set for different recipients. It is too complicated to keep
2057 track of this in the cache, so now Exim always runs a verification when a
2058 callout is required, relying on the callout cache for the optimization.
2059 The overhead is duplication of the address routing, but this should not be
2062 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2063 to be verified with callout=postmaster, and the main callout worked but the
2064 postmaster check failed, the verification correctly failed. However, if a
2065 subsequent RCPT command asked for sender verification *without* the
2066 postmaster check, incorrect caching caused this verification also to fail,
2069 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2070 it was not caching the DNS options (qualify_single, search_parents) that
2071 were used when the lookup failed. A subsequent lookup with different
2072 options therefore always gave the same answer, though there were cases
2073 where it should not have. (Example: a "domains = !$mx_any" option on a
2074 dnslookup router: the "domains" option is always processed without any
2075 widening, but the router might have qualify_single set.) Now Exim uses the
2076 cached value only when the same options are set.
2078 32. Added John Jetmore's "exipick" utility to the distribution.
2080 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2081 than a timeout (e.g. a certificate is required, and is not provided), an
2082 Exim server now closes the connection immediately. Previously it waited for
2083 the client to close - but if the client is SSL, it seems that they each
2084 wait for each other, leading to a delay before one of them times out.
2086 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2087 maintained 0.8.x compatibility because I don't think many are using it, and
2088 it is clearly obsolete.
2090 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2093 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2094 client certificate was expired. A simple patch fixes this, though I don't
2095 understand the full logic of why the verify callback is called multiple
2098 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2099 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2100 which causes problems with some clients (such as the Certicom SSL Plus
2101 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2102 disables the coutermeasure allowing Eudora to connect."
2104 38. Exim was not checking that a write() to a log file succeeded. This could
2105 lead to Bad Things if a log got too big, in particular if it hit a file
2106 size limit. Exim now panics and dies if it cannot write to a log file, just
2107 as it does if it cannot open a log file.
2109 39. Modified OS/Makefile-Linux so that it now contains
2111 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2113 The two -D definitions ensure that Exim is compiled with large file
2114 support, which makes it possible to handle log files that are bigger than
2117 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2118 instance) a domain was checked against a named list that involved a lookup,
2119 causing $domain_data to be set, then another domain was checked against the
2120 same list, then the first domain was re-checked, the value of $domain_data
2121 after the final check could be wrong. In particular, if the second check
2122 failed, it could be set empty. This bug probably also applied to
2125 41. The strip_trailing_dot option was not being applied to the address given
2126 with the -f command-line option.
2128 42. The code for reading a message's header from the spool was incrementing
2129 $received_count, but never initializing it. This meant that the value was
2130 incorrect (doubled) while delivering a message in the same process in which
2131 it was received. In the most common configuration of Exim, this never
2132 happens - a fresh exec is done - but it can happen when
2133 deliver_drop_privilege is set.
2135 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2136 it now includes up to 150 characters of the unexpected data in the log
2139 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2140 and building data strings. The size of both of these buffers was 10 000
2141 bytes - far larger than anybody would *ever* want, thought I. Needless to
2142 say, somebody hit the limit. I have increased the maximum line length to
2143 20 000 and the maximum data length of concatenated lines to 100 000. I have
2144 also fixed two bugs, because there was no checking on these buffers. Tsk,
2145 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2146 buffer is too small.
2148 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2149 lsearch lookups. Now it does.
2151 46. When parsing a route_list item in a manualroute router, a fixed-length
2152 buffer was used for the list of hosts. I made this 1024 bytes long,
2153 thinking that nobody would ever have a list of hosts that long. Wrong.
2154 Somebody had a whole pile of complicated expansion conditions, and the
2155 string was silently truncated, leading to an expansion error. It turns out
2156 that it is easier to change to an unlimited length (owing to other changes
2157 that have happened since this code was originally written) than to build
2158 structure for giving a limitation error. The length of the item that
2159 expands into the list of hosts is now unlimited.
2161 47. The lsearch lookup could not handle data where the length of text line was
2162 more than 4095 characters. Such lines were truncated, leading to shortened
2163 data being returned. It should now handle lines of any length.
2165 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2166 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2169 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2170 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2171 no longer happens in comment lines. A list of replacements is now placed
2172 at the head of all of the source files, except those whose only change is
2173 to replace PERL_COMMAND in the very first #! line.
2175 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2176 messages on the queue, with a bottom-up merge sort, using code contributed
2177 by Michael Haardt. This should make operations like -bp somewhat faster on
2178 large queues. It won't affect queue runners, except when queue_run_in_order
2181 51. Installed eximstats 1.31 in the distribution.
2183 52. Added support for SRV lookups to the dnslookup router.
2185 53. If an ACL referred to $message_body or $message_body_end, the value was not
2186 reset for any messages that followed in the same SMTP session.
2188 54. The store-handling optimization for building very long strings was not
2189 differentiating between the different store pools. I don't think this
2190 actually made any difference in practice, but I've tidied it.
2192 55. While running the routers to verify a sender address, $sender_address
2193 was still set to the sender address. This is wrong, because when routing to
2194 send a bounce to the sender, it would be empty. Therefore, I have changed
2195 it so that, while verifying a sender address, $sender_address is set to <>.
2196 (There is no change to what happens when verifying a recipient address.)
2198 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2199 target A or AAAA records (if not already returned) without resetting the
2200 qualify_single or search_parents options of the DNS resolver. These are
2201 inappropriate in this case because the targets of MX and SRV records must
2202 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2203 target that, when qualified, matched something in the local domain. These
2204 two options are now turned off when doing these lookups.
2206 57. It seems that at least some releases of Reiserfs (which does not have the
2207 concept of a fixed number of inodes) returns zero and not -1 for the
2208 number of available inodes. This interacted badly with check_spool_inodes,
2209 which assumed that -1 was the "no such thing" setting. What I have done is
2210 to check that the total number of inodes is greater than zero before doing
2211 the test of how many are available.
2213 58. When a "warn" ACL statement has a log_message modifier, the message is
2214 remembered, and not repeated. This is to avoid a lot of repetition when a
2215 message has many recipients that cause the same warning to be written.
2216 Howewer, Exim was preserving the list of already written lines for an
2217 entire SMTP session, which doesn't seem right. The memory is now reset if a
2218 new message is started.
2220 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2221 debugging output unless log_rewrite was also set.
2223 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2224 of (int)(handle) when we know that handle contains (void *)(-1).
2226 61. The Exim daemon panic-logs an error return when it closes the incoming
2227 connection. However "connection reset by peer" seems to be common, and
2228 isn't really an error worthy of noting specially, so that particular error
2231 62. When Exim is trying to find all the local interfaces, it used to panic and
2232 die if the ioctl to get the interface flags failed. However, it seems that
2233 on at least one OS (Solaris 9) it is possible to have an interface that is
2234 included in the list of interfaces, but for which you get a failure error
2235 for this call. This happens when the interface is not "plumbed" into a
2236 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2237 failure of the "get flags" call assumes that the interface is down.
2239 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2240 makes life easier for people who are doing arithmetic on fields extracted
2241 from dates, where you often get leading zeros that should not be
2242 interpreted as octal.
2244 64. Added qualify_domain to the redirect router, to override the global
2247 65. If a pathologically long header line contained very many addresses (the
2248 report of this problem mentioned 10 000) and each of them was rewritten,
2249 Exim could use up a very large amount of memory. (It kept on making new
2250 copies of the header line as it rewrote, and never released the old ones.)
2251 At the expense of a bit more processing, the header rewriting function has
2252 been changed so that it no longer eats memory in this way.
2254 66. The generation of the Received: header has been moved from the time that a
2255 message starts to be received, to the time that it finishes. The timestamp
2256 in the Received: header should now be very close to that of the <= log
2257 line. There are two side-effects of this change:
2259 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2260 logged header lines no longer include the local Received: line, because
2261 it has not yet been created. The same applies to a copy of the message
2262 that is returned to a non-SMTP sender when a message is rejected.
2264 (b) When a filter file is tested using -bf, no additional Received: header
2265 is added to the test message. After some thought, I decided that this
2268 This change does not affect the value of $received_for. It is still set
2269 after address rewriting, but before local_scan() is called.
2271 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2273 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2274 gave an unhelpful panic error message, and a defer error. I have managed to
2275 change this behaviour so that it now rejects any supplied certificate,
2276 which seems right, as the list of acceptable certificates is empty.
2278 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2279 gave an unhelpful defer error. I have not managed to make this reject any
2280 supplied certificates, but the error message it gives is "no certificate
2281 supplied", which is not helpful.
2283 70. exigrep's output now also includes lines that are not associated with any
2284 message, but which match the given pattern. Implemented by a patch from
2285 Martin Sluka, which also tidied up the Perl a bit.
2287 71. Recipient callout verification, like sender verification, was using <> in
2288 the MAIL FROM command. This isn't really the right thing, since the actual
2289 sender may affect whether the remote host accepts the recipient or not. I
2290 have changed it to use the actual sender in the callout; this means that
2291 the cache record is now keyed on a recipient/sender pair, not just the
2292 recipient address. There doesn't seem to be a real danger of callout loops,
2293 since a callout by the remote host to check the sender would use <>.
2294 [SEE ABOVE: changed after hitting problems.]
2296 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2297 temporary errors. However, in the case of such a code being given after
2298 the end of a data transmission (i.e. after ".") Exim was failing to write
2299 a retry record for the message. (Yes, there was some broken host that was
2300 actually sending 8xx at this point.)
2302 73. An unknown lookup type in a host list could cause Exim to panic-die when
2303 the list was checked. (An example that provoked this was putting <; in the
2304 middle of a list instead of at the start.) If this happened during a DATA
2305 ACL check, a -D file could be left lying around. This kind of configuration
2306 error no longer causes Exim to die; instead it causes a defer errror. The
2307 incident is still logged to the main and panic logs.
2309 74. Buglet left over from Exim 3 conversion. The message "too many messages
2310 in one connection" was written to the rejectlog but not the mainlog, except
2311 when address rewriting (yes!) was being logged.
2313 75. Added write_rejectlog option.
2315 76. When a system filter was run not as root (that is, when system_filter_user
2316 was set), the values of the $n variables were not being returned to the
2317 main process; thus, they were not subsequently available in the $sn
2320 77. Added +return_path_on_delivery log selector.
2322 78. A connection timeout was being treated differently from recipients deferred
2323 when testing hosts_max_try with a message that was older than the host's
2324 retry timeout. (The host should not be counted, thus allowing all hosts to
2325 be tried at least once before bouncing.) This may have been the cause of an
2326 occasionally reported bug whereby a message would remain on the queue
2327 longer than the retry timeout, but would be bounced if a delivery was
2328 forced. I say "may" because I never totally pinned down the problem;
2329 setting up timeout/retry tests is difficult. See also the next item.
2331 79. The ultimate address timeout was not being applied to errors that involved
2332 a combination of host plus message (for example, a timeout on a MAIL
2333 command). When an address resolved to a number of possible hosts, and they
2334 were not all tried for each delivery (e.g. because of hosts_max_try), a
2335 message could remain on the queue longer than the retry timeout.
2337 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2340 81. Fixed an obscure SMTP outgoing bug which required at least the following
2341 conditions: (a) there was another message waiting for the same server;
2342 (b) the server returned 5xx to all RCPT commands in the first message so
2343 that the message was not completed; (c) the server dropped the connection
2344 or gave a negative response to the RSET that Exim sends to abort the
2345 transaction. The observed case was a dropped connection after DATA that had
2346 been sent in pipelining mode. That is, the server had advertised PIPELINING
2347 but was not implementing it correctly. The effect of the bug was incorrect
2348 behaviour, such as trying another host, and this could lead to a crash.
2354 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2355 and daemon.c were passed as pointers to ints; they should have been
2356 pointers to socklen_t variables (which are typically unsigned ints).
2358 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2361 3. Fixed a really odd bug that affected only the testing scheme; patching a
2362 certain fixed string in the binary changed the value of another string that
2363 happened to be identical to the end of the original first string.
2365 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2366 name", it returns that address as the IP address. On some operating
2367 systems (e.g. Solaris), it also passes back the IP address string as the
2368 "host name". However, on others (e.g. Linux), it passes back an empty
2369 string. Exim wasn't checking for this, and was changing the host name to an
2370 empty string, assuming it had been canonicized.
2372 5. Although rare, it is permitted to have more than one PTR record for a given
2373 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2374 all the names associated with an address, because they do in Solaris.
2375 However, it seems that they do not in Linux for data that comes from the
2376 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2377 I found this out when I moved to a new Linux workstation and tried to run
2378 the Exim test suite.
2380 To get round this problem I have changed the code so that it now does its
2381 own call to the DNS to look up PTR records when searching for a host name.
2382 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2383 addresses that are only in /etc/hosts are still found.
2385 This behaviour is, however, controlled by an option called host_lookup_
2386 order, which defaults to "bydns:byaddr". If people want to use the other
2387 order, or indeed, just use one or the other means of lookup, they can
2388 specify it in this variable.
2390 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2391 some operating systems, this comes back from gethostbyaddr() as an empty
2392 string, and this is what Exim used to test for. However, it seems that in
2393 other systems, "." is yielded. Exim now tests for this case too.
2395 7. The values of check_spool_space and check_log_space are now held internally
2396 as a number of kilobytes instead of an absolute number of bytes. If a
2397 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2398 kilobyte. This means that much larger values can be stored.
2400 8. Exim monitor: an attempt to get the action menu when not actually pointing
2401 at a message produces an empty menu entitled "No message selected". This
2402 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2403 no entries in it ("Shell widget menu has zero width and/or height"). So I
2404 have added a single, blank menu entry in this case.
2406 9. Added ${quote_local_part.
2408 10. MIME decoding is now applied to the contents of Subject: header lines when
2411 11. Now that a reference to $sender_host_address automatically causes a reverse
2412 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2413 host lookup before query-style lookups in lists that might use this
2414 variable. This has therefore been abolished, and the "net-" prefix is no
2415 longer necessary for query-style lookups.
2417 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2418 have been a typo for LFLAGS, so it has been changed.
2420 13. The install script calls Exim with "-C /dev/null" in order to find the
2421 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2422 to be output. Howeve, since Exim outputs its version number before the
2423 error, it didn't break the script. It just looked ugly. I fixed this by
2424 always allowing "-C /dev/null" if the caller is root.
2426 14. Ignore overlarge ACL variable number when reading spool file - insurance
2427 against a later release with more variables having written the file.
2429 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2430 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2431 this, as well as the form without the "IPv6" on the front.
2433 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2434 OS/Makefile-Darwin file.
2436 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2437 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2439 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2440 variable. However, some OS use ints for inodes. I've added cast to long int
2441 to get rid of the compiler warning.
2443 19. I had forgotten to lock out "/../" in configuration file names when
2444 ALT_CONFIG_PREFIX was set.
2446 20. Routers used for verification do not need to specify transports. However,
2447 if such a router generated a host list, and callout was configured, Exim
2448 crashed, because it could not find a port number from the (non-existent)
2449 transport. It now assumes port 25 in this circumstance.
2451 21. Added the -t option to exigrep.
2453 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2454 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2455 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2456 them is set, LOOKUP_LSEARCH is forced.
2458 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2460 24. Added sender and host information to the "rejected by local_scan()" log
2461 line; previously there was no indication of these.
2463 25. Added .include_if_exists.
2465 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2466 Linux. It turns out that not all file systems support this. Apparently some
2467 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2468 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2469 not supported on the file descriptor, is now ignored when Exim is trying to
2470 sync a directory. This applies only to Linux.
2472 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2474 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2475 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2478 29. Added support for maildirsize files from supplied patch (modified a bit).
2480 30. The use of :fail: followed by an empty string could lead Exim to respond to
2481 sender verification failures with (e.g.):
2483 550 Verification failed for <xxx>
2484 550 Sender verify failed
2486 where the first response line was missing the '-' that indicates it is not
2487 the final line of the response.
2489 31. The loop for finding the name of the user that called Exim had a hardwired
2490 limit of 10; it now uses the value of finduser_retries, which is used for
2491 all other user lookups.
2493 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2496 33. Exim was neglecting to zero errno before one call of strtol() when
2497 expanding a string and expecting an integer value. On some systems this
2498 resulted in spurious "integer overflow" errors. Also, it was casting the
2499 result into an int without checking.
2501 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2502 did not work. The code looks as if it has *never* worked, though it appears
2503 to have been documented since at least releast 1.62. I have made it work.
2505 35. The "timeout_DNS" error in retry rules, also documented since at least
2506 1.62, also never worked. As it isn't clear exactly what this means, and
2507 clearly it isn't a major issue, I have abolished the feature by treating it
2508 as "timeout", and writing a warning to the main and panic logs.
2510 36. The display of retry rules for -brt wasn't always showing the error code
2513 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2514 timeout_connect_A, timeout_connect_MX.
2516 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2517 to the empty sender.
2519 39. The daemon was not analysing the content of -oX till after it had closed
2520 stderr and disconnected from the controlling terminal. This meant that any
2521 syntax errors were only noted on the panic log, and the return code from
2522 the command was 0. By re-arranging the code a little, I've made the
2523 decoding happen first, so such errors now appear on stderr, and the return
2524 code is 1. However, the actual setting up of the sockets still happens in
2525 the disconnected process, so errors there are still only recorded on the
2528 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2529 connections (as happens on some IP stacks) was logged at start up time as
2530 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2531 it from "IPv6 and IPv4", which means that two separate sockets are being
2534 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2535 says whether AF_INET or AF_INET6 was passed as an argument.
2537 42. Exiwhat output was messed up when time zones were included in log
2540 43. Exiwhat now gives more information about the daemon's listening ports,
2541 and whether -tls-on-connect was used.
2543 44. The "port" option of the smtp transport is now expanded.
2545 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2546 silently ignored. Now an error message is written to the main and panic
2549 46. There's a new ACL modifier called "logwrite" which writes to a log file
2550 as soon as it is encountered.
2552 47. Added $local_user_uid and $local_user_gid at routing time.
2554 48. Exim crashed when trying to verify a sender address that was being
2557 49. Exim was recognizing only a space character after ".include". It now also
2558 recognizes a tab character.
2560 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2561 extracting the relevant information from the specification. The man page no
2562 longer contains scrambled data for the -d option, and I've added a section
2563 at the front about calling Exim under different names.
2565 51. Added "extra_headers" argument to the "mail" command in filter files.
2567 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2570 53. Installed eximstats 1.29.
2572 54. Added transport_filter_timeout as a generic transport option.
2574 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2575 Cc: header lines. This was required by RFC 822, but it not required by RFC
2578 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2579 incoming messages that did not have them. Now it does so only if the
2580 message originates locally, that is, if there is no associated remote host
2581 address. When Resent- header lines are present, this applies to the Resent-
2582 lines rather than the non-Resent- lines.
2584 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2585 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2587 58. Messages for configuration errors now include the name of the main
2588 configuration file - useful now that there may be more than one file in a
2589 list (.included file names were always shown).
2591 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2592 for those rare installations that do not start the daemon as root or run it
2593 setuid root. I've cut out the call to initgroups() if the daemon is not
2596 60. The Exim user and group can now be bound into the binary as text strings
2597 that are looked up at the start of Exim's processing.
2599 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2601 62. Added $mailstore_basename variable.
2603 63. Installed patch to sieve.c from Michael Haardt.
2605 64. When Exim failed to open the panic log after failing to open the main log,
2606 the original message it was trying to log was written to stderr and debug
2607 output, but if they were not available (the usual case in production), it
2608 was lost. Now it is written to syslog before the two lines that record the
2609 failures to open the logs.
2611 65. Users' Exim filters run in subprocesses under the user's uid. It is
2612 possible for a "deliver" command or an alias in a "personal" command to
2613 provoke an address rewrite. If logging of address rewriting is configured,
2614 this fails because the process is not running as root or exim. There may be
2615 a better way of dealing with this, but for the moment (because 4.30 needs
2616 to be released), I have disabled address rewrite logging when running a
2617 filter in a non-root, non-exim process.
2623 1. The buildconfig auxiliary program wasn't quoting the value set for
2624 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2625 not defined. This bug was masked in 4.22 by the effect that was fixed in
2628 2. Some messages that were rejected after a message id was allocated were
2629 shown as "incomplete" by exigrep. It no longer does this for messages that
2630 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2632 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2633 have allow_domain_literals set, the ID did not get logged in the <= line.
2634 Domain literals are now always recognized in Message-ID: header lines.
2636 4. The first argument for a ${extract expansion item is the key name or field
2637 number. Leading and trailing spaces in this item were not being ignored,
2638 causing some misleading effects.
2640 5. When deliver_drop_privilege was set, single queue runner processes started
2641 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2642 same command in the process it spins off) were not dropping privilege.
2644 6. When the daemon running as "exim" started a queue runner, it always
2645 re-executed Exim in the spun-off process. This is a waste of effort when
2646 deliver_drop_privilege is set. The new process now just calls the
2647 queue-runner function directly.
2653 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2656 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2657 ignored. Though the use of -oP was forcing the writing of a pid file, it
2658 was always written to the default place.
2660 3. If the message "no IP address found for host xxxx" is generated during
2661 incoming verification, it is now followed by identification of the incoming
2662 connection (so you can more easily find what provoked it).
2664 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2666 5. Added some features to "harden" Exim a bit more against certain attacks:
2668 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2669 be put in Local/Makefile. This is like the never_users runtime option,
2670 but it cannot be overridden. The default setting is "root".
2672 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2673 prefix string with which any file named in a -C command line option
2676 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2677 is retained for -C and -D only if the caller of Exim is root. Without
2678 it, the exim user may also use -C and -D and retain privilege.
2680 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2681 command line option is disabled.
2683 6. Macro names set by the -D option must start with an upper case letter, just
2684 like macro names defined in the configuration file.
2686 7. Added "dereference=" facility to LDAP.
2688 8. Two instances of the typo "uknown" in the source files are fixed.
2690 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2691 the Configure-Makefile script screwed up while processing it.
2693 10. Incorporated PCRE 4.4.
2695 11. The SMTP synchronization check was not operating right at the start of an
2696 SMTP session. For example, it could not catch a HELO sent before the client
2697 waited for the greeting. There is now a check for outstanding input at the
2698 point when the greeting is written. Because of the duplex, asynchronous
2699 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2700 way, but not yet received, when the check is performed.
2702 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2703 on TCP/IP sockets, because this apparently causes some broken clients to
2706 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2707 unchanged) from the Cygwin maintainer.
2709 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2710 was supported instead of testing for mbx. Effectively a typo.
2712 15. The spa authenticator server code was not checking that the input it
2713 received was valid base64.
2715 16. The debug output line for the "set" modifier in ACLs was not showing the
2716 name of the variable that was being set.
2718 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2720 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2721 lookup on its own. Something else was needed to trigger the lookup. For
2722 example, a match in host_lookup or the need for a host name in a host list.
2723 Now, if $sender_host_name is referenced and the host name has not yet been
2724 looked up, a lookup is performed. If the lookup fails, the variable remains
2725 empty, and $host_lookup_failed is set to "1".
2727 19. Added "eqi" as a case-independent comparison operator.
2729 20. The saslauthd authentication condition could segfault if neither service
2730 nor realm was specified.
2732 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2733 error message that was logged was misleading, and incoming SMTP
2734 connections were dropped. The message is now more accurate, and temporary
2735 errors are given to SMTP connections.
2737 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2738 (except RSET) with a 421 error, until QUIT is received. However, it was
2739 failing to send a response to QUIT.
2741 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2742 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2743 ACL is now run after the helo_try_verify_hosts code.
2745 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2746 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2747 case-independent, so other case variants are also recognized). Apparently
2748 some systems use these upper case variants.
2750 25. If more than two messages were waiting for the same host, and a transport
2751 filter was specified for the transport, Exim sent two messages over the
2752 same TCP/IP connection, and then failed with "socket operation on non-
2753 socket" when it tried to send the third.
2755 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2757 27. The extern definition of crypt16() in expand.c was not being excluded when
2758 the OS had its own crypt16() function.
2760 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2761 as a preferred synonym for return_size_limit, both as an option and as an
2764 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2766 30. Changed the default configuration ACL to relax the local part checking rule
2767 for addresses that are not in any local domains. For these addresses,
2768 slashes and pipe symbols are allowed within local parts, but the sequence
2769 /../ is explicitly forbidden.
2771 31. SPA server authentication was not clearing the challenge buffer before
2774 32. log_message in a "warn" ACL statement was writing to the reject log as
2775 well as to the main log, which contradicts the documentation and doesn't
2776 seem right (because no rejection is happening). So I have stopped it.
2778 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2779 However, I am unable to do any testing of this.
2781 34. Fixed an infelicity in the appendfile transport. When checking directories
2782 for a mailbox, to see if any needed to be created, it was accidentally
2783 using path names with one or more superfluous leading slashes; tracing
2784 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2786 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2787 contents are added to the log line that is written for every discarded
2788 recipient. (Previously a log_message setting was ignored.)
2790 36. The ${quote: operator now quotes the string if it is empty.
2792 37. The install script runs exim in order to find its version number. If for
2793 some reason other than non-existence or emptiness, which it checks, it
2794 could not run './exim', it was installing it with an empty version number,
2795 i.e. as "exim-". This error state is now caught, and the installation is
2798 38. An argument was missing from the function that creates an error message
2799 when Exim fails to connect to the socket for saslauthd authentication.
2800 This could cause Exim to crash, or give a corrupted message.
2802 39. Added isip, isip4, and isip6 to ${if conditions.
2804 40. The ACL variables $acl_xx are now saved with the message, and can be
2805 accessed later in routers, transports, and filters.
2807 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2808 strings in the file are not string-expanded.
2810 42. If a MAIL command specified a SIZE value that was too large to fit into an
2811 int variable, the check against message_size_limit failed. Such values are
2812 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2813 day this will have to be increased, but I don't think I want to be around
2814 when emails are that large.
2821 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2822 iconv() is not standard in FreeBSD.
2824 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2825 IPv6 enabled. The observed symptom was a segmentation fault on return from
2826 the function os_common_find_running_interfaces() in src/os.c.
2828 3. In the check_special_case() function in daemon.c I had used "errno" as an
2829 argument name, which causes warnings on some systems. This was basically a
2830 typo, since it was named "eno" in the comments!
2832 4. The code that waits for the clock to tick (at a resolution of some fraction
2833 of a second) so as to ensure message-id uniqueness was always waiting for
2834 at least one whole tick, when it could have waited for less. [This is
2835 almost certainly not relevant at current processor speeds, where it is
2836 unlikely to ever wait at all. But we try to future-proof.]
2838 5. The function that sleeps for a time interval that includes fractions of a
2839 second contained a race. It did not block SIGALRM between setting the
2840 timer, and suspending (a couple of lines later). If the interval was short
2841 and the sigsuspend() was delayed until after it had expired, the suspension
2842 never ended. On busy systems this could lead to processes getting stuck for
2845 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2846 process, before it forks any delivery processes. The open lookup caching
2847 mechanism meant that the open file or database connection was passed into
2848 the delivery process. The problem was that delivery processes always tidy
2849 up cached lookup data. This could cause a problem for the next delivery
2850 process started by the queue runner, because the external queue runner
2851 process does not know about the closure. So the next delivery process
2852 still has data in the lookup cache. In the case of a file lookup, there was
2853 no problem because closing a file descriptor in a subprocess doesn't affect
2854 the parent. However, if the lookup was caching a connection to a database,
2855 the connection was closed, and the second delivery process was likely to
2856 see errors such as "PGSQL: query failed: server closed the connection
2857 unexpectedly". The problem has been fixed by closing all cached lookups
2858 in a queue runner before running a delivery process.
2860 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2861 seem to have the "const" qualifier which it has on other OS. I've
2864 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2865 *of the same type* (client or server) with the same public name that an
2866 error should be diagnosed.
2868 9. When Exim looked up a host name for an IP address, but failed to find the
2869 original IP address when looking up the host name (a safety check), it
2870 output the message "<ip address> does not match any IP for NULL", which was
2871 confusing, to say the least. The bug was that the host name should have
2872 appeared instead of "NULL".
2874 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2875 user that is built into the binary, and the -C or -D options is used, root
2876 privilege is dropped before the configuration file is read. In addition,
2877 logging is switched to stderr instead of the normal log files. If the
2878 configuration then re-defines the Exim user, the unprivileged environment
2879 is probably not what is expected, so Exim logs a panic warning message (but
2882 However, if deliver_drop_privilege is set, the unprivileged state may well
2883 be exactly what is intended, so the warning has been cut out in that case,
2884 and Exim is allowed to try to write to its normal log files.
2890 1. smtp_return_error_details was not giving details for temporary sender
2891 or receiver verification errors.
2893 2. Diagnose a configuration error if two authenticators have the same public
2896 3. Exim used not to create the message log file for a message until the first
2897 delivery attempt. This could be confusing when incoming messages were held
2898 for policy or load reasons. The message log file is now created at the time
2899 the message is received, and an initial "Received" line is written to it.
2901 4. The automatically generated man page for command line options had a minor
2902 bug that caused no ill effects; however, a more serious problem was that
2903 the procedure for building the man page automatically didn't always
2904 operate. Consequently, release 4.20 contains an out-of-date version. This
2905 shouldn't happen again.
2907 5. When building Exim with embedded Perl support, the script that builds the
2908 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2909 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2911 6. The freeze_tell option was not being used for messages that were frozen on
2912 arrival, either by an ACL or by local_scan().
2914 7. Added the smtp_incomplete_transaction log selector.
2916 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2917 was accepting AUTH without a new EHLO.
2919 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2920 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2923 10. Logging of TCP/IP connections (when configured) now happens in the main
2924 daemon process instead of the child process, so that the TCP/IP connection
2925 count is more accurate (but it can never be perfect).
2927 11. The use of "drop" in a nested ACL was not being handled correctly in the
2928 outer ACL. Now, if condition failure induced by the nested "drop" causes
2929 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2930 or "require"), the connection is dropped.
2932 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2933 that yield "discard" can now be used with an "accept" or a "discard" verb,
2934 but an error is generated for any others (because I can't see a useful way
2935 to define what should happen).
2937 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2938 are now processed in the same way as lines in the Exim configuration file.
2939 In particular, continuation lines are supported.
2941 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2943 15. Added -ti meaning -t -i.
2945 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2946 domains, and warn by logging if others are found.
2948 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2949 returned from GIFCONF when Exim is trying to find the list of interfaces on
2950 a host. The code in os.c has been modified to copy each ifreq to an aligned
2951 structure in all cases.
2953 Also, in some cases, the returned ifreq's were being copied to a 'struct
2954 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2955 means the last couple of bytes of an IPv6 address could be chopped if the
2956 ifreq contained only a normal sockaddr (14 bytes storage).
2958 18. Named domain lists were not supported in the hosts_treat_as_local option.
2959 An entry such as +xxxx was not recognized, and was treated as a literal
2962 19. Ensure that header lines added by a DATA ACL are included in the reject log
2963 if the ACL subsequently rejects the message.
2965 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2966 MD5 (which is deprecated).
2968 21. When testing a filter file using -bf, Exim was writing a message when it
2969 took the sender from a "From " line in the message, but it was not doing so
2970 when it took $return_path from a Return-Path: header line. It now does.
2972 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2973 with a valid header line field name (a series of printing characters
2974 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2976 23. Changed "disc" in the source to "disk" to conform to the documentation and
2977 the book and for uniformity.
2979 24. Ignore Sendmail's -Ooption=value command line item.
2981 25. When execve() failed while trying to run a command in a pipe transport,
2982 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2983 could be confused with a return value of 69 from the command itself. This
2984 has been changed to 127, the value the shell returns if it is asked to run
2985 a non-existent command. The wording for the related log line suggests a
2986 non-existent command as the problem.
2988 26. If received_header_text expands to an empty string, do not add a Received:
2989 header line to the message. (Well, it adds a token one on the spool, but
2990 marks it "old" so that it doesn't get used or transmitted.)
2992 27. Installed eximstats 1.28 (addition of -nt option).
2994 28. There was no check for failure on the call to getsockname() in the daemon
2995 code. This can fail if there is a shortage of resources on the system, with
2996 ENOMEM, for example. A temporary error is now given on failure.
2998 29. Contrary to the C standard, it seems that in some environments, the
2999 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3000 program. Exim now does this explicitly; it affects the formatting of
3001 timestamps using strftime().
3003 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3004 complaints. I've now initialized all the variables, to avoid this.
3006 32. Header lines added by a system filter were not being "seen" during
3007 transport-time rewrites.
3009 33. The info_callback() function passed to OpenSSL is set up with type void
3010 (*)(SSL *, int, int), as described somewhere. However, when calling the
3011 function (actually a macro) that sets it up, the type void(*)() is
3012 expected. I've put in a cast to prevent warnings from picky compilers.
3014 34. If a DNS black list lookup found a CNAME record, but there were no A
3015 records associated with the domain it pointed at, Exim crashed.
3017 35. If a DNS black list lookup returned more than one A record, Exim ignored
3018 all but the first. It now scans all returned addresses if a particular IP
3019 value is being sought. In this situation, the contents of the
3020 $dnslist_value variable are a list of all the addresses, separated by a
3023 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3024 to accept a host name and all its aliases if the forward lookup for any of
3025 them yielded the IP address of the incoming connection. Now it accepts only
3026 those names whose forward lookup yields the correct IP address. Any other
3027 names are discarded. This closes a loophole whereby a rogue DNS
3028 administrator could create reverse DNS records to break through a
3029 wildcarded host restriction in an ACL.
3031 37. If a user filter or a system filter that ran in a subprocess used any of
3032 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3033 the wrong values were passed to the pipe command ($thisaddress had the
3034 value of $0, $0 had the value of $1, etc). This bug was introduced by
3035 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3037 38. Improved the line breaking for long SMTP error messages from ACLs.
3038 Previously, if there was no break point between 40 and 75 characters, Exim
3039 left the rest of the message alone. Two changes have been made: (a) I've
3040 reduced the minimum length to 35 characters; (b) if it can't find a break
3041 point between 35 and 75 characters, it looks ahead and uses the first one
3042 that it finds. This may give the occasional overlong line, but at least the
3043 remaining text gets split now.
3045 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3046 file descriptors might be low, and that setting 1000 would always raise it.
3047 It turns out that in some environments, the limit is already over 1000 and
3048 that lowering it causes trouble. So now Exim takes care not to decrease it.
3050 40. When delivering a message, the value of $return_path is set to $sender_
3051 address at the start of routing (routers may change the value). By an
3052 oversight, this default was not being set up when an address was tested by
3053 -bt or -bv, which affected the outcome if any router or filter referred to
3056 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3057 log only when "message" or "log_message" are set. However, if one of the
3058 conditions was an address verification, or a call to a nested ACL, the
3059 messages generated by the underlying test were being passed through. This
3060 no longer happens. The underlying message is available in $acl_verify_
3061 message for both "message" and "log_message" expansions, so it can be
3062 passed through if needed.
3064 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3065 new expansion $bh_ to give the encoded byte string without charset
3066 translation. Translation happens only if iconv() is available; HAVE_ICONV
3067 indicates this at build time. HEADERS_CHARSET gives the charset to
3068 translate to; headers_charset can change it in the configuration, and
3069 "headers charset" can change it in an individual filter file.
3071 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3072 that creates RFC 2047 encoded "words" labels them as that charset instead
3073 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3074 expansion operator; (ii) when Exim creates a From: line for a local
3075 message; (iii) when a header line is rewritten to include a "phrase" part.
3077 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3078 buggy, causing it to skip the first lines of messages whose message ID
3079 ended in 'D'. This would not have bitten before Exim release 4.14, because
3080 message IDs were unlikely to end in 'D' before then. The effect was to have
3081 incorrect size information for certain domains.
3083 45. #include "config.h" was missing at the start of the crypt16.c module. This
3084 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3087 46. If there was a timeout during a "random" callout check, Exim treated it as
3088 a failure of the random address, and carried on sending RSET and the real
3089 address. If the delay was just some slowness somewhere, the response to the
3090 original RCPT would be taken as a response to RSET and so on, causing
3091 mayhem of various kinds.
3093 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3094 when I implemented it. It didn't allow for the fact that some option values
3095 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3096 the right test for positive values.
3098 48. Domain names in DNS records are case-independent. Exim always looks them up
3099 in lower case. Some resolvers return domain names in exactly the case they
3100 appear in the zone file, that is, they may contain uppercase letters. Not
3101 all resolvers do this - some return always lower case. Exim was treating a
3102 change of case by a resolver as a change of domain, similar to a widening
3103 of a domain abbreviation. This triggered its re-routing code and so it was
3104 trying to route what was effectively the same domain again. This normally
3105 caused routing to fail (because the router wouldn't handle the domain
3106 twice). Now Exim checks for this case specially, and just changes the
3107 casing of the domain that it ultimately uses when it transmits the message
3110 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3113 50. If a filter generated a file delivery with a non-absolute name (possible if
3114 no home directory exists for the router), the forbid_file option was not
3117 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3118 the existing equality matching.
3120 52. Exim was using ints instead of ino_t variables in some places where it was
3121 dealing with inode numbers.
3123 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3124 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3125 called TMPDIR, and if it finds it is different, it changes its value.
3127 54. The smtp_printf() function is now made available to local_scan() so
3128 additional output lines can be written before returning. There is also an
3129 smtp_fflush() function to enable the detection of a dropped connection.
3130 The variables smtp_input and smtp_batched_input are exported to
3133 55. Changed the default runtime configuration: the message "Unknown user"
3134 has been removed from the ACL, and instead placed on the localuser router,
3135 using the cannot_route_message feature. This means that any verification
3136 failures that generate their own messages won't get overridden. Similarly,
3137 the "Unrouteable address" message that was in the ACL for unverifiable
3138 relay addresses has also been removed.
3140 56. Added hosts_avoid_esmtp to the smtp transport.
3142 57. The exicyclog script was not checking for the esoteric option
3143 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3144 will work only if exicyclog is run under the appropriate euid.
3146 58. Following a discussion on the list, the rules by which Exim recognises line
3147 endings on incoming messages have been changed. The -dropcr and drop_cr
3148 options are now no-ops, retained only for backwards compatibility. The
3149 following line terminators are recognized: LF CRLF CR. However, special
3150 processing applies to CR:
3152 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3153 nor a local message in the state where . is a terminator.
3155 (ii) If a bare CR is encountered in a header line, an extra space is added
3156 after the line terminator so as not to end the header. The reasoning
3157 behind this is that bare CRs in header lines are most likely either
3158 to be mistakes, or people trying to play silly games.
3160 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3161 was being incorrectly given as 18 bytes larger than it should have been.
3162 This is a VOB (very old bug).
3164 60. This may never have affected anything current, but just in case it has:
3165 When the local host is found other than at the start of a list of hosts,
3166 the local host, those with the same MX, and any that follow, are discarded.
3167 When the list in question was part of a longer list of hosts, the following
3168 hosts (not currently being processed) were also being discarded. This no
3169 longer happens. I'm not sure if this situation could ever has previously
3172 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3175 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3176 header. This is something that is recommended in a new Internet Draft, and
3177 is something that is documented as being done by Sendmail. There are two
3178 possible values. For messages generated by the autoreply transport, Exim
3181 Auto-Submitted: auto-replied
3183 whereas for all other generated messages (e.g. bounces) it adds
3185 Auto-Submitted: auto-generated
3187 63. The "personal" condition in filters now includes a test for the
3188 Auto-Submitted: header. If it contains the string "auto-" the message it
3189 not considered personal.
3191 64. Added rcpt_include_affixes as a generic transport option.
3193 65. Added queue_only_override (default true).
3195 66. Added the syslog_duplication option.
3197 67. If what should have been the first header line of a message consisted of
3198 a space followed by a colon, Exim was mis-interpreting it as a header line.
3199 It isn't of course - it is syntactically invalid and should therefore be
3200 treated as the start of the message body. The misbehaviour could have
3201 caused a number of strange effects, including loss of data in subsequent
3202 header lines, and spool format errors.
3204 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3205 client host had authenticated. This control can now be exercised by an ACL
3206 for more flexibility.
3208 69. By default, callouts do not happen when testing with -bh. There is now a
3209 variant, -bhc, which does actually run the callout code, including
3210 consulting and updating the callout cache.
3212 70. Added support for saslauthd authentication, courtesy of Alexander
3215 71. If statvfs() failed on the spool or log directories while checking their
3216 size for availability, Exim confusingly gave the error "space shortage".
3217 Furthermore, in debugging mode it crashed with a floating point exception.
3218 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3219 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3220 serious problem, Exim now writes to the main and panic logs when this
3221 happens, with details of the failure. It then refuses to accept the
3222 incoming message, giving the message "spool directory problem" or "log
3223 directory problem" with a 421 code for SMTP messages.
3225 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3226 0, 1, and 2 exist, because some OS complain for execs without them (see
3227 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3228 descriptors. However, the code omitted to check that the open succeeded,
3229 causing mysterious errors if for some reason the permissions on /dev/null
3230 got screwed. Now Exim writes a message to the main and panic logs, and
3231 bombs out if it can't open /dev/null.
3233 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3234 interact so that it is all more flexible. It is supposed to remain
3235 backwards compatible. Also added extra_local_interfaces.
3237 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3238 to bomb out with an assertion failure - to the client this appears as a
3239 connection drop. This problem occurs in the part of the code that was taken
3240 from the Samba project. Fortunately, the assertion is in a very simple
3241 function, so I have fixed this by reproducing the function inline in the
3242 one place where it is called, and arranging for authentication to fail
3243 instead of killing the process with assert().
3245 75. The SPA client code was not working when the server requested OEM rather
3246 than Unicode encoding.
3248 76. Added code to make require_files with a specific uid setting more usable in
3249 the case where statting the file as root fails - usually a non-root-mounted
3250 NFS file system. When this happens and the failure is EACCES, Exim now
3251 forks a subprocess and does the per-uid checking as the relevant uid.
3253 77. Added process_log_path.
3255 78. If log_file_path was not explicitly set, a setting of check_log_space or
3256 check_log_inodes was ignored.
3258 79. If a space check for the spool or log partitions fails, the incident is now
3259 logged. Of course, in the latter case the data may get lost...
3261 80. Added the %p formatting code to string_format() so that it can be used to
3262 print addresses in debug_print(). Adjusted all the address printing in the
3263 debugging in store.c to use %p rather than %d.
3265 81. There was a concern that a line of code in smtp_in.c could overflow a
3266 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3267 initially expressed, the concern was not well-founded, because trailing
3268 spaces are removed early. However, if the trailing spaces were followed by
3269 a NULL, they did not get removed, so the overflow was possible. Two fixes
3272 (a) I re-wrote the offending code in a cleaner fashion.
3273 (b) If an incoming SMTP command contains a NULL character, it is rejected
3276 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3277 runs initgroups(), so that if the Exim user is in any additional groups,
3278 they will be used during message reception.
3284 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob