1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.197 2005/08/02 16:12:14 jetmore 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)
65 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
67 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
68 command itself ended in error, the underlying message about the transport
69 filter timeout was being overwritten with the pipe command error. Now the
70 underlying error message should be appended to the second error message.
72 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
74 PH/02 Reduce the amount of output that "make" produces by default. Full output
75 can still be requested.
77 PH/03 The warning log line about a condition test deferring for a "warn" verb
78 was being output only once per connection, rather than after each
79 occurrence (because it was using the same function as for successful
80 "warn" verbs). This seems wrong, so I have changed it.
82 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
83 it should not have, which might have caused a crash in the right
84 circumstances, but probably never did.
86 PH/04 Installed a modified version of Tony Finch's patch to make submission
87 mode fix the return path as well as the Sender: header line, and to
88 add a /name= option so that you can make the user's friendly name appear
91 TF/03 Added the control = fakedefer ACL modifier.
93 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
94 Mark Lowes for thorough testing.
96 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
98 TK/03 Merged latest SRS patch from Miles Wilton.
100 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
101 with the definition in sysexits.h (which is #included earlier).
102 Fortunately, Exim does not actually use EX_OK. The code used to try to
103 preserve the sysexits.h value, by assumimg that macro definitions were
104 scanned for macro replacements. I have been disabused of this notion,
105 so now the code just undefines EX_OK before #including unistd.h.
107 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
108 in the smtp transport. When a block could not be written in a single
109 write() function, the timeout was being re-applied to each part-write.
110 This seems wrong - if the receiver was accepting one byte at a time it
111 would take for ever. The timeout is now adjusted when this happens. It
112 doesn't have to be particularly precise.
114 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
115 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
117 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
118 without a domain if the check to <postmaster@domain> fails.
120 SC/01 Eximstats: added -xls and the ability to specify output files
121 (patch written by Frank Heydlauf).
123 SC/02 Eximstats: use FileHandles for outputing results.
125 SC/03 Eximstats: allow any combination of xls, txt, and html output.
127 SC/04 Eximstats: fixed display of large numbers with -nvr option
129 SC/05 Eximstats: fixed merging of reports with empty tables.
131 SC/06 Eximstats: added the -include_original_destination flag
133 SC/07 Eximstats: removed tabs and trailing whitespace.
135 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
137 TK/06 MBOX spool code: Add real "From " MBOX separator line
138 so the .eml file is really in mbox format (even though
139 most programs do not really care). Patch from Alex Miller.
141 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
142 The latter is generated from $received_to and is only set if the
143 message has one envelope recipient. SA can use these headers,
144 obviously out-of-the-box. Patch from Alex Miller.
146 PH/08 The ${def test on a variable was returning false if the variable's
147 value was "0", contrary to what the specification has always said!
148 The result should be true unless the variable is empty.
150 PH/09 The syntax error of a character other than { following "${if
151 def:variable_name" (after optional whitespace) was not being diagnosed.
152 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
153 accidental colon was present, for example, could give incorrect results.
155 PH/10 Tidied the code in a number of places where the st_size field of a stat()
156 result is used (not including appendfile, where other changes are about
159 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
160 This involved changing a lot of size variables from int to off_t. It
161 should work with maildirs and everything.
163 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
164 spamd dying while we are connected to it.
166 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
169 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
170 function for BDB 4.3.
172 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
174 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
175 the sources for PCRE have been re-arranged and more files are now
178 PH/15 The code I had for printing potentially long long variables in PH/11
179 above was not the best (it lost precision). The length of off_t variables
180 is now inspected at build time, and an appropriate printing format (%ld
181 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
182 to be "long long int" or "long int". This is needed for the internal
183 formatting function string_vformat().
185 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
186 the configuration file to be ":syslog", then the script "guesses" where
187 the logs files are, rather than using the compiled in default. In our
188 case the guess is not the same as the compiled default, so the script
189 suddenly stopped working when I started to use syslog. The patch checks
190 to see if log_file_path is "". If so, it attempts to read it from exim
191 with no configuration file to get the compiled in version, before it
192 falls back to the previous guessing code."
194 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
195 implementing BATV in an Exim configuration. See NewStuff for the gory
198 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
199 Makefile that are specific to HP-UX.
201 PH/18 If the "use_postmaster" option was set for a recipient callout together
202 with the "random" option, the postmaster address was used as the MAIL
203 FROM address for the random test, but not for the subsequent recipient
204 test. It is now used for both.
206 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
207 patch removes a few documentation additions to RFC 3028, because the
208 latest draft now contains them. It adds the new en;ascii-case comparator
209 and a new error check for 8bit text in MIME parts. Comparator and
210 require names are now matched exactly. I enabled the subaddress
211 extension, but it is not well tested yet (read: it works for me)."
213 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
214 rework some of the code of TK/09 above to avoid the hardwired use of
215 "%lld" and "long long". Replaced the call to snprintf() with a call to
218 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
219 records point to non-existent hosts", "retry timeout exceeded", and
220 "retry time not reached for any host after a long failure period".
222 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
223 experimental DomainKeys support:
225 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
226 (2) On an error such as an illegally used "control", the wrong name for
227 the control was given.
229 These problems did NOT occur unless DomainKeys support was compiled.
231 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
233 PH/24 Added ${if match_ip condition.
235 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
236 databases so that it will be absolutely obvious if a crash occurs in the
237 DB library. This is a regular occurrence (often caused by mis-matched
240 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
241 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
242 picked up on a user's system that detects such things. There doesn't seem
243 to be a gcc warning option for this - only an attribute that has to be
244 put on the function's prototype. It seems that in Fedora Core 4 they have
245 set this on a number of new functions. No doubt there will be more in due
248 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
249 specify a transport. However, if an address that was verified by such a
250 router was the subject of a callout, Exim crashed because it tried to
251 read the rcpt_include_affixes from the non-existent transport. Now it
252 just assumes that the setting of that option is false. This bug was
253 introduced by 4.51/PH/31.
255 PH/28 Changed -d+all to exclude +memory, because that information is very
256 rarely of interest, but it makes the output a lot bigger. People tend to
257 do -d+all out of habit.
259 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
260 code in os-type was giving problems when libc.so lives in lib64, like on
263 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
264 aren't the modern standard, and it seems that some systems' include files
265 don't always have them. Exim was already checking for some of the newer
266 ones like T_AAAA, and defining it itself. I've added checks for all the
267 record types that Exim uses.
269 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
270 not automatically generating a new one, as it is supposed to. This
271 prevented TLS from working. If the file did exist, but contained invalid
272 data, a new version was generated, as expected. It was only the case of a
273 non-existent file that was broken.
275 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
276 with a change in libdomainkeys > 0.64.
278 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
279 from DNS. If the selector record carries the flag, it now has
280 precedence over the domain-wide flag.
282 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
284 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
285 the use of a port name that isn't defined in /etc/services) occurred, the
286 message was deferred as in a normal delivery, and thus remained on the
287 spool, instead of being failed because of the mua_wrapper setting. This
288 is now fixed, and I tidied up some of the mua_wrapper messages at the
291 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
292 the messages in a hash of arrays rather than using individual hashes.
293 This is a bit cleaner and results in dramatic memory savings, albeit
294 at a slight CPU cost.
296 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
297 as requested by Marc Sherman.
299 SC/10 Eximstats: added histograms for user specified patterns as requested
302 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
304 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
305 fopen() in the content-scanning modules that did not already have it.
311 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
312 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
314 TK/02 Fix ACL "control" statement not being available in MIME ACL.
316 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
318 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
319 to test Sieve filters that use "vacation".
321 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
322 that changes the way the GnuTLS parameters are stored in the cache file.
323 The new format can be generated externally. For backward compatibility,
324 if the data in the cache doesn't make sense, Exim assumes it has read an
325 old-format file, and it generates new data and writes a new file. This
326 means that you can't go back to an older release without removing the
329 PH/03 A redirect router that has both "unseen" and "one_time" set does not
330 work if there are any delivery delays because "one_time" forces the
331 parent to be marked "delivered", so its unseen clone is never tried
332 again. For this reason, Exim now forbids the simultaneous setting of
335 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
336 redirected to themselves ("homonym" addresses). Read the long ChangeLog
337 entry if you want to know the details. The fix, however, neglected to
338 consider the case when local delivery batching is involved. The test for
339 "previously delivered" was not happening when checking to see if an
340 address could be batched with a previous (undelivered) one; under
341 certain circumstances this could lead to multiple deliveries to the same
344 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
345 in its include files, and this causes problems building Exim.
347 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
348 header_syntax) but Exim was just ignoring anything given after a slash.
349 In particular, this caused confusion with an attempt to use "verify =
350 reverse_host_lookup/defer_ok". An error is now given when options are
351 supplied for verify items that do not have them. (Maybe reverse_host_
352 lookup should have a defer_ok option, but that's a different point.)
354 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
355 defined by RFC 821) to 2048, because there were problems with some AUTH
356 commands, and RFC 1869 says the size should be increased for extended
357 SMTP commands that take arguments.
359 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
362 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
363 "unknown" error; now it says that the functionality isn't in the binary.
365 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
366 an address' error message when a string expansion fails (syntax or
367 whatever). Otherwise the password may appear in the log. Following change
368 PH/42 below, there is no longer a chance of it appearing in a bounce
371 PH/11 Installed exipick version 20050225.0 from John Jetmore.
373 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
374 of its addresses was ever tried. (Bugzilla bug #2.)
376 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
377 the result incorrectly in the debug output. (It correctly added a newline
378 to what was transported.)
380 TF/01 Added $received_time.
382 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
383 commented out examples of how to interface to a virus scanner and to
384 SpamAssassin. Also added commented examples of av_scanner and
385 spamd_address settings.
387 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
388 and controls are allowed in which ACLs. There were a couple of minor
389 errors. Some of the entries in the conditions table (which is a table of
390 where they are NOT allowed) were getting very unwieldy; rewrote them as a
391 negation of where the condition IS allowed.
393 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
395 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
396 header file does not have a version number, so I've had to invent a new
397 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
398 API. The code is untested by me (my Linux distribution still has 0.3.2 of
399 radiusclient), but it was contributed by a Radius user.
401 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
402 files or directories, for OpenSSL.
404 PH/19 When an Exim process that is running as root has to create an Exim log
405 file, it does so in a subprocess that runs as exim:exim so as to get the
406 ownership right at creation (otherwise, other Exim processes might see
407 the file with the wrong ownership). There was no test for failure of this
408 fork() call, which would lead to the process getting stuck as it waited
409 for a non-existent subprocess. Forks do occasionally fail when resources
410 run out. I reviewed all the other calls to fork(); they all seem to check
413 PH/20 When checking for unexpected SMTP input at connect time (before writing
414 the banner), Exim was not dealing correctly with a non-positive return
415 from the read() function. If the client had disconnected by this time,
416 the result was a log entry for a synchronization error with an empty
417 string after "input=" when read() returned zero. If read() returned -1
418 (an event I could not check), uninitialized data bytes were printed.
419 There were reports of junk text (parts of files, etc) appearing after
422 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
424 PH/22 Added support for macro redefinition, and (re)definition in between
425 driver and ACL definitions.
427 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
428 forgetting to use the resulting value; it was using the unexpanded value.
430 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
431 hadn't been configured. The fix is from Juergen Kreileder, who
432 understands it better than I do:
434 "Here's what I see happening with three configured cyrus_sasl
435 authenticators configured (plain, login, cram-md5):
437 On startup auth_cyrus_sasl_init() gets called for each of these.
438 This means three calls to sasl_listmech() without a specified mech_list.
439 => SASL tests which mechs of all available mechs actually work
440 => three warnings about OTP not working
441 => the returned list contains: plain, login, cram-md5, digest-md5, ...
443 With the patch, sasl_listmech() also gets called three times. But now
444 SASL's mech_list option is set to the server_mech specified in the the
445 authenticator. Or in other words, the answer from sasl_listmech()
446 gets limited to just the mech you're testing for (which is different
448 => the return list contains just 'plain' or 'login', 'cram-md5' or
449 nothing depending on the value of ob->server_mech.
451 I've just tested the patch: Authentication still works fine,
452 unavailable mechs specified in the exim configuration are still
453 caught, and the auth.log warnings about OTP are gone."
455 PH/25 When debugging is enabled, the contents of the command line are added
456 to the debugging output, even when log_selector=+arguments is not
459 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
460 answer is "GNU", and only if the return is "GNU/something" is the answer
463 PH/27 $acl_verify_message is now set immediately after the failure of a
464 verification in an ACL, and so is available in subsequent modifiers. In
465 particular, the message can be preserved by coding like this:
467 warn !verify = sender
468 set acl_m0 = $acl_verify_message
470 Previously, $acl_verify_message was set only while expanding "message"
471 and "log_message" when a very denied access.
473 PH/28 Modified OS/os.c-Linux with
475 -#ifndef OS_LOAD_AVERAGE
476 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
478 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
479 nomenclature these days.)
481 PH/29 Installed patch from the Sieve maintainer that adds the options
482 sieve_useraddress and sieve_subaddress to the redirect router.
484 PH/30 In these circumstances:
485 . Two addresses routed to the same list of hosts;
486 . First host does not offer TLS;
487 . First host accepts first address;
488 . First host gives temporary error to second address;
489 . Second host offers TLS and a TLS session is established;
490 . Second host accepts second address.
491 Exim incorrectly logged both deliveries with the TLS parameters (cipher
492 and peerdn, if requested) that were in fact used only for the second
495 PH/31 When doing a callout as part of verifying an address, Exim was not paying
496 attention to any local part prefix or suffix that was matched by the
497 router that accepted the address. It now behaves in the same way as it
498 does for delivery: the affixes are removed from the local part unless
499 rcpt_include_affixes is set on the transport.
501 PH/32 Add the sender address, as F=<...>, to the log line when logging a
502 timeout during the DATA phase of an incoming message.
504 PH/33 Sieve envelope tests were broken for match types other than :is. I have
505 applied a patch sanctioned by the Sieve maintainer.
507 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
508 the uid or gid is negative. A case of a negative gid caused this to be
509 noticed. The fix allows for either to be negative.
511 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
512 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
515 PH/36 The change PH/12 above was broken. Fixed it.
517 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
518 the grounds that routing the same address twice would always produce the
519 same answer. This might have been true once, but it is certainly no
520 longer true now. Routing a child address may depend on the previous
521 routing that produced that child. Some complicated redirection strategies
522 went wrong when messages had multiple recipients, and made Exim's
523 behaviour dependent on the order in which the addresses were given.
525 I have moved the duplicate checking until after the routing is complete.
526 Exim scans the addresses that are assigned to local and remote
527 transports, and removes any duplicates. This means that more work will be
528 done, as duplicates will always all be routed, but duplicates are
529 presumably rare, so I don't expect this is of any significance.
531 For deliveries to pipes, files, and autoreplies, the duplicate checking
532 still happens during the routing process, since they are not going to be
535 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
536 It corrects a timeout issue with spamd. This is Ian's comment: "The
537 background is that sometimes spamd either never reads data from a
538 connection it has accepted, or it never writes response data. The exiscan
539 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
540 blindly assumes that writes won't block so it may never time out."
542 PH/39 Allow G after quota size as well as K and M.
544 PH/40 The value set for $authenticated_id in an authenticator may not contain
545 binary zeroes or newlines because the value is written to log lines and
546 to spool files. There was no check on this. Now the value is run through
547 the string_printing() function so that such characters are converted to
548 printable escape sequences.
550 PH/41 $message_linecount is a new variable that contains the total number of
551 lines in the message. Compare $body_linecount, which is the count for the
554 PH/42 Exim no longer gives details of delivery errors for specific addresses in
555 bounce and delay warning messages, except in certain special cases, which
558 (a) An SMTP error message from a remote host;
559 (b) A message specified in a :fail: redirection;
560 (c) A message specified in a "fail" command in a system filter;
561 (d) A message specified in a FAIL return from the queryprogram router;
562 (e) A message specified by the cannot_route_message router option.
564 In these cases only, Exim does include the error details in bounce and
565 warning messages. There are also a few cases where bland messages such
566 as "unrouteable address" or "local delivery error" are given.
568 PH/43 $value is now also set for the "else" part of a ${run expansion.
570 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
571 being worked on, but at least Exim now implements the latest version to
574 PH/45 In a pipe transport, although a timeout while waiting for the pipe
575 process to complete was treated as a delivery failure, a timeout while
576 writing the message to the pipe was logged, but erroneously treated as a
577 successful delivery. Such timeouts include transport filter timeouts. For
578 consistency with the overall process timeout, these timeouts are now
579 treated as errors, giving rise to delivery failures by default. However,
580 there is now a new Boolean option for the pipe transport called
581 timeout_defer, which, if set TRUE, converts the failures into defers for
582 both kinds of timeout. A transport filter timeout is now identified in
585 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
586 systems where "make" and "gmake" are different, calling "gmake" at top
587 level broke things. I've arranged for the value of $(MAKE) to be passed
588 from the Makefile to this script so that it can call the same version of
592 A note about Exim versions 4.44 and 4.50
593 ----------------------------------------
595 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
596 changes of various kinds. As a consequence, a big documentation update was
597 needed. This delayed the release for rather longer than seemed good, especially
598 in the light of a couple of (minor) security issues. Therefore, the changes
599 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
600 release. So 4.44 and 4.50 are in effect two different branches that both start
603 I have left the 4.50 change log unchanged; it contains all the changes since
604 4.43. The change log for 4.44 is below; many of its items are identical to
605 those for 4.50. This seems to be the most sensible way to preserve the
606 historical information.
612 1. Minor wording change to the doc/README.SIEVE file.
614 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
615 computation of the current number of files was incorrect.
617 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
618 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
621 4. Give more explanation in the error message when the command for a transport
622 filter fails to execute.
624 5. There are several places where Exim runs a non-Exim command in a
625 subprocess. The SIGUSR1 signal should be disabled for these processes. This
626 was being done only for the command run by the queryprogram router. It is
627 now done for all such subprocesses. The other cases are: ${run, transport
628 filters, and the commands run by the lmtp and pipe transports.
630 6. Added CONFIGURE_GROUP build-time option.
632 7. Some older OS have a limit of 256 on the maximum number of file
633 descriptors. Exim was using setrlimit() to set 1000 as a large value
634 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
635 systems. I've change it so that if it can't get 1000, it tries for 256.
637 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
638 was an oversight, and furthermore, ever since the addition of extra
639 controls (e.g. 4.43/32), the checks on when to allow different forms of
640 "control" were broken. There should now be diagnostics for all cases when a
641 control that does not make sense is encountered.
643 9. Added the /retain_sender option to "control=submission".
645 10. $recipients is now available in the predata ACL (oversight).
647 11. Tidy the search cache before the fork to do a delivery from a message
648 received from the command line. Otherwise the child will trigger a lookup
649 failure and thereby defer the delivery if it tries to use (for example) a
650 cached ldap connection that the parent has called unbind on.
652 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
653 of $address_data from the recipient verification was clobbered by the
656 13. The value of address_data from a sender verification is now available in
657 $sender_address_data in subsequent conditions in the ACL statement.
659 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
661 15. Added a new option "connect=<time>" to callout options, to set a different
664 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
665 was its contents. (It was OK if the option was not defined at all.)
667 17. A "Completed" log line is now written for messages that are removed from
668 the spool by the -Mrm option.
670 18. New variables $sender_verify_failure and $recipient_verify_failure contain
671 information about exactly what failed.
673 19. Added -dd to debug only the daemon process.
675 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
676 handles timeouts, both on the server side and network timeouts. Renamed the
677 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
679 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
680 It is now set to "smtps".
682 22. $host_address is now set to the target address during the checking of
685 23. When checking ignore_target_hosts for an ipliteral router, no host name was
686 being passed; this would have caused $sender_host_name to have been used if
687 matching the list had actually called for a host name (not very likely,
688 since this list is usually IP addresses). A host name is now passed as
691 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
692 code that specifies a non-restarting handler (typically sigaction() in
693 modern systems) in an attempt to fix a rare and obscure crash bug.
695 25. Narrowed the window for a race in the daemon that could cause it to ignore
696 SIGCHLD signals. This is not a major problem, because they are used only to
697 wake it up if nothing else does.
699 26. A malformed maildirsize file could cause Exim to calculate negative values
700 for the mailbox size or file count. Odd effects could occur as a result.
701 The maildirsize information is now recalculated if the size or filecount
704 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
705 support for a long time. Removed HAVE_SYS_VFS_H.
707 28. Installed the latest version of exipick from John Jetmore.
709 29. In an address list, if the pattern was not a regular expression, an empty
710 subject address (from a bounce message) matched only if the pattern was an
711 empty string. Non-empty patterns were not even tested. This was the wrong
712 because it is perfectly reasonable to use an empty address as part of a
713 database query. An empty address is now tested by patterns that are
714 lookups. However, all the other forms of pattern expect the subject to
715 contain a local part and a domain, and therefore, for them, an empty
716 address still always fails if the pattern is not itself empty.
718 30. Exim went into a mad DNS loop when attempting to do a callout where the
719 host was specified on an smtp transport, and looking it up yielded more
722 31. Re-factored the code for checking spool and log partition space into a
723 function that finds that data and another that does the check. The former
724 is then used to implement four new variables: $spool_space, $log_space,
725 $spool_inodes, and $log_inodes.
727 32. The RFC2047 encoding function was originally intended for short strings
728 such as real names; it was not keeping to the 75-character limit for
729 encoded words that the RFC imposes. It now respects the limit, and
730 generates multiple encoded words if necessary. To be on the safe side, I
731 have increased the buffer size for the ${rfc2047: expansion operator from
734 33. It is now permitted to omit both strings after an "if" condition; if the
735 condition is true, the result is "true". As before, when the second string
736 is omitted, a false condition yields an empty string. This makes it less
737 cumbersome to write custom ACL and router conditions.
739 34. Failure to deliver a bounce message always caused it to be frozen, even if
740 there was an errors_to setting on the router. The errors_to setting is now
743 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
744 canonical form (fully expanded) before being placed in
745 $sender_host_address.
747 36. The table in the code that translates DNS record types into text (T_A to
748 "A" for instance) was missing entries for NS and CNAME. It is just possible
749 that this could have caused confusion if both these types were looked up
750 for the same domain, because the text type is used as part of Exim's
751 per-process caching. But the chance of anyone hitting this buglet seems
754 37. The dnsdb lookup has been extended in a number of ways.
756 (1) There is a new type, "zns", which walks up the domain tree until it
757 finds some nameserver records. It should be used with care.
759 (2) There is a new type, "mxh", which is like "mx" except that it returns
760 just the host names, not the priorities.
762 (3) It is now possible to give a list of domains (or IP addresses) to be
763 looked up. The behaviour when one of the lookups defers can be
764 controlled by a keyword.
766 (4) It is now possible to specify the separator character for use when
767 multiple records are returned.
769 38. The dnslists ACL condition has been extended: it is now possible to supply
770 a list of IP addresses and/or domains to be looked up in a particular DNS
773 39. Added log_selector=+queue_time_overall.
775 40. When running the queue in the test harness, wait just a tad after forking a
776 delivery process, to get repeatability of debugging output.
778 41. Include certificate and key file names in error message when GnuTLS fails
779 to set them up, because the GnuTLS error message doesn't include the name
780 of the failing file when there is a problem reading it.
782 42. Allow both -bf and -bF in the same test run.
784 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
786 44. The "Exiscan patch" is now merged into the mainline Exim source.
788 45. Sometimes the final signoff response after QUIT could fail to get
789 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
790 < 0 before doing a fflush(). This bug looks as though it goes back to the
791 introduction of TLS in release 3.20, but "sometimes" must have been rare
792 because the tests only now provoked it.
794 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
795 (this can affect the format of dates).
797 47. exim_tidydb, when checking for the continued existence of a message for
798 which it has found a message-specific retry record, was not finding
799 messages that were in split spool directories. Consequently, it was
800 deleting retry records that should have stayed in existence.
802 48. Steve fixed some bugs in eximstats.
804 49. The SPA authentication driver was not abandoning authentication and moving
805 on to the next authenticator when an expansion was forced to fail,
806 contradicting the general specification for all authenticators. Instead it
807 was generating a temporary error. It now behaves as specified.
809 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
810 (the order specifies the preference for clients). The order is now AES256,
811 AES128, 3DES, ARCFOUR128.
813 51. Small patch to Sieve code - explicitly set From: when generating an
816 52. Exim crashed if a remote delivery caused a very long error message to be
817 recorded - for instance if somebody sent an entire SpamAssassin report back
818 as a large number of 550 error lines. This bug was coincidentally fixed by
819 increasing the size of one of Exim's internal buffers (big_buffer) that
820 happened as part of the Exiscan merge. However, to be on the safe side, I
821 have made the code more robust (and fixed the comments that describe what
824 53. Now that there can be additional text after "Completed" in log lines (if
825 the queue_time_overall log selector is set), a one-byte patch to exigrep
826 was needed to allow it to recognize "Completed" as not the last thing in
829 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
830 patch that reportedly fixes this has been added. I am not expert enough to
831 create a test for it. This is what the patch creator wrote:
833 "I found a little strange behaviour of ldap code when working with
834 Windows 2003 AD Domain, where users was placed in more than one
835 Organization Units. When I tried to give exim partial DN, the exit code
836 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
837 But simultaneously result of request was absolutely normal ldap result,
838 so I produce this patch..."
840 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
841 REFERENCE, so I have modified the code to exclude the patch when that macro
844 55. Some experimental protocols are using DNS PTR records for new purposes. The
845 keys for these records are domain names, not reversed IP addresses. The
846 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
847 leaves it alone. Component reversal etc. now happens only for IP addresses.
850 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
852 57. Double the size of the debug message buffer (to 2048) so that more of very
853 long debug lines gets shown.
855 58. The exicyclog utility now does better if the number of log files to keep
856 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
858 59. Two changes related to the smtp_active_hostname option:
860 (1) $smtp_active_hostname is now available as a variable.
861 (2) The default for smtp_banner uses $smtp_active_hostname instead
862 of $primary_hostname.
864 60. The host_aton() function is supposed to be passed a string that is known
865 to be a valid IP address. However, in the case of IPv6 addresses, it was
866 not checking this. This is a hostage to fortune. Exim now panics and dies
867 if the condition is not met. A case was found where this could be provoked
868 from a dnsdb PTR lookup with an IPv6 address that had more than 8
869 components; fortuitously, this particular loophole had already been fixed
870 by change 4.50/55 above.
872 If there are any other similar loopholes, the new check in host_aton()
873 itself should stop them being exploited. The report I received stated that
874 data on the command line could provoke the exploit when Exim was running as
875 exim, but did not say which command line option was involved. All I could
876 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
880 61. There was a buffer overflow vulnerability in the SPA authentication code
881 (which came originally from the Samba project). I have added a test to the
882 spa_base64_to_bits() function which I hope fixes it.
885 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
886 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
888 63. The daemon start-up calls getloadavg() while still root for those OS that
889 need the first call to be done as root, but it missed one case: when
890 deliver_queue_load_max is set with deliver_drop_privilege. This is
891 necessary for the benefit of the queue runner, because there is no re-exec
892 when deliver_drop_privilege is set.
894 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
897 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
898 was in use, was not putting the data itself into the right store pool;
899 consequently, it could be overwritten for a subsequent message in the same
900 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
903 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
905 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
906 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
907 the function were treating the return as a boolean value, which happened to
908 work because 0=false and not-0=true, but is not correct code.
910 68. The host_aton() function was not handling scoped IPv6 addresses (those
911 with, for example, "%eth0" on the end) correctly.
913 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
914 negated items (that is, ~something) in unsigned ints. Some compilers
915 apparently mutter when there is no cast.
917 70. If an address verification called from an ACL failed, and did not produce a
918 user-specific message (i.e. there was only a "system" message), nothing was
919 put in $acl_verify_message. In this situation, it now puts the system
922 71. Change 4.23/11 added synchronization checking at the start of an SMTP
923 session; change 4.31/43 added the unwanted input to the log line - except
924 that it did not do this in the start of session case. It now does.
926 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
927 This is wrong and can cause the other end to generate a synchronization
928 error if it is another Exim or anything else that does the synchronization
929 check. A QUIT command is no longer sent after a timeout.
931 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
934 74. The defer_ok option of callout verification was not working if it was used
935 when verifying addresses in header lines, that is, for this case:
937 verify = header_sender/callout=defer_ok
939 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
940 those file descriptors could be used for SMTP connections. If anything
941 wrote to stderr (the example that came up was "warn" in embedded Perl), it
942 could be sent to the SMTP client, causing chaos. The daemon now opens
943 stdin, stdout, and stderr to /dev/null when it puts itself into the
946 76. Arrange for output from Perl's "warn" command to be written to Exim's main
947 log by default. The user can override this with suitable Perl magic.
949 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
950 the log message when discard triggers, was not working for the DATA ACL or
951 for the non-SMTP ACL.
953 78. Error message wording change in sieve.c.
955 79. If smtp_accept_max_per_host was set, the number of connections could be
956 restricted to fewer than expected, because the daemon was trying to set up
957 a new connection before checking whether the processes handling previous
958 connections had finished. The check for completed processes is now done
959 earlier. On busy systems, this bug wouldn't be noticed because something
960 else would have woken the daemon, and it would have reaped the completed
963 80. If a message was submitted locally by a user whose login name contained one
964 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
965 It caused a spool format error. I have fixed the spool reading code. A
966 related problem was that the "from" clause in the Received: line became
967 illegal because of the space(s). It is now covered by ${quote_local_part.
969 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
972 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
974 83. Patch from Sieve maintainer to fix unterminated string problem in
977 84. Some minor changes to the Linux configuration files to help with other
978 OS variants using glibc.
980 85. One more patch for Sieve to update vacation handling to latest spec.
983 ----------------------------------------------------
984 See the note above about the 4.44 and 4.50 releases.
985 ----------------------------------------------------
991 1. Change 4.43/35 introduced a bug that caused file counts to be
992 incorrectly computed when quota_filecount was set in an appendfile
995 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
996 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
999 3. Give more explanation in the error message when the command for a transport
1000 filter fails to execute.
1002 4. There are several places where Exim runs a non-Exim command in a
1003 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1004 was being done only for the command run by the queryprogram router. It is
1005 now done for all such subprocesses. The other cases are: ${run, transport
1006 filters, and the commands run by the lmtp and pipe transports.
1008 5. Some older OS have a limit of 256 on the maximum number of file
1009 descriptors. Exim was using setrlimit() to set 1000 as a large value
1010 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1011 systems. I've change it so that if it can't get 1000, it tries for 256.
1013 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1014 was an oversight, and furthermore, ever since the addition of extra
1015 controls (e.g. 4.43/32), the checks on when to allow different forms of
1016 "control" were broken. There should now be diagnostics for all cases when a
1017 control that does not make sense is encountered.
1019 7. $recipients is now available in the predata ACL (oversight).
1021 8. Tidy the search cache before the fork to do a delivery from a message
1022 received from the command line. Otherwise the child will trigger a lookup
1023 failure and thereby defer the delivery if it tries to use (for example) a
1024 cached ldap connection that the parent has called unbind on.
1026 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1027 of $address_data from the recipient verification was clobbered by the
1028 sender verification.
1030 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1031 was its contents. (It was OK if the option was not defined at all.)
1033 11. A "Completed" log line is now written for messages that are removed from
1034 the spool by the -Mrm option.
1036 12. $host_address is now set to the target address during the checking of
1037 ignore_target_hosts.
1039 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1040 being passed; this would have caused $sender_host_name to have been used if
1041 matching the list had actually called for a host name (not very likely,
1042 since this list is usually IP addresses). A host name is now passed as
1045 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1046 code that specifies a non-restarting handler (typically sigaction() in
1047 modern systems) in an attempt to fix a rare and obscure crash bug.
1049 15. Narrowed the window for a race in the daemon that could cause it to ignore
1050 SIGCHLD signals. This is not a major problem, because they are used only to
1051 wake it up if nothing else does.
1053 16. A malformed maildirsize file could cause Exim to calculate negative values
1054 for the mailbox size or file count. Odd effects could occur as a result.
1055 The maildirsize information is now recalculated if the size or filecount
1058 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1059 support for a long time. Removed HAVE_SYS_VFS_H.
1061 18. Updated exipick to current release from John Jetmore.
1063 19. Allow an empty sender to be matched against a lookup in an address list.
1064 Previously the only cases considered were a regular expression, or an
1067 20. Exim went into a mad DNS lookup loop when doing a callout where the
1068 host was specified on the transport, if the DNS lookup yielded more than
1071 21. The RFC2047 encoding function was originally intended for short strings
1072 such as real names; it was not keeping to the 75-character limit for
1073 encoded words that the RFC imposes. It now respects the limit, and
1074 generates multiple encoded words if necessary. To be on the safe side, I
1075 have increased the buffer size for the ${rfc2047: expansion operator from
1078 22. Failure to deliver a bounce message always caused it to be frozen, even if
1079 there was an errors_to setting on the router. The errors_to setting is now
1082 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1083 canonical form (fully expanded) before being placed in
1084 $sender_host_address.
1086 24. Updated eximstats to version 1.33
1088 25. Include certificate and key file names in error message when GnuTLS fails
1089 to set them up, because the GnuTLS error message doesn't include the name
1090 of the failing file when there is a problem reading it.
1092 26. Expand error message when OpenSSL has problems setting up cert/key files.
1095 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1096 (this can affect the format of dates).
1098 28. exim_tidydb, when checking for the continued existence of a message for
1099 which it has found a message-specific retry record, was not finding
1100 messages that were in split spool directories. Consequently, it was
1101 deleting retry records that should have stayed in existence.
1103 29. eximstats updated to version 1.35
1104 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1105 1.35 - bugfix such that pie charts by volume are generated correctly
1107 30. The SPA authentication driver was not abandoning authentication and moving
1108 on to the next authenticator when an expansion was forced to fail,
1109 contradicting the general specification for all authenticators. Instead it
1110 was generating a temporary error. It now behaves as specified.
1112 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1113 (the order specifies the preference for clients). The order is now AES256,
1114 AES128, 3DES, ARCFOUR128.
1116 31. Small patch to Sieve code - explicitly set From: when generating an
1119 32. Exim crashed if a remote delivery caused a very long error message to be
1120 recorded - for instance if somebody sent an entire SpamAssassin report back
1121 as a large number of 550 error lines. This bug was coincidentally fixed by
1122 increasing the size of one of Exim's internal buffers (big_buffer) that
1123 happened as part of the Exiscan merge. However, to be on the safe side, I
1124 have made the code more robust (and fixed the comments that describe what
1127 33. Some experimental protocols are using DNS PTR records for new purposes. The
1128 keys for these records are domain names, not reversed IP addresses. The
1129 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1130 leaves it alone. Component reversal etc. now happens only for IP addresses.
1133 34. The host_aton() function is supposed to be passed a string that is known
1134 to be a valid IP address. However, in the case of IPv6 addresses, it was
1135 not checking this. This is a hostage to fortune. Exim now panics and dies
1136 if the condition is not met. A case was found where this could be provoked
1137 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1138 components; fortuitously, this particular loophole had already been fixed
1139 by change 4.50/55 or 4.44/33 above.
1141 If there are any other similar loopholes, the new check in host_aton()
1142 itself should stop them being exploited. The report I received stated that
1143 data on the command line could provoke the exploit when Exim was running as
1144 exim, but did not say which command line option was involved. All I could
1145 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1146 running as the user.
1149 35. There was a buffer overflow vulnerability in the SPA authentication code
1150 (which came originally from the Samba project). I have added a test to the
1151 spa_base64_to_bits() function which I hope fixes it.
1154 36. The daemon start-up calls getloadavg() while still root for those OS that
1155 need the first call to be done as root, but it missed one case: when
1156 deliver_queue_load_max is set with deliver_drop_privilege. This is
1157 necessary for the benefit of the queue runner, because there is no re-exec
1158 when deliver_drop_privilege is set.
1160 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1161 was in use, was not putting the data itself into the right store pool;
1162 consequently, it could be overwritten for a subsequent message in the same
1163 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1166 38. Sometimes the final signoff response after QUIT could fail to get
1167 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1168 < 0 before doing a fflush(). This bug looks as though it goes back to the
1169 introduction of TLS in release 3.20, but "sometimes" must have been rare
1170 because the tests only now provoked it.
1176 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1177 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1178 Now it returns an integer. A number of calls were still expecting a T/F
1179 return. Fortuitously, in all cases, the tests worked in OK situations,
1180 which is the norm. However, things would have gone wrong on any write
1181 failures on the smtp file descriptor. This function is used when sending
1182 messages over SMTP and also when doing verify callouts.
1184 2. When Exim is called to do synchronous delivery of a locally submitted
1185 message (the -odf or -odi options), it no longer closes stderr before doing
1188 3. Implemented the mua_wrapper option.
1190 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1192 5. Implemented the functions header_remove(), header_testname(),
1193 header_add_at_position(), and receive_remove_recipient(), and exported them
1196 6. If an ACL "warn" statement specified the addition of headers, Exim already
1197 inserted X-ACL-Warn: at the start if there was no header name. However, it
1198 was not making this test for the second and subsequent header lines if
1199 there were newlines in the string. This meant that an invalid header could
1200 be inserted if Exim was badly configured.
1202 7. Allow an ACL "warn" statement to add header lines at the start or after all
1203 the Received: headers, as well as at the end.
1205 8. Added the rcpt_4xx retry error code.
1207 9. Added postmaster_mailfrom=xxx to callout verification option.
1209 10. Added mailfrom=xxxx to the callout verification option, for verify=
1212 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1213 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1215 12. Inserted some casts to stop certain compilers warning when using pointer
1216 differences as field lengths or precisions in printf-type calls (mostly
1217 affecting debugging statements).
1219 13. Added optional readline() support for -be (dynamically loaded).
1221 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1222 same clock tick as a message's arrival, so that its received time was the
1223 same as the "first fail" time on the retry record, and that message
1224 remained on the queue past the ultimate address timeout, every queue runner
1225 would try a delivery (because it was past the ultimate address timeout) but
1226 after another failure, the ultimate address timeout, which should have then
1227 bounced the address, did not kick in. This was a "< instead of <=" error;
1228 in most cases the first failure would have been in the next clock tick
1229 after the received time, and all would be well.
1231 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1232 being recognized when the domain list was tested by the match_domain
1233 condition in an expansion string.
1235 16. Added the ${str2b64: operator.
1237 17. Exim was always calling setrlimit() to set a large limit for the number of
1238 processes, without checking whether the existing limit was already
1239 adequate. (It did check for the limit on file descriptors.) Furthermore,
1240 errors from getrlimit() and setrlimit() were being ignored. Now they are
1241 logged to the main and panic logs, but Exim does carry on, to try to do its
1242 job under whatever limits there are.
1244 18. Imported PCRE 5.0.
1246 19. Trivial typo in log message " temporarily refused connection" (the leading
1249 20. If the log selector return_path_on_delivery was set and an address was
1250 redirected to /dev/null, the delivery process crashed because it assumed
1251 that a return path would always be set for a "successful" delivery. In this
1252 case, the whole delivery is bypassed as an optimization, and therefore no
1255 21. Internal re-arrangement: the function for sending a challenge and reading
1256 a response while authentication was assuming a zero-terminated challenge
1257 string. It's now changed to take a pointer and a length, to allow for
1258 binary data in such strings.
1260 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1262 23. Exim was not respecting finduser_retries when seeking the login of the
1263 uid under which it was called; it was always trying 10 times. (The default
1264 setting of finduser_retries is zero.) Also, it was sleeping after the final
1265 failure, which is pointless.
1267 24. Implemented tls_on_connect_ports.
1269 25. Implemented acl_smtp_predata.
1271 26. If the domain in control=submission is set empty, Exim assumes that the
1272 authenticated id is a complete email address when it generates From: or
1273 Sender: header lines.
1275 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1276 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1277 chown and chgrp in /bin and hostname in /usr/bin.
1279 28. Exim was keeping the "process log" file open after each use, just as it
1280 does for the main log. This opens the possibility of it remaining open for
1281 long periods when the USR1 signal hits a daemon. Occasional processlog
1282 errors were reported, that could have been caused by this. Anyway, it seems
1283 much more sensible not to leave this file open at all, so that is what now
1286 29. The long-running daemon process does not normally write to the log once it
1287 has entered its main loop, and it closes the log before doing so. This is
1288 so that log files can straightforwardly be renamed and moved. However,
1289 there are a couple of unusual error situations where the daemon does write
1290 log entries, and I had neglected to close the log afterwards.
1292 30. The text of an SMTP error response that was received during a remote
1293 delivery was being truncated at 512 bytes. This is too short for some of
1294 the long messages that one sometimes sees. I've increased the limit to
1297 31. It is now possible to make retry rules that apply only when a message has a
1298 specific sender, in particular, an empty sender.
1300 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1301 it possible to be selective about when SMTP synchronization is enforced.
1303 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1305 32. Implemented hosts_connection_nolog.
1307 33. Added an ACL for QUIT.
1309 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1312 35. Added mailbox_size and mailbox_filecount to appendfile.
1314 36. Added control = no_multiline_responses to ACLs.
1316 37. There was a bug in the logic of the code that waits for the clock to tick
1317 in the case where the clock went backwards by a substantial amount such
1318 that the microsecond fraction of "now" was more than the microsecond
1319 fraction of "then" (but the whole seconds number was less).
1321 38. Added support for the libradius Radius client library this is found on
1322 FreeBSD (previously only the radiusclient library was supported).
1328 1. When certain lookups returned multiple values in the form name=value, the
1329 quoting of the values was not always being done properly. Specifically:
1330 (a) If the value started with a double quote, but contained no whitespace,
1332 (b) If the value contained whitespace other than a space character (i.e.
1333 tabs or newlines or carriage returns) it was not quoted.
1334 This fix has been applied to the mysql and pgsql lookups by writing a
1335 separate quoting function and calling it from the lookup code. The fix
1336 should probably also be applied to nisplus, ibase and oracle lookups, but
1337 since I cannot test any of those, I have not disturbed their existing code.
1339 2. A hit in the callout cache for a specific address caused a log line with no
1340 reason for rejecting RCPT. Now it says "Previous (cached) callout
1341 verification failure".
1343 3. There was an off-by-one bug in the queryprogram router. An over-long
1344 return line was truncated at 256 instead of 255 characters, thereby
1345 overflowing its buffer with the terminating zero. As well as fixing this, I
1346 have increased the buffer size to 1024 (and made a note to document this).
1348 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1349 when Exim is waiting for an SMTP response from a remote server, Exim
1350 restarts its select() call on the socket, thereby resetting its timeout.
1351 This is not a problem when such interrupts are rare. Somebody set up a cron
1352 job to run exiwhat every 2 minutes, which is less than the normal select()
1353 timeout (5 or 10 minutes). This meant that the select() timeout never
1354 kicked in because it was always reset. I have fixed this by comparing the
1355 time when an interrupt arrives with the time at the start of the first call
1356 to select(). If more time than the timeout has elapsed, the interrupt is
1357 treated as a timeout.
1359 5. Some internal re-factoring in preparation for the addition of Sieve
1360 extensions (by MH). In particular, the "personal" test is moved to a
1361 separate function, and given an option for scanning Cc: and Bcc: (which is
1362 not set for Exim filters).
1364 6. When Exim created an email address using the login of the caller as the
1365 local part (e.g. when creating a From: or Sender: header line), it was not
1366 quoting the local part when it contained special characters such as @.
1368 7. Installed new OpenBSD configuration files.
1370 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1371 try to make them clearer.
1373 9. Callout options, other than the timeout value, were being ignored when
1374 verifying sender addresses in header lines. For example, when using
1376 verify = header_sender/callout=no_cache
1378 the cache was (incorrectly) being used.
1380 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1381 only the Cygwin environment.
1383 11. When return_path_on_delivery was set as a log selector, if different remote
1384 addresses in the same message used different return paths and parallel
1385 remote delivery occurred, the wrong values would sometimes be logged.
1386 (Whenever a remote delivery process finished, the return path value from
1387 the most recently started remote delivery process was logged.)
1389 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1390 lines when AUTH and/or TLS are in use. This is the "received protocol"
1391 field. Exim used to use "asmtp" for authenticated SMTP, without any
1392 indication (in the protocol name) for TLS use. Now it follows the RFC and
1393 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1394 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1395 names appear in log lines as well as in Received: header lines.
1397 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1398 extensions, and comparison tests, and to fix some bugs.
1400 14. Changes to the "personal" filter test:
1402 (1) The test was buggy in that it was just doing the equivalent of
1403 "contains" tests on header lines. For example, if a user's address was
1404 anne@some.where, the "personal" test would incorrectly be true for
1406 To: susanne@some.where
1408 This test is now done by extracting each address from the header in turn,
1409 and checking the entire address. Other tests that are part of "personal"
1410 are now done using regular expressions (for example, to check local parts
1411 of addresses in From: header lines).
1413 (2) The list of non-personal local parts in From: addresses has been
1414 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1415 taken from the Sieve specification recommendations.
1417 (3) If the message contains any header line starting with "List-" it is
1418 treated as non-personal.
1420 (4) The test for "circular" in the Subject: header line has been removed
1421 because it now seems ill-conceived.
1423 15. Minor typos in src/EDITME comments corrected.
1425 16. Installed latest exipick from John Jetmore.
1427 17. If headers_add on a router specified a text string that was too long for
1428 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1429 of string_sprintf() is now avoided.
1431 18. $message_body_size was not set (it was always zero) when running the DATA
1432 ACL and the local_scan() function.
1434 19. For the "mail" command in an Exim filter, no default was being set for
1435 the once_repeat time, causing a random time value to be used if "once" was
1436 specified. (If the value happened to be <= 0, no repeat happened.) The
1437 default is now 0s, meaning "never repeat". The "vacation" command was OK
1438 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1439 (I found it when inspecting the code).
1441 20. There is now an overall timeout for performing a callout verification. It
1442 defaults to 4 times the callout timeout, which applies to individual SMTP
1443 commands during the callout. The overall timeout applies when there is more
1444 than one host that can be tried. The timeout is checked before trying the
1445 next host. This prevents very long delays if there are a large number of
1446 hosts and all are timing out (e.g. when the network connections are timing
1447 out). The value of the overall timeout can be changed by specifying an
1448 additional sub-option for "callout", called "maxwait". For example:
1450 verify = sender/callout=5s,maxwait=20s
1452 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1453 to the end before writing, but this should make it even safer).
1455 22. Exim was forgetting that it had advertised PIPELINING for the second and
1456 subsequent messages on an SMTP connection. It was also not resetting its
1457 memory on STARTTLS and an internal HELO.
1459 23. When Exim logs an SMTP synchronization error within a session, it now
1460 records whether PIPELINING has been advertised or not.
1462 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1463 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1464 rather than long int.
1466 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1468 26. Added the never_mail option to autoreply.
1474 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1475 crash if the getsockname() call failed; this can happen if a connection is
1476 closed very soon after it is established. The problem was simply in the
1477 order in which certain operations were done, causing Exim to try to write
1478 to the SMTP stream before it had set up the file descriptor. The bug has
1479 been fixed by making things happen in the correct order.
1485 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1486 before the connection was closed, thus losing the rejection response.
1488 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1489 some early Solaris releases, but causes trouble in current releases where
1490 socklen_t is defined.
1492 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1495 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1498 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1499 that releasing the top of it at the end releases what was used for sub-
1500 expansions (unless the block got too big). However, discard this block if
1501 the first thing is a variable or header, so that we can use its block when
1502 it is dynamic (useful for very large $message_headers, for example).
1504 6. Lookups now cache *every* query, not just the most recent. A new, separate
1505 store pool is used for this. It can be recovered when all lookup caches are
1506 flushed. Lookups now release memory at the end of their result strings.
1507 This has involved some general refactoring of the lookup sources.
1509 7. Some code has been added to the store_xxx() functions to reduce the amount
1510 of flapping under certain conditions.
1512 8. log_incoming_interface used to affect only the <= reception log lines. Now
1513 it causes the local interface and port to be added to several more SMTP log
1514 lines, for example "SMTP connection from", and rejection lines.
1516 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1518 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1520 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1521 could be overwritten at the end of the current message (or the start of a
1522 new message if it was set in a HELO ACL). The value is now preserved for
1523 the duration of the SMTP connection.
1525 12. If a transport had a headers_rewrite setting, and a matching header line
1526 contained an unqualified address, that address was qualified, even if it
1527 did not match any rewriting rules. The underlying bug was that the values
1528 of the flags that permit the existence of unqualified sender and recipient
1529 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1530 non-local messages, and by -bnq for local messages) were not being
1531 preserved with the message after it was received.
1533 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1534 "next input=" as part of the text comprising the host identity instead of
1535 the correct text. The code was using the same buffer for two different
1536 strings. However, depending on which order the printing function evaluated
1537 its arguments, the bug did not always show up. Under Linux, for example, my
1538 test suite worked just fine.
1540 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1541 doesn't work with some older versions of Perl. It has been changed to "my",
1542 which in any case is probably the better facility to use.
1544 15. A really picky compiler found some instances of statements for creating
1545 error messages that either had too many or two few arguments for the format
1548 16. The size of the buffer for calls to the DNS resolver has been increased
1549 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1550 for addresses that have a lot of PTR records. This alleviates a problem; it
1551 does not fully solve it.
1553 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1554 buffer now truncates the list and logs the incident, which is the same
1555 action as happens when Exim is looking up a host name and its aliases.
1556 Previously in this situation something unpredictable would happen;
1557 sometimes it was "internal error: store_reset failed".
1559 18. If a server dropped the connection unexpectedly when an Exim client was
1560 using GnuTLS and trying to read a response, the client delivery process
1561 crashed while trying to generate an error log message.
1563 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1564 string, for example:
1566 warn message = H1: something\nH2: something
1568 the text was added as a single header line from Exim's point of view
1569 though it ended up OK in the delivered message. However, searching for the
1570 second and subsequent header lines using $h_h2: did not work. This has been
1571 fixed. Similarly, if a system filter added multiple headers in this way,
1572 the routers could not see them.
1574 20. Expanded the error message when iplsearch is called with an invalid key to
1575 suggest using net-iplsearch in a host list.
1577 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1578 ACLs are no longer actually imposed (and a message to that effect is
1581 22. If a "gecos" field in a passwd entry contained escaped characters, in
1582 particular, if it contained a \" sequence, Exim got it wrong when building
1583 a From: or a Sender: header from that name. A second bug also caused
1584 incorrect handling when an unquoted " was present following a character
1585 that needed quoting.
1587 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1588 was not being matched caselessly.
1590 24. Arranged for all hyphens in the exim.8 source to be escaped with
1593 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1594 properly. Recipient callout cache records were still being keyed to include
1595 the sender, even when use_sender was set false. This led to far more
1596 callouts that were necessary. The sender is no longer included in the key
1597 when use_sender is false.
1599 26. Added "control = submission" modifier to ACLs.
1601 27. Added the ${base62d: operator to decode base 62 numbers.
1603 28. dnsdb lookups can now access SRV records.
1605 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1606 the configuration file.
1608 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1609 (-v) mode. This makes the output for a verbose queue run more intelligible.
1611 31. Added a use_postmaster feature to recipient callouts.
1613 32. Added the $body_zerocount variable, containing the number of binary zero
1614 bytes in the message body.
1616 33. The time of last modification of the "new" subdirectory is now used as the
1617 "mailbox time last read" when there is a quota error for a maildir
1620 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1622 35. Added +ignore_unknown as a special item in host lists.
1624 36. Code for decoding IPv6 addresses in host lists is now included, even if
1625 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1626 address was recognized as an IP address, but was then not correctly decoded
1627 into binary, causing unexpected and incorrect effects when compared with
1634 1. Very minor rewording of debugging text in manualroute to say "list of
1635 hosts" instead of "hostlist".
1637 2. If verify=header_syntax was set, and a header line with an unqualified
1638 address (no domain) and a large number of spaces between the end of the
1639 name and the colon was received, the reception process suffered a buffer
1640 overflow, and (when I tested it) crashed. This was caused by some obsolete
1641 code that should have been removed. The fix is to remove it!
1643 3. When running in the test harness, delay a bit after writing a bounce
1644 message to get a bit more predictability in the log output.
1646 4. Added a call to search_tidyup() just before forking a reception process. In
1647 theory, someone could use a lookup in the expansion of smtp_accept_max_
1648 per_host which, without the tidyup, could leave open a database connection.
1650 5. Added the variables $recipient_data and $sender_data which get set from a
1651 lookup success in an ACL "recipients" or "senders" condition, or a router
1652 "senders" option, similar to $domain_data and $local_part_data.
1654 6. Moved the writing of debug_print from before to after the "senders" test
1657 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1658 problems for message scanning, either using a data ACL, or using
1659 local_scan() because the Received: header was not generated till after they
1660 were called (in order to set the time as the time of reception completion).
1661 I have revised the way this works. The header is now generated after the
1662 body is received, but before the ACL or local_scan() are called. After they
1663 are run, the timestamp in the header is updated.
1669 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1670 before starting a queue runner without re-exec. This happened only when
1671 deliver_drop_privilege was set or when the Exim user was set to root. The
1672 effect of the bug was that timeouts during subsequent deliveries caused
1673 crashes instead of being properly handled. The handler is now left at its
1674 default (and expected) setting.
1676 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1677 message, again when deliver_drop_privilege is set or Exim is run as root.
1678 The bug described in (1) was not present in this case, but the tidying up
1679 of the other signals was missing. I have made the two cases consistent.
1681 3. The ignore_target_hosts setting on a manualroute router was being ignored
1682 for hosts that were looked up using the /MX notation.
1684 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1687 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1688 operated on the sender address. After changing the $sender_address to <>
1689 for the sender address verify, Exim was re-instated it as the original
1690 (before rewriting) address, but remembering that it had rewritten it, so it
1691 wasn't rewriting it again. This bug also had the effect of breaking the
1692 sender address verification caching when the sender address was rewritten.
1694 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1695 This has been changed so that if the ip literal address matches
1696 ignore_target_hosts, the router declines.
1698 7. Added expansion conditions match_domain, match_address, and match_local_
1699 part (NOT match_host).
1701 8. The placeholder for the Received: header didn't have a length field set.
1703 9. Added code to Exim itself and to exim_lock to test for a specific race
1704 condition that could lead to file corruption when using MBX delivery. The
1705 issue is with the lockfile that is created in /tmp. If this file is removed
1706 after a process has opened it but before that process has acquired a lock,
1707 there is the potential for a second process to recreate the file and also
1708 acquire a lock. This could lead to two Exim processes writing to the file
1709 at the same time. The added code performs the same test as UW imapd; it
1710 checks after acquiring the lock that its file descriptor still refers to
1711 the same named file.
1713 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1714 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1717 11. Added the smtp_active_hostname option. If used, this will typically be made
1718 to depend on the incoming interface address. Because $interface_address is
1719 not set up until the daemon has forked a reception process, error responses
1720 that can happen earlier (such as "too many connections") no longer contain
1723 12. If an expansion in a condition on a "warn" statement fails because a lookup
1724 defers, the "warn" statement is abandoned, and the next ACL statement is
1725 processed. Previously this caused the whole ACL to be aborted.
1727 13. Added the iplsearch lookup type.
1729 14. Added ident_timeout as a log selector.
1731 15. Added tls_certificate_verified as a log selector.
1733 16. Added a global option tls_require_ciphers (compare the smtp transport
1734 option of the same name). This controls incoming TLS connections.
1736 17. I finally figured out how to make tls_require_ciphers do a similar thing
1737 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1738 before starting the TLS session.
1740 18. Tabs are now shown as \t in -bP output.
1742 19. If the log selector return_path_on_delivery was set, Exim crashed when
1743 bouncing a message because it had too many Received: header lines.
1745 20. If two routers both had headers_remove settings, and the first one included
1746 a superfluous trailing colon, the final name in the first list and the
1747 first name in the second list were incorrectly joined into one item (with a
1748 colon in the middle).
1754 1. Added -C and -D options to the exinext utility, mainly to make it easier
1755 to include in the automated testing, but these could be helpful when
1756 multiple configurations are in use.
1758 2. The exinext utility was not formatting the output nicely when there was
1759 an alternate port involved in the retry record key, nor when there was a
1760 message id as well (for retries that were specific to a specific message
1761 and a specific host). It was also confused by IPv6 addresses, because of
1762 the additional colons they contain. I have fixed the IPv4 problem, and
1763 patched it up to do a reasonable job for IPv6.
1765 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1766 delivery, the log line now contains "pipelined" if PIPELINING was used.
1768 4. An SMTP transport process used to panic and die if the bind() call to set
1769 an explicit outgoing interface failed. This has been changed; it is now
1770 treated in the same way as a connect() failure.
1772 5. A reference to $sender_host_name in the part of a conditional expansion
1773 that was being skipped was still causing a DNS lookup. This no longer
1776 6. The def: expansion condition was not recognizing references to header lines
1777 that used bh_ and bheader_.
1779 7. Added the _cache feature to named lists.
1781 8. The code for checking quota_filecount in the appendfile transport was
1782 allowing one more file than it should have been.
1784 9. For compatibility with Sendmail, the command line option
1792 and sets the incoming protocol and host name (for trusted callers). The
1793 host name and its colon can be omitted when only the protocol is to be set.
1794 Note the Exim already has two private options, -pd and -ps, that refer to
1795 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1796 "s", but I don't think that's a major issue.
1798 10. A number of refactoring changes to the code, none of which should affect
1801 (a) The number of logging options was getting close to filling up the
1802 32-bit word that was used as a bit map. I have split them into two classes:
1803 those that are passed in the argument to log_write(), and those that are
1804 only ever tested independently outside of that function. These are now in
1805 separate 32-bit words, so there is plenty of room for expansion again.
1806 There is no change in the user interface or the logging behaviour.
1808 (b) When building, for example, log lines, the code previously used a
1809 macro that called string_cat() twice, in order to add two strings. This is
1810 not really sufficiently general. Furthermore, there was one instance where
1811 it was actually wrong because one of the argument was used twice, and in
1812 one call a function was used. (As it happened, calling the function twice
1813 did not affect the overall behaviour.) The macro has been replaced by a
1814 function that can join an arbitrary number of extra strings onto a growing
1817 (c) The code for expansion conditions now uses a table and a binary chop
1818 instead of a serial search (which was left over from when there were very
1819 few conditions). Also, it now recognizes conditions like "pam" even when
1820 the relevant support is not compiled in: a suitably worded error message is
1821 given if an attempt is made to use such a condition.
1823 11. Added ${time_interval:xxxxx}.
1825 12. A bug was causing one of the ddress fields not to be passed back correctly
1826 from remote delivery subprocesses. The field in question was not being
1827 subsequently used, so this caused to problems in practice.
1829 13. Added new log selectors queue_time and deliver_time.
1831 14. Might have fixed a bug in maildirsizefile handling that threw up
1832 "unexpected character" debug warnings, and recalculated the data
1833 unnecessarily. In any case, I expanded the warning message to give more
1836 15. Added the message "Restricted characters in address" to the statements in
1837 the default ACL that block characters like @ and % in local parts.
1839 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1840 Three changes have been made:
1842 (a) There was a serious bug; a negative response to MAIL caused the whole
1843 recipient domain to be cached as invalid, thereby blocking all messages
1844 to all local parts at the same domain, from all senders. This bug has
1845 been fixed. The domain is no longer cached after a negative response to
1846 MAIL if the sender used is not empty.
1848 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1851 (c) A new callout option, "use_sender" has been added for people who want
1852 the modified behaviour.
1858 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1861 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1862 indeed breaks things for older releases.
1864 3. Added additional logging to the case where there is a problem reading data
1865 from a filter that is running in a subprocess using a pipe, in order to
1866 try to track down a specific problem.
1868 4. Testing facility fudge: when running in the test harness and attempting
1869 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1870 getting "No route to host". Convert this to a timeout.
1872 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1875 6. Some OS don't have socklen_t but use size_t instead. This affects the
1876 fifth argument of getsockopt() amongst other things. This is now
1877 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1878 can be set for individual OS. I have set it for SunOS5, OSF1, and
1879 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1880 some earlier ones do not.
1882 7. Change 4.30/15 was not doing the test caselessly.
1884 8. The standard form for an IPv6 address literal was being rejected by address
1885 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1886 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1887 this, as well as the form without the "IPv6" on the front (but only when
1888 address literals are enabled, of course).
1890 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1892 10. Exim crashed if a message with an empty sender address specified by -f
1893 encountered a router with an errors_to setting. This could be provoked only
1894 by a command such as
1898 where an empty string was supplied; "<>" did not hit this bug.
1900 11. Installed PCRE release 4.5.
1902 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1903 remained set. It is now erased.
1905 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1906 times from message ids (which are base 36 rather than the normal 62).
1908 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1909 were being counted as actual protocol errors, and logged if the log
1910 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1911 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1912 and DATA following a set of rejected RCPTs do not count as protocol errors.
1913 In other words, Exim assumes they were pipelined, though this may not
1914 actually be the case. Of course, in all cases the client gets an
1915 appropriate error code.
1917 15. If a lookup fails in an ACL condition, a message about the failure may
1918 be available; it is used if testing the ACL cannot continue, because most
1919 such messages specify what the cause of the deferral is. However, some
1920 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1921 that caused an old message to be retained and used if a later statement
1922 caused a defer, replacing the real cause of the deferral.
1924 16. If an IP address had so many PTR records that the DNS lookup buffer
1925 was not large enough to hold them, Exim could crash while trying to process
1926 the truncated data. It now detects and logs this case.
1928 17. Further to 4.21/58, another change has been made: if (and only if) the
1929 first line of a message (the first header line) ends with CRLF, a bare LF
1930 in a subsequent header line has a space inserted after it, so as not to
1931 terminate the header.
1933 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1934 unnecessarily, used atoi() instead of strtol(), and didn't check the
1935 termination when getting file sizes from file names by regex.
1937 19. Completely re-implemented the support for maildirsize files, in the light
1938 of a number of problems with the previous contributed implementation
1939 (4.30/29). In particular:
1941 . If the quota is zero, the maildirsize file is maintained, but no quota is
1944 . If the maildir directory does not exist, it is created before any attempt
1945 to write a maildirsize file.
1947 . The quota value in the file is just a cache; if the quota is changed in
1948 the transport, the new value overrides.
1950 . A regular expression is available for excluding directories from the
1953 20. The autoreply transport checks the characters in options that define the
1954 message's headers; it allows continued headers, but it was checking with
1955 isspace() after an embedded newline instead of explicitly looking for a
1958 21. If all the "regular" hosts to which an address was routed had passed their
1959 expiry times, and had not reached their retry times, the address was
1960 bounced, even if fallback hosts were defined. Now Exim should go on to try
1963 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1964 equivalent code in the SMTP transport. Some hosts send humungous responses
1965 to HELO/EHLO, more than 1024 it seems.
1967 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1968 gives compiler warnings in some environments. I've now done it "properly",
1971 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1972 (because of problems with the built-in one) was declared to return uschar *
1973 instead of char *, causing compiler failure.
1975 25. Fixed a file descriptor leak when processing alias/forward files.
1977 26. Fixed a minor format string issue in dbfn.c.
1979 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1981 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1982 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1985 29. When a sender address is verified, it is cached, to save repeating the test
1986 when there is more than one recipient in a message. However, when the
1987 verification involves a callout, it is possible for different callout
1988 options to be set for different recipients. It is too complicated to keep
1989 track of this in the cache, so now Exim always runs a verification when a
1990 callout is required, relying on the callout cache for the optimization.
1991 The overhead is duplication of the address routing, but this should not be
1994 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
1995 to be verified with callout=postmaster, and the main callout worked but the
1996 postmaster check failed, the verification correctly failed. However, if a
1997 subsequent RCPT command asked for sender verification *without* the
1998 postmaster check, incorrect caching caused this verification also to fail,
2001 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2002 it was not caching the DNS options (qualify_single, search_parents) that
2003 were used when the lookup failed. A subsequent lookup with different
2004 options therefore always gave the same answer, though there were cases
2005 where it should not have. (Example: a "domains = !$mx_any" option on a
2006 dnslookup router: the "domains" option is always processed without any
2007 widening, but the router might have qualify_single set.) Now Exim uses the
2008 cached value only when the same options are set.
2010 32. Added John Jetmore's "exipick" utility to the distribution.
2012 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2013 than a timeout (e.g. a certificate is required, and is not provided), an
2014 Exim server now closes the connection immediately. Previously it waited for
2015 the client to close - but if the client is SSL, it seems that they each
2016 wait for each other, leading to a delay before one of them times out.
2018 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2019 maintained 0.8.x compatibility because I don't think many are using it, and
2020 it is clearly obsolete.
2022 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2025 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2026 client certificate was expired. A simple patch fixes this, though I don't
2027 understand the full logic of why the verify callback is called multiple
2030 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2031 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2032 which causes problems with some clients (such as the Certicom SSL Plus
2033 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2034 disables the coutermeasure allowing Eudora to connect."
2036 38. Exim was not checking that a write() to a log file succeeded. This could
2037 lead to Bad Things if a log got too big, in particular if it hit a file
2038 size limit. Exim now panics and dies if it cannot write to a log file, just
2039 as it does if it cannot open a log file.
2041 39. Modified OS/Makefile-Linux so that it now contains
2043 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2045 The two -D definitions ensure that Exim is compiled with large file
2046 support, which makes it possible to handle log files that are bigger than
2049 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2050 instance) a domain was checked against a named list that involved a lookup,
2051 causing $domain_data to be set, then another domain was checked against the
2052 same list, then the first domain was re-checked, the value of $domain_data
2053 after the final check could be wrong. In particular, if the second check
2054 failed, it could be set empty. This bug probably also applied to
2057 41. The strip_trailing_dot option was not being applied to the address given
2058 with the -f command-line option.
2060 42. The code for reading a message's header from the spool was incrementing
2061 $received_count, but never initializing it. This meant that the value was
2062 incorrect (doubled) while delivering a message in the same process in which
2063 it was received. In the most common configuration of Exim, this never
2064 happens - a fresh exec is done - but it can happen when
2065 deliver_drop_privilege is set.
2067 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2068 it now includes up to 150 characters of the unexpected data in the log
2071 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2072 and building data strings. The size of both of these buffers was 10 000
2073 bytes - far larger than anybody would *ever* want, thought I. Needless to
2074 say, somebody hit the limit. I have increased the maximum line length to
2075 20 000 and the maximum data length of concatenated lines to 100 000. I have
2076 also fixed two bugs, because there was no checking on these buffers. Tsk,
2077 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2078 buffer is too small.
2080 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2081 lsearch lookups. Now it does.
2083 46. When parsing a route_list item in a manualroute router, a fixed-length
2084 buffer was used for the list of hosts. I made this 1024 bytes long,
2085 thinking that nobody would ever have a list of hosts that long. Wrong.
2086 Somebody had a whole pile of complicated expansion conditions, and the
2087 string was silently truncated, leading to an expansion error. It turns out
2088 that it is easier to change to an unlimited length (owing to other changes
2089 that have happened since this code was originally written) than to build
2090 structure for giving a limitation error. The length of the item that
2091 expands into the list of hosts is now unlimited.
2093 47. The lsearch lookup could not handle data where the length of text line was
2094 more than 4095 characters. Such lines were truncated, leading to shortened
2095 data being returned. It should now handle lines of any length.
2097 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2098 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2101 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2102 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2103 no longer happens in comment lines. A list of replacements is now placed
2104 at the head of all of the source files, except those whose only change is
2105 to replace PERL_COMMAND in the very first #! line.
2107 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2108 messages on the queue, with a bottom-up merge sort, using code contributed
2109 by Michael Haardt. This should make operations like -bp somewhat faster on
2110 large queues. It won't affect queue runners, except when queue_run_in_order
2113 51. Installed eximstats 1.31 in the distribution.
2115 52. Added support for SRV lookups to the dnslookup router.
2117 53. If an ACL referred to $message_body or $message_body_end, the value was not
2118 reset for any messages that followed in the same SMTP session.
2120 54. The store-handling optimization for building very long strings was not
2121 differentiating between the different store pools. I don't think this
2122 actually made any difference in practice, but I've tidied it.
2124 55. While running the routers to verify a sender address, $sender_address
2125 was still set to the sender address. This is wrong, because when routing to
2126 send a bounce to the sender, it would be empty. Therefore, I have changed
2127 it so that, while verifying a sender address, $sender_address is set to <>.
2128 (There is no change to what happens when verifying a recipient address.)
2130 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2131 target A or AAAA records (if not already returned) without resetting the
2132 qualify_single or search_parents options of the DNS resolver. These are
2133 inappropriate in this case because the targets of MX and SRV records must
2134 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2135 target that, when qualified, matched something in the local domain. These
2136 two options are now turned off when doing these lookups.
2138 57. It seems that at least some releases of Reiserfs (which does not have the
2139 concept of a fixed number of inodes) returns zero and not -1 for the
2140 number of available inodes. This interacted badly with check_spool_inodes,
2141 which assumed that -1 was the "no such thing" setting. What I have done is
2142 to check that the total number of inodes is greater than zero before doing
2143 the test of how many are available.
2145 58. When a "warn" ACL statement has a log_message modifier, the message is
2146 remembered, and not repeated. This is to avoid a lot of repetition when a
2147 message has many recipients that cause the same warning to be written.
2148 Howewer, Exim was preserving the list of already written lines for an
2149 entire SMTP session, which doesn't seem right. The memory is now reset if a
2150 new message is started.
2152 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2153 debugging output unless log_rewrite was also set.
2155 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2156 of (int)(handle) when we know that handle contains (void *)(-1).
2158 61. The Exim daemon panic-logs an error return when it closes the incoming
2159 connection. However "connection reset by peer" seems to be common, and
2160 isn't really an error worthy of noting specially, so that particular error
2163 62. When Exim is trying to find all the local interfaces, it used to panic and
2164 die if the ioctl to get the interface flags failed. However, it seems that
2165 on at least one OS (Solaris 9) it is possible to have an interface that is
2166 included in the list of interfaces, but for which you get a failure error
2167 for this call. This happens when the interface is not "plumbed" into a
2168 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2169 failure of the "get flags" call assumes that the interface is down.
2171 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2172 makes life easier for people who are doing arithmetic on fields extracted
2173 from dates, where you often get leading zeros that should not be
2174 interpreted as octal.
2176 64. Added qualify_domain to the redirect router, to override the global
2179 65. If a pathologically long header line contained very many addresses (the
2180 report of this problem mentioned 10 000) and each of them was rewritten,
2181 Exim could use up a very large amount of memory. (It kept on making new
2182 copies of the header line as it rewrote, and never released the old ones.)
2183 At the expense of a bit more processing, the header rewriting function has
2184 been changed so that it no longer eats memory in this way.
2186 66. The generation of the Received: header has been moved from the time that a
2187 message starts to be received, to the time that it finishes. The timestamp
2188 in the Received: header should now be very close to that of the <= log
2189 line. There are two side-effects of this change:
2191 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2192 logged header lines no longer include the local Received: line, because
2193 it has not yet been created. The same applies to a copy of the message
2194 that is returned to a non-SMTP sender when a message is rejected.
2196 (b) When a filter file is tested using -bf, no additional Received: header
2197 is added to the test message. After some thought, I decided that this
2200 This change does not affect the value of $received_for. It is still set
2201 after address rewriting, but before local_scan() is called.
2203 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2205 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2206 gave an unhelpful panic error message, and a defer error. I have managed to
2207 change this behaviour so that it now rejects any supplied certificate,
2208 which seems right, as the list of acceptable certificates is empty.
2210 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2211 gave an unhelpful defer error. I have not managed to make this reject any
2212 supplied certificates, but the error message it gives is "no certificate
2213 supplied", which is not helpful.
2215 70. exigrep's output now also includes lines that are not associated with any
2216 message, but which match the given pattern. Implemented by a patch from
2217 Martin Sluka, which also tidied up the Perl a bit.
2219 71. Recipient callout verification, like sender verification, was using <> in
2220 the MAIL FROM command. This isn't really the right thing, since the actual
2221 sender may affect whether the remote host accepts the recipient or not. I
2222 have changed it to use the actual sender in the callout; this means that
2223 the cache record is now keyed on a recipient/sender pair, not just the
2224 recipient address. There doesn't seem to be a real danger of callout loops,
2225 since a callout by the remote host to check the sender would use <>.
2226 [SEE ABOVE: changed after hitting problems.]
2228 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2229 temporary errors. However, in the case of such a code being given after
2230 the end of a data transmission (i.e. after ".") Exim was failing to write
2231 a retry record for the message. (Yes, there was some broken host that was
2232 actually sending 8xx at this point.)
2234 73. An unknown lookup type in a host list could cause Exim to panic-die when
2235 the list was checked. (An example that provoked this was putting <; in the
2236 middle of a list instead of at the start.) If this happened during a DATA
2237 ACL check, a -D file could be left lying around. This kind of configuration
2238 error no longer causes Exim to die; instead it causes a defer errror. The
2239 incident is still logged to the main and panic logs.
2241 74. Buglet left over from Exim 3 conversion. The message "too many messages
2242 in one connection" was written to the rejectlog but not the mainlog, except
2243 when address rewriting (yes!) was being logged.
2245 75. Added write_rejectlog option.
2247 76. When a system filter was run not as root (that is, when system_filter_user
2248 was set), the values of the $n variables were not being returned to the
2249 main process; thus, they were not subsequently available in the $sn
2252 77. Added +return_path_on_delivery log selector.
2254 78. A connection timeout was being treated differently from recipients deferred
2255 when testing hosts_max_try with a message that was older than the host's
2256 retry timeout. (The host should not be counted, thus allowing all hosts to
2257 be tried at least once before bouncing.) This may have been the cause of an
2258 occasionally reported bug whereby a message would remain on the queue
2259 longer than the retry timeout, but would be bounced if a delivery was
2260 forced. I say "may" because I never totally pinned down the problem;
2261 setting up timeout/retry tests is difficult. See also the next item.
2263 79. The ultimate address timeout was not being applied to errors that involved
2264 a combination of host plus message (for example, a timeout on a MAIL
2265 command). When an address resolved to a number of possible hosts, and they
2266 were not all tried for each delivery (e.g. because of hosts_max_try), a
2267 message could remain on the queue longer than the retry timeout.
2269 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2272 81. Fixed an obscure SMTP outgoing bug which required at least the following
2273 conditions: (a) there was another message waiting for the same server;
2274 (b) the server returned 5xx to all RCPT commands in the first message so
2275 that the message was not completed; (c) the server dropped the connection
2276 or gave a negative response to the RSET that Exim sends to abort the
2277 transaction. The observed case was a dropped connection after DATA that had
2278 been sent in pipelining mode. That is, the server had advertised PIPELINING
2279 but was not implementing it correctly. The effect of the bug was incorrect
2280 behaviour, such as trying another host, and this could lead to a crash.
2286 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2287 and daemon.c were passed as pointers to ints; they should have been
2288 pointers to socklen_t variables (which are typically unsigned ints).
2290 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2293 3. Fixed a really odd bug that affected only the testing scheme; patching a
2294 certain fixed string in the binary changed the value of another string that
2295 happened to be identical to the end of the original first string.
2297 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2298 name", it returns that address as the IP address. On some operating
2299 systems (e.g. Solaris), it also passes back the IP address string as the
2300 "host name". However, on others (e.g. Linux), it passes back an empty
2301 string. Exim wasn't checking for this, and was changing the host name to an
2302 empty string, assuming it had been canonicized.
2304 5. Although rare, it is permitted to have more than one PTR record for a given
2305 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2306 all the names associated with an address, because they do in Solaris.
2307 However, it seems that they do not in Linux for data that comes from the
2308 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2309 I found this out when I moved to a new Linux workstation and tried to run
2310 the Exim test suite.
2312 To get round this problem I have changed the code so that it now does its
2313 own call to the DNS to look up PTR records when searching for a host name.
2314 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2315 addresses that are only in /etc/hosts are still found.
2317 This behaviour is, however, controlled by an option called host_lookup_
2318 order, which defaults to "bydns:byaddr". If people want to use the other
2319 order, or indeed, just use one or the other means of lookup, they can
2320 specify it in this variable.
2322 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2323 some operating systems, this comes back from gethostbyaddr() as an empty
2324 string, and this is what Exim used to test for. However, it seems that in
2325 other systems, "." is yielded. Exim now tests for this case too.
2327 7. The values of check_spool_space and check_log_space are now held internally
2328 as a number of kilobytes instead of an absolute number of bytes. If a
2329 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2330 kilobyte. This means that much larger values can be stored.
2332 8. Exim monitor: an attempt to get the action menu when not actually pointing
2333 at a message produces an empty menu entitled "No message selected". This
2334 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2335 no entries in it ("Shell widget menu has zero width and/or height"). So I
2336 have added a single, blank menu entry in this case.
2338 9. Added ${quote_local_part.
2340 10. MIME decoding is now applied to the contents of Subject: header lines when
2343 11. Now that a reference to $sender_host_address automatically causes a reverse
2344 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2345 host lookup before query-style lookups in lists that might use this
2346 variable. This has therefore been abolished, and the "net-" prefix is no
2347 longer necessary for query-style lookups.
2349 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2350 have been a typo for LFLAGS, so it has been changed.
2352 13. The install script calls Exim with "-C /dev/null" in order to find the
2353 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2354 to be output. Howeve, since Exim outputs its version number before the
2355 error, it didn't break the script. It just looked ugly. I fixed this by
2356 always allowing "-C /dev/null" if the caller is root.
2358 14. Ignore overlarge ACL variable number when reading spool file - insurance
2359 against a later release with more variables having written the file.
2361 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2362 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2363 this, as well as the form without the "IPv6" on the front.
2365 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2366 OS/Makefile-Darwin file.
2368 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2369 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2371 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2372 variable. However, some OS use ints for inodes. I've added cast to long int
2373 to get rid of the compiler warning.
2375 19. I had forgotten to lock out "/../" in configuration file names when
2376 ALT_CONFIG_PREFIX was set.
2378 20. Routers used for verification do not need to specify transports. However,
2379 if such a router generated a host list, and callout was configured, Exim
2380 crashed, because it could not find a port number from the (non-existent)
2381 transport. It now assumes port 25 in this circumstance.
2383 21. Added the -t option to exigrep.
2385 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2386 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2387 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2388 them is set, LOOKUP_LSEARCH is forced.
2390 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2392 24. Added sender and host information to the "rejected by local_scan()" log
2393 line; previously there was no indication of these.
2395 25. Added .include_if_exists.
2397 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2398 Linux. It turns out that not all file systems support this. Apparently some
2399 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2400 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2401 not supported on the file descriptor, is now ignored when Exim is trying to
2402 sync a directory. This applies only to Linux.
2404 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2406 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2407 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2410 29. Added support for maildirsize files from supplied patch (modified a bit).
2412 30. The use of :fail: followed by an empty string could lead Exim to respond to
2413 sender verification failures with (e.g.):
2415 550 Verification failed for <xxx>
2416 550 Sender verify failed
2418 where the first response line was missing the '-' that indicates it is not
2419 the final line of the response.
2421 31. The loop for finding the name of the user that called Exim had a hardwired
2422 limit of 10; it now uses the value of finduser_retries, which is used for
2423 all other user lookups.
2425 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2428 33. Exim was neglecting to zero errno before one call of strtol() when
2429 expanding a string and expecting an integer value. On some systems this
2430 resulted in spurious "integer overflow" errors. Also, it was casting the
2431 result into an int without checking.
2433 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2434 did not work. The code looks as if it has *never* worked, though it appears
2435 to have been documented since at least releast 1.62. I have made it work.
2437 35. The "timeout_DNS" error in retry rules, also documented since at least
2438 1.62, also never worked. As it isn't clear exactly what this means, and
2439 clearly it isn't a major issue, I have abolished the feature by treating it
2440 as "timeout", and writing a warning to the main and panic logs.
2442 36. The display of retry rules for -brt wasn't always showing the error code
2445 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2446 timeout_connect_A, timeout_connect_MX.
2448 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2449 to the empty sender.
2451 39. The daemon was not analysing the content of -oX till after it had closed
2452 stderr and disconnected from the controlling terminal. This meant that any
2453 syntax errors were only noted on the panic log, and the return code from
2454 the command was 0. By re-arranging the code a little, I've made the
2455 decoding happen first, so such errors now appear on stderr, and the return
2456 code is 1. However, the actual setting up of the sockets still happens in
2457 the disconnected process, so errors there are still only recorded on the
2460 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2461 connections (as happens on some IP stacks) was logged at start up time as
2462 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2463 it from "IPv6 and IPv4", which means that two separate sockets are being
2466 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2467 says whether AF_INET or AF_INET6 was passed as an argument.
2469 42. Exiwhat output was messed up when time zones were included in log
2472 43. Exiwhat now gives more information about the daemon's listening ports,
2473 and whether -tls-on-connect was used.
2475 44. The "port" option of the smtp transport is now expanded.
2477 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2478 silently ignored. Now an error message is written to the main and panic
2481 46. There's a new ACL modifier called "logwrite" which writes to a log file
2482 as soon as it is encountered.
2484 47. Added $local_user_uid and $local_user_gid at routing time.
2486 48. Exim crashed when trying to verify a sender address that was being
2489 49. Exim was recognizing only a space character after ".include". It now also
2490 recognizes a tab character.
2492 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2493 extracting the relevant information from the specification. The man page no
2494 longer contains scrambled data for the -d option, and I've added a section
2495 at the front about calling Exim under different names.
2497 51. Added "extra_headers" argument to the "mail" command in filter files.
2499 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2502 53. Installed eximstats 1.29.
2504 54. Added transport_filter_timeout as a generic transport option.
2506 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2507 Cc: header lines. This was required by RFC 822, but it not required by RFC
2510 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2511 incoming messages that did not have them. Now it does so only if the
2512 message originates locally, that is, if there is no associated remote host
2513 address. When Resent- header lines are present, this applies to the Resent-
2514 lines rather than the non-Resent- lines.
2516 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2517 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2519 58. Messages for configuration errors now include the name of the main
2520 configuration file - useful now that there may be more than one file in a
2521 list (.included file names were always shown).
2523 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2524 for those rare installations that do not start the daemon as root or run it
2525 setuid root. I've cut out the call to initgroups() if the daemon is not
2528 60. The Exim user and group can now be bound into the binary as text strings
2529 that are looked up at the start of Exim's processing.
2531 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2533 62. Added $mailstore_basename variable.
2535 63. Installed patch to sieve.c from Michael Haardt.
2537 64. When Exim failed to open the panic log after failing to open the main log,
2538 the original message it was trying to log was written to stderr and debug
2539 output, but if they were not available (the usual case in production), it
2540 was lost. Now it is written to syslog before the two lines that record the
2541 failures to open the logs.
2543 65. Users' Exim filters run in subprocesses under the user's uid. It is
2544 possible for a "deliver" command or an alias in a "personal" command to
2545 provoke an address rewrite. If logging of address rewriting is configured,
2546 this fails because the process is not running as root or exim. There may be
2547 a better way of dealing with this, but for the moment (because 4.30 needs
2548 to be released), I have disabled address rewrite logging when running a
2549 filter in a non-root, non-exim process.
2555 1. The buildconfig auxiliary program wasn't quoting the value set for
2556 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2557 not defined. This bug was masked in 4.22 by the effect that was fixed in
2560 2. Some messages that were rejected after a message id was allocated were
2561 shown as "incomplete" by exigrep. It no longer does this for messages that
2562 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2564 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2565 have allow_domain_literals set, the ID did not get logged in the <= line.
2566 Domain literals are now always recognized in Message-ID: header lines.
2568 4. The first argument for a ${extract expansion item is the key name or field
2569 number. Leading and trailing spaces in this item were not being ignored,
2570 causing some misleading effects.
2572 5. When deliver_drop_privilege was set, single queue runner processes started
2573 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2574 same command in the process it spins off) were not dropping privilege.
2576 6. When the daemon running as "exim" started a queue runner, it always
2577 re-executed Exim in the spun-off process. This is a waste of effort when
2578 deliver_drop_privilege is set. The new process now just calls the
2579 queue-runner function directly.
2585 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2588 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2589 ignored. Though the use of -oP was forcing the writing of a pid file, it
2590 was always written to the default place.
2592 3. If the message "no IP address found for host xxxx" is generated during
2593 incoming verification, it is now followed by identification of the incoming
2594 connection (so you can more easily find what provoked it).
2596 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2598 5. Added some features to "harden" Exim a bit more against certain attacks:
2600 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2601 be put in Local/Makefile. This is like the never_users runtime option,
2602 but it cannot be overridden. The default setting is "root".
2604 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2605 prefix string with which any file named in a -C command line option
2608 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2609 is retained for -C and -D only if the caller of Exim is root. Without
2610 it, the exim user may also use -C and -D and retain privilege.
2612 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2613 command line option is disabled.
2615 6. Macro names set by the -D option must start with an upper case letter, just
2616 like macro names defined in the configuration file.
2618 7. Added "dereference=" facility to LDAP.
2620 8. Two instances of the typo "uknown" in the source files are fixed.
2622 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2623 the Configure-Makefile script screwed up while processing it.
2625 10. Incorporated PCRE 4.4.
2627 11. The SMTP synchronization check was not operating right at the start of an
2628 SMTP session. For example, it could not catch a HELO sent before the client
2629 waited for the greeting. There is now a check for outstanding input at the
2630 point when the greeting is written. Because of the duplex, asynchronous
2631 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2632 way, but not yet received, when the check is performed.
2634 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2635 on TCP/IP sockets, because this apparently causes some broken clients to
2638 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2639 unchanged) from the Cygwin maintainer.
2641 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2642 was supported instead of testing for mbx. Effectively a typo.
2644 15. The spa authenticator server code was not checking that the input it
2645 received was valid base64.
2647 16. The debug output line for the "set" modifier in ACLs was not showing the
2648 name of the variable that was being set.
2650 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2652 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2653 lookup on its own. Something else was needed to trigger the lookup. For
2654 example, a match in host_lookup or the need for a host name in a host list.
2655 Now, if $sender_host_name is referenced and the host name has not yet been
2656 looked up, a lookup is performed. If the lookup fails, the variable remains
2657 empty, and $host_lookup_failed is set to "1".
2659 19. Added "eqi" as a case-independent comparison operator.
2661 20. The saslauthd authentication condition could segfault if neither service
2662 nor realm was specified.
2664 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2665 error message that was logged was misleading, and incoming SMTP
2666 connections were dropped. The message is now more accurate, and temporary
2667 errors are given to SMTP connections.
2669 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2670 (except RSET) with a 421 error, until QUIT is received. However, it was
2671 failing to send a response to QUIT.
2673 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2674 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2675 ACL is now run after the helo_try_verify_hosts code.
2677 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2678 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2679 case-independent, so other case variants are also recognized). Apparently
2680 some systems use these upper case variants.
2682 25. If more than two messages were waiting for the same host, and a transport
2683 filter was specified for the transport, Exim sent two messages over the
2684 same TCP/IP connection, and then failed with "socket operation on non-
2685 socket" when it tried to send the third.
2687 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2689 27. The extern definition of crypt16() in expand.c was not being excluded when
2690 the OS had its own crypt16() function.
2692 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2693 as a preferred synonym for return_size_limit, both as an option and as an
2696 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2698 30. Changed the default configuration ACL to relax the local part checking rule
2699 for addresses that are not in any local domains. For these addresses,
2700 slashes and pipe symbols are allowed within local parts, but the sequence
2701 /../ is explicitly forbidden.
2703 31. SPA server authentication was not clearing the challenge buffer before
2706 32. log_message in a "warn" ACL statement was writing to the reject log as
2707 well as to the main log, which contradicts the documentation and doesn't
2708 seem right (because no rejection is happening). So I have stopped it.
2710 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2711 However, I am unable to do any testing of this.
2713 34. Fixed an infelicity in the appendfile transport. When checking directories
2714 for a mailbox, to see if any needed to be created, it was accidentally
2715 using path names with one or more superfluous leading slashes; tracing
2716 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2718 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2719 contents are added to the log line that is written for every discarded
2720 recipient. (Previously a log_message setting was ignored.)
2722 36. The ${quote: operator now quotes the string if it is empty.
2724 37. The install script runs exim in order to find its version number. If for
2725 some reason other than non-existence or emptiness, which it checks, it
2726 could not run './exim', it was installing it with an empty version number,
2727 i.e. as "exim-". This error state is now caught, and the installation is
2730 38. An argument was missing from the function that creates an error message
2731 when Exim fails to connect to the socket for saslauthd authentication.
2732 This could cause Exim to crash, or give a corrupted message.
2734 39. Added isip, isip4, and isip6 to ${if conditions.
2736 40. The ACL variables $acl_xx are now saved with the message, and can be
2737 accessed later in routers, transports, and filters.
2739 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2740 strings in the file are not string-expanded.
2742 42. If a MAIL command specified a SIZE value that was too large to fit into an
2743 int variable, the check against message_size_limit failed. Such values are
2744 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2745 day this will have to be increased, but I don't think I want to be around
2746 when emails are that large.
2753 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2754 iconv() is not standard in FreeBSD.
2756 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2757 IPv6 enabled. The observed symptom was a segmentation fault on return from
2758 the function os_common_find_running_interfaces() in src/os.c.
2760 3. In the check_special_case() function in daemon.c I had used "errno" as an
2761 argument name, which causes warnings on some systems. This was basically a
2762 typo, since it was named "eno" in the comments!
2764 4. The code that waits for the clock to tick (at a resolution of some fraction
2765 of a second) so as to ensure message-id uniqueness was always waiting for
2766 at least one whole tick, when it could have waited for less. [This is
2767 almost certainly not relevant at current processor speeds, where it is
2768 unlikely to ever wait at all. But we try to future-proof.]
2770 5. The function that sleeps for a time interval that includes fractions of a
2771 second contained a race. It did not block SIGALRM between setting the
2772 timer, and suspending (a couple of lines later). If the interval was short
2773 and the sigsuspend() was delayed until after it had expired, the suspension
2774 never ended. On busy systems this could lead to processes getting stuck for
2777 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2778 process, before it forks any delivery processes. The open lookup caching
2779 mechanism meant that the open file or database connection was passed into
2780 the delivery process. The problem was that delivery processes always tidy
2781 up cached lookup data. This could cause a problem for the next delivery
2782 process started by the queue runner, because the external queue runner
2783 process does not know about the closure. So the next delivery process
2784 still has data in the lookup cache. In the case of a file lookup, there was
2785 no problem because closing a file descriptor in a subprocess doesn't affect
2786 the parent. However, if the lookup was caching a connection to a database,
2787 the connection was closed, and the second delivery process was likely to
2788 see errors such as "PGSQL: query failed: server closed the connection
2789 unexpectedly". The problem has been fixed by closing all cached lookups
2790 in a queue runner before running a delivery process.
2792 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2793 seem to have the "const" qualifier which it has on other OS. I've
2796 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2797 *of the same type* (client or server) with the same public name that an
2798 error should be diagnosed.
2800 9. When Exim looked up a host name for an IP address, but failed to find the
2801 original IP address when looking up the host name (a safety check), it
2802 output the message "<ip address> does not match any IP for NULL", which was
2803 confusing, to say the least. The bug was that the host name should have
2804 appeared instead of "NULL".
2806 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2807 user that is built into the binary, and the -C or -D options is used, root
2808 privilege is dropped before the configuration file is read. In addition,
2809 logging is switched to stderr instead of the normal log files. If the
2810 configuration then re-defines the Exim user, the unprivileged environment
2811 is probably not what is expected, so Exim logs a panic warning message (but
2814 However, if deliver_drop_privilege is set, the unprivileged state may well
2815 be exactly what is intended, so the warning has been cut out in that case,
2816 and Exim is allowed to try to write to its normal log files.
2822 1. smtp_return_error_details was not giving details for temporary sender
2823 or receiver verification errors.
2825 2. Diagnose a configuration error if two authenticators have the same public
2828 3. Exim used not to create the message log file for a message until the first
2829 delivery attempt. This could be confusing when incoming messages were held
2830 for policy or load reasons. The message log file is now created at the time
2831 the message is received, and an initial "Received" line is written to it.
2833 4. The automatically generated man page for command line options had a minor
2834 bug that caused no ill effects; however, a more serious problem was that
2835 the procedure for building the man page automatically didn't always
2836 operate. Consequently, release 4.20 contains an out-of-date version. This
2837 shouldn't happen again.
2839 5. When building Exim with embedded Perl support, the script that builds the
2840 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2841 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2843 6. The freeze_tell option was not being used for messages that were frozen on
2844 arrival, either by an ACL or by local_scan().
2846 7. Added the smtp_incomplete_transaction log selector.
2848 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2849 was accepting AUTH without a new EHLO.
2851 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2852 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2855 10. Logging of TCP/IP connections (when configured) now happens in the main
2856 daemon process instead of the child process, so that the TCP/IP connection
2857 count is more accurate (but it can never be perfect).
2859 11. The use of "drop" in a nested ACL was not being handled correctly in the
2860 outer ACL. Now, if condition failure induced by the nested "drop" causes
2861 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2862 or "require"), the connection is dropped.
2864 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2865 that yield "discard" can now be used with an "accept" or a "discard" verb,
2866 but an error is generated for any others (because I can't see a useful way
2867 to define what should happen).
2869 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2870 are now processed in the same way as lines in the Exim configuration file.
2871 In particular, continuation lines are supported.
2873 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2875 15. Added -ti meaning -t -i.
2877 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2878 domains, and warn by logging if others are found.
2880 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2881 returned from GIFCONF when Exim is trying to find the list of interfaces on
2882 a host. The code in os.c has been modified to copy each ifreq to an aligned
2883 structure in all cases.
2885 Also, in some cases, the returned ifreq's were being copied to a 'struct
2886 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2887 means the last couple of bytes of an IPv6 address could be chopped if the
2888 ifreq contained only a normal sockaddr (14 bytes storage).
2890 18. Named domain lists were not supported in the hosts_treat_as_local option.
2891 An entry such as +xxxx was not recognized, and was treated as a literal
2894 19. Ensure that header lines added by a DATA ACL are included in the reject log
2895 if the ACL subsequently rejects the message.
2897 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2898 MD5 (which is deprecated).
2900 21. When testing a filter file using -bf, Exim was writing a message when it
2901 took the sender from a "From " line in the message, but it was not doing so
2902 when it took $return_path from a Return-Path: header line. It now does.
2904 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2905 with a valid header line field name (a series of printing characters
2906 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2908 23. Changed "disc" in the source to "disk" to conform to the documentation and
2909 the book and for uniformity.
2911 24. Ignore Sendmail's -Ooption=value command line item.
2913 25. When execve() failed while trying to run a command in a pipe transport,
2914 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2915 could be confused with a return value of 69 from the command itself. This
2916 has been changed to 127, the value the shell returns if it is asked to run
2917 a non-existent command. The wording for the related log line suggests a
2918 non-existent command as the problem.
2920 26. If received_header_text expands to an empty string, do not add a Received:
2921 header line to the message. (Well, it adds a token one on the spool, but
2922 marks it "old" so that it doesn't get used or transmitted.)
2924 27. Installed eximstats 1.28 (addition of -nt option).
2926 28. There was no check for failure on the call to getsockname() in the daemon
2927 code. This can fail if there is a shortage of resources on the system, with
2928 ENOMEM, for example. A temporary error is now given on failure.
2930 29. Contrary to the C standard, it seems that in some environments, the
2931 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2932 program. Exim now does this explicitly; it affects the formatting of
2933 timestamps using strftime().
2935 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2936 complaints. I've now initialized all the variables, to avoid this.
2938 32. Header lines added by a system filter were not being "seen" during
2939 transport-time rewrites.
2941 33. The info_callback() function passed to OpenSSL is set up with type void
2942 (*)(SSL *, int, int), as described somewhere. However, when calling the
2943 function (actually a macro) that sets it up, the type void(*)() is
2944 expected. I've put in a cast to prevent warnings from picky compilers.
2946 34. If a DNS black list lookup found a CNAME record, but there were no A
2947 records associated with the domain it pointed at, Exim crashed.
2949 35. If a DNS black list lookup returned more than one A record, Exim ignored
2950 all but the first. It now scans all returned addresses if a particular IP
2951 value is being sought. In this situation, the contents of the
2952 $dnslist_value variable are a list of all the addresses, separated by a
2955 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2956 to accept a host name and all its aliases if the forward lookup for any of
2957 them yielded the IP address of the incoming connection. Now it accepts only
2958 those names whose forward lookup yields the correct IP address. Any other
2959 names are discarded. This closes a loophole whereby a rogue DNS
2960 administrator could create reverse DNS records to break through a
2961 wildcarded host restriction in an ACL.
2963 37. If a user filter or a system filter that ran in a subprocess used any of
2964 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2965 the wrong values were passed to the pipe command ($thisaddress had the
2966 value of $0, $0 had the value of $1, etc). This bug was introduced by
2967 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2969 38. Improved the line breaking for long SMTP error messages from ACLs.
2970 Previously, if there was no break point between 40 and 75 characters, Exim
2971 left the rest of the message alone. Two changes have been made: (a) I've
2972 reduced the minimum length to 35 characters; (b) if it can't find a break
2973 point between 35 and 75 characters, it looks ahead and uses the first one
2974 that it finds. This may give the occasional overlong line, but at least the
2975 remaining text gets split now.
2977 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2978 file descriptors might be low, and that setting 1000 would always raise it.
2979 It turns out that in some environments, the limit is already over 1000 and
2980 that lowering it causes trouble. So now Exim takes care not to decrease it.
2982 40. When delivering a message, the value of $return_path is set to $sender_
2983 address at the start of routing (routers may change the value). By an
2984 oversight, this default was not being set up when an address was tested by
2985 -bt or -bv, which affected the outcome if any router or filter referred to
2988 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2989 log only when "message" or "log_message" are set. However, if one of the
2990 conditions was an address verification, or a call to a nested ACL, the
2991 messages generated by the underlying test were being passed through. This
2992 no longer happens. The underlying message is available in $acl_verify_
2993 message for both "message" and "log_message" expansions, so it can be
2994 passed through if needed.
2996 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
2997 new expansion $bh_ to give the encoded byte string without charset
2998 translation. Translation happens only if iconv() is available; HAVE_ICONV
2999 indicates this at build time. HEADERS_CHARSET gives the charset to
3000 translate to; headers_charset can change it in the configuration, and
3001 "headers charset" can change it in an individual filter file.
3003 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3004 that creates RFC 2047 encoded "words" labels them as that charset instead
3005 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3006 expansion operator; (ii) when Exim creates a From: line for a local
3007 message; (iii) when a header line is rewritten to include a "phrase" part.
3009 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3010 buggy, causing it to skip the first lines of messages whose message ID
3011 ended in 'D'. This would not have bitten before Exim release 4.14, because
3012 message IDs were unlikely to end in 'D' before then. The effect was to have
3013 incorrect size information for certain domains.
3015 45. #include "config.h" was missing at the start of the crypt16.c module. This
3016 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3019 46. If there was a timeout during a "random" callout check, Exim treated it as
3020 a failure of the random address, and carried on sending RSET and the real
3021 address. If the delay was just some slowness somewhere, the response to the
3022 original RCPT would be taken as a response to RSET and so on, causing
3023 mayhem of various kinds.
3025 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3026 when I implemented it. It didn't allow for the fact that some option values
3027 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3028 the right test for positive values.
3030 48. Domain names in DNS records are case-independent. Exim always looks them up
3031 in lower case. Some resolvers return domain names in exactly the case they
3032 appear in the zone file, that is, they may contain uppercase letters. Not
3033 all resolvers do this - some return always lower case. Exim was treating a
3034 change of case by a resolver as a change of domain, similar to a widening
3035 of a domain abbreviation. This triggered its re-routing code and so it was
3036 trying to route what was effectively the same domain again. This normally
3037 caused routing to fail (because the router wouldn't handle the domain
3038 twice). Now Exim checks for this case specially, and just changes the
3039 casing of the domain that it ultimately uses when it transmits the message
3042 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3045 50. If a filter generated a file delivery with a non-absolute name (possible if
3046 no home directory exists for the router), the forbid_file option was not
3049 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3050 the existing equality matching.
3052 52. Exim was using ints instead of ino_t variables in some places where it was
3053 dealing with inode numbers.
3055 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3056 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3057 called TMPDIR, and if it finds it is different, it changes its value.
3059 54. The smtp_printf() function is now made available to local_scan() so
3060 additional output lines can be written before returning. There is also an
3061 smtp_fflush() function to enable the detection of a dropped connection.
3062 The variables smtp_input and smtp_batched_input are exported to
3065 55. Changed the default runtime configuration: the message "Unknown user"
3066 has been removed from the ACL, and instead placed on the localuser router,
3067 using the cannot_route_message feature. This means that any verification
3068 failures that generate their own messages won't get overridden. Similarly,
3069 the "Unrouteable address" message that was in the ACL for unverifiable
3070 relay addresses has also been removed.
3072 56. Added hosts_avoid_esmtp to the smtp transport.
3074 57. The exicyclog script was not checking for the esoteric option
3075 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3076 will work only if exicyclog is run under the appropriate euid.
3078 58. Following a discussion on the list, the rules by which Exim recognises line
3079 endings on incoming messages have been changed. The -dropcr and drop_cr
3080 options are now no-ops, retained only for backwards compatibility. The
3081 following line terminators are recognized: LF CRLF CR. However, special
3082 processing applies to CR:
3084 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3085 nor a local message in the state where . is a terminator.
3087 (ii) If a bare CR is encountered in a header line, an extra space is added
3088 after the line terminator so as not to end the header. The reasoning
3089 behind this is that bare CRs in header lines are most likely either
3090 to be mistakes, or people trying to play silly games.
3092 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3093 was being incorrectly given as 18 bytes larger than it should have been.
3094 This is a VOB (very old bug).
3096 60. This may never have affected anything current, but just in case it has:
3097 When the local host is found other than at the start of a list of hosts,
3098 the local host, those with the same MX, and any that follow, are discarded.
3099 When the list in question was part of a longer list of hosts, the following
3100 hosts (not currently being processed) were also being discarded. This no
3101 longer happens. I'm not sure if this situation could ever has previously
3104 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3107 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3108 header. This is something that is recommended in a new Internet Draft, and
3109 is something that is documented as being done by Sendmail. There are two
3110 possible values. For messages generated by the autoreply transport, Exim
3113 Auto-Submitted: auto-replied
3115 whereas for all other generated messages (e.g. bounces) it adds
3117 Auto-Submitted: auto-generated
3119 63. The "personal" condition in filters now includes a test for the
3120 Auto-Submitted: header. If it contains the string "auto-" the message it
3121 not considered personal.
3123 64. Added rcpt_include_affixes as a generic transport option.
3125 65. Added queue_only_override (default true).
3127 66. Added the syslog_duplication option.
3129 67. If what should have been the first header line of a message consisted of
3130 a space followed by a colon, Exim was mis-interpreting it as a header line.
3131 It isn't of course - it is syntactically invalid and should therefore be
3132 treated as the start of the message body. The misbehaviour could have
3133 caused a number of strange effects, including loss of data in subsequent
3134 header lines, and spool format errors.
3136 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3137 client host had authenticated. This control can now be exercised by an ACL
3138 for more flexibility.
3140 69. By default, callouts do not happen when testing with -bh. There is now a
3141 variant, -bhc, which does actually run the callout code, including
3142 consulting and updating the callout cache.
3144 70. Added support for saslauthd authentication, courtesy of Alexander
3147 71. If statvfs() failed on the spool or log directories while checking their
3148 size for availability, Exim confusingly gave the error "space shortage".
3149 Furthermore, in debugging mode it crashed with a floating point exception.
3150 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3151 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3152 serious problem, Exim now writes to the main and panic logs when this
3153 happens, with details of the failure. It then refuses to accept the
3154 incoming message, giving the message "spool directory problem" or "log
3155 directory problem" with a 421 code for SMTP messages.
3157 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3158 0, 1, and 2 exist, because some OS complain for execs without them (see
3159 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3160 descriptors. However, the code omitted to check that the open succeeded,
3161 causing mysterious errors if for some reason the permissions on /dev/null
3162 got screwed. Now Exim writes a message to the main and panic logs, and
3163 bombs out if it can't open /dev/null.
3165 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3166 interact so that it is all more flexible. It is supposed to remain
3167 backwards compatible. Also added extra_local_interfaces.
3169 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3170 to bomb out with an assertion failure - to the client this appears as a
3171 connection drop. This problem occurs in the part of the code that was taken
3172 from the Samba project. Fortunately, the assertion is in a very simple
3173 function, so I have fixed this by reproducing the function inline in the
3174 one place where it is called, and arranging for authentication to fail
3175 instead of killing the process with assert().
3177 75. The SPA client code was not working when the server requested OEM rather
3178 than Unicode encoding.
3180 76. Added code to make require_files with a specific uid setting more usable in
3181 the case where statting the file as root fails - usually a non-root-mounted
3182 NFS file system. When this happens and the failure is EACCES, Exim now
3183 forks a subprocess and does the per-uid checking as the relevant uid.
3185 77. Added process_log_path.
3187 78. If log_file_path was not explicitly set, a setting of check_log_space or
3188 check_log_inodes was ignored.
3190 79. If a space check for the spool or log partitions fails, the incident is now
3191 logged. Of course, in the latter case the data may get lost...
3193 80. Added the %p formatting code to string_format() so that it can be used to
3194 print addresses in debug_print(). Adjusted all the address printing in the
3195 debugging in store.c to use %p rather than %d.
3197 81. There was a concern that a line of code in smtp_in.c could overflow a
3198 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3199 initially expressed, the concern was not well-founded, because trailing
3200 spaces are removed early. However, if the trailing spaces were followed by
3201 a NULL, they did not get removed, so the overflow was possible. Two fixes
3204 (a) I re-wrote the offending code in a cleaner fashion.
3205 (b) If an incoming SMTP command contains a NULL character, it is rejected
3208 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3209 runs initgroups(), so that if the Exim user is in any additional groups,
3210 they will be used during message reception.
3216 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob