1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.199 2005/08/08 09:57:29 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TK/01 Added the "success_on_redirect" address verification option. See
10 NewStuff for rationale and an example.
12 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
14 PH/02 Patch to exigrep to allow it to work on syslog lines.
16 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
17 fread() to skip over the body file's header line, because in Cygwin the
18 header line is locked and is inaccessible.
20 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
21 co-exist for some time) to make it clear that it is the Exim ID that is
22 referenced, not the Message-ID: header line.
24 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
25 string_format() function, because snprintf() does not exist on all
28 PH/06 The use of forbid_filter_existstest now also locks out the use of the
29 ${stat: expansion item.
31 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
32 protocol synchronization error", to keep the pedants happy.
34 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
35 well as for IRIX systems, when gcc is being used. See the host.c source
38 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
40 PH/10 Named domain lists were not working if used in a queue_smtp_domains
43 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
44 transport and to the smtp transport in LMTP mode.
46 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
48 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
49 run a filter in a subprocess. This could lead to confusion in subsequent
50 lookups in the parent process. There should also be a search_tidyup() at
51 the end of the subprocess.
53 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
54 only if the host matched helo_try_verify_hosts, which caused the
55 verification to occur when the EHLO/HELO command was issued. The ACL just
56 tested the remembered result. Now, if a previous verification attempt has
57 not happened, "verify = helo" does it there and then.
59 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
61 TK/03 Fix log output including CR from clamd.
63 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
64 exist provoked a memory error which could cause a segfault.
70 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
72 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
73 command itself ended in error, the underlying message about the transport
74 filter timeout was being overwritten with the pipe command error. Now the
75 underlying error message should be appended to the second error message.
77 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
79 PH/02 Reduce the amount of output that "make" produces by default. Full output
80 can still be requested.
82 PH/03 The warning log line about a condition test deferring for a "warn" verb
83 was being output only once per connection, rather than after each
84 occurrence (because it was using the same function as for successful
85 "warn" verbs). This seems wrong, so I have changed it.
87 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
88 it should not have, which might have caused a crash in the right
89 circumstances, but probably never did.
91 PH/04 Installed a modified version of Tony Finch's patch to make submission
92 mode fix the return path as well as the Sender: header line, and to
93 add a /name= option so that you can make the user's friendly name appear
96 TF/03 Added the control = fakedefer ACL modifier.
98 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
99 Mark Lowes for thorough testing.
101 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
103 TK/03 Merged latest SRS patch from Miles Wilton.
105 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
106 with the definition in sysexits.h (which is #included earlier).
107 Fortunately, Exim does not actually use EX_OK. The code used to try to
108 preserve the sysexits.h value, by assumimg that macro definitions were
109 scanned for macro replacements. I have been disabused of this notion,
110 so now the code just undefines EX_OK before #including unistd.h.
112 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
113 in the smtp transport. When a block could not be written in a single
114 write() function, the timeout was being re-applied to each part-write.
115 This seems wrong - if the receiver was accepting one byte at a time it
116 would take for ever. The timeout is now adjusted when this happens. It
117 doesn't have to be particularly precise.
119 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
120 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
122 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
123 without a domain if the check to <postmaster@domain> fails.
125 SC/01 Eximstats: added -xls and the ability to specify output files
126 (patch written by Frank Heydlauf).
128 SC/02 Eximstats: use FileHandles for outputing results.
130 SC/03 Eximstats: allow any combination of xls, txt, and html output.
132 SC/04 Eximstats: fixed display of large numbers with -nvr option
134 SC/05 Eximstats: fixed merging of reports with empty tables.
136 SC/06 Eximstats: added the -include_original_destination flag
138 SC/07 Eximstats: removed tabs and trailing whitespace.
140 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
142 TK/06 MBOX spool code: Add real "From " MBOX separator line
143 so the .eml file is really in mbox format (even though
144 most programs do not really care). Patch from Alex Miller.
146 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
147 The latter is generated from $received_to and is only set if the
148 message has one envelope recipient. SA can use these headers,
149 obviously out-of-the-box. Patch from Alex Miller.
151 PH/08 The ${def test on a variable was returning false if the variable's
152 value was "0", contrary to what the specification has always said!
153 The result should be true unless the variable is empty.
155 PH/09 The syntax error of a character other than { following "${if
156 def:variable_name" (after optional whitespace) was not being diagnosed.
157 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
158 accidental colon was present, for example, could give incorrect results.
160 PH/10 Tidied the code in a number of places where the st_size field of a stat()
161 result is used (not including appendfile, where other changes are about
164 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
165 This involved changing a lot of size variables from int to off_t. It
166 should work with maildirs and everything.
168 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
169 spamd dying while we are connected to it.
171 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
174 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
175 function for BDB 4.3.
177 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
179 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
180 the sources for PCRE have been re-arranged and more files are now
183 PH/15 The code I had for printing potentially long long variables in PH/11
184 above was not the best (it lost precision). The length of off_t variables
185 is now inspected at build time, and an appropriate printing format (%ld
186 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
187 to be "long long int" or "long int". This is needed for the internal
188 formatting function string_vformat().
190 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
191 the configuration file to be ":syslog", then the script "guesses" where
192 the logs files are, rather than using the compiled in default. In our
193 case the guess is not the same as the compiled default, so the script
194 suddenly stopped working when I started to use syslog. The patch checks
195 to see if log_file_path is "". If so, it attempts to read it from exim
196 with no configuration file to get the compiled in version, before it
197 falls back to the previous guessing code."
199 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
200 implementing BATV in an Exim configuration. See NewStuff for the gory
203 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
204 Makefile that are specific to HP-UX.
206 PH/18 If the "use_postmaster" option was set for a recipient callout together
207 with the "random" option, the postmaster address was used as the MAIL
208 FROM address for the random test, but not for the subsequent recipient
209 test. It is now used for both.
211 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
212 patch removes a few documentation additions to RFC 3028, because the
213 latest draft now contains them. It adds the new en;ascii-case comparator
214 and a new error check for 8bit text in MIME parts. Comparator and
215 require names are now matched exactly. I enabled the subaddress
216 extension, but it is not well tested yet (read: it works for me)."
218 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
219 rework some of the code of TK/09 above to avoid the hardwired use of
220 "%lld" and "long long". Replaced the call to snprintf() with a call to
223 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
224 records point to non-existent hosts", "retry timeout exceeded", and
225 "retry time not reached for any host after a long failure period".
227 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
228 experimental DomainKeys support:
230 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
231 (2) On an error such as an illegally used "control", the wrong name for
232 the control was given.
234 These problems did NOT occur unless DomainKeys support was compiled.
236 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
238 PH/24 Added ${if match_ip condition.
240 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
241 databases so that it will be absolutely obvious if a crash occurs in the
242 DB library. This is a regular occurrence (often caused by mis-matched
245 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
246 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
247 picked up on a user's system that detects such things. There doesn't seem
248 to be a gcc warning option for this - only an attribute that has to be
249 put on the function's prototype. It seems that in Fedora Core 4 they have
250 set this on a number of new functions. No doubt there will be more in due
253 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
254 specify a transport. However, if an address that was verified by such a
255 router was the subject of a callout, Exim crashed because it tried to
256 read the rcpt_include_affixes from the non-existent transport. Now it
257 just assumes that the setting of that option is false. This bug was
258 introduced by 4.51/PH/31.
260 PH/28 Changed -d+all to exclude +memory, because that information is very
261 rarely of interest, but it makes the output a lot bigger. People tend to
262 do -d+all out of habit.
264 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
265 code in os-type was giving problems when libc.so lives in lib64, like on
268 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
269 aren't the modern standard, and it seems that some systems' include files
270 don't always have them. Exim was already checking for some of the newer
271 ones like T_AAAA, and defining it itself. I've added checks for all the
272 record types that Exim uses.
274 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
275 not automatically generating a new one, as it is supposed to. This
276 prevented TLS from working. If the file did exist, but contained invalid
277 data, a new version was generated, as expected. It was only the case of a
278 non-existent file that was broken.
280 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
281 with a change in libdomainkeys > 0.64.
283 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
284 from DNS. If the selector record carries the flag, it now has
285 precedence over the domain-wide flag.
287 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
289 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
290 the use of a port name that isn't defined in /etc/services) occurred, the
291 message was deferred as in a normal delivery, and thus remained on the
292 spool, instead of being failed because of the mua_wrapper setting. This
293 is now fixed, and I tidied up some of the mua_wrapper messages at the
296 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
297 the messages in a hash of arrays rather than using individual hashes.
298 This is a bit cleaner and results in dramatic memory savings, albeit
299 at a slight CPU cost.
301 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
302 as requested by Marc Sherman.
304 SC/10 Eximstats: added histograms for user specified patterns as requested
307 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
309 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
310 fopen() in the content-scanning modules that did not already have it.
316 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
317 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
319 TK/02 Fix ACL "control" statement not being available in MIME ACL.
321 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
323 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
324 to test Sieve filters that use "vacation".
326 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
327 that changes the way the GnuTLS parameters are stored in the cache file.
328 The new format can be generated externally. For backward compatibility,
329 if the data in the cache doesn't make sense, Exim assumes it has read an
330 old-format file, and it generates new data and writes a new file. This
331 means that you can't go back to an older release without removing the
334 PH/03 A redirect router that has both "unseen" and "one_time" set does not
335 work if there are any delivery delays because "one_time" forces the
336 parent to be marked "delivered", so its unseen clone is never tried
337 again. For this reason, Exim now forbids the simultaneous setting of
340 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
341 redirected to themselves ("homonym" addresses). Read the long ChangeLog
342 entry if you want to know the details. The fix, however, neglected to
343 consider the case when local delivery batching is involved. The test for
344 "previously delivered" was not happening when checking to see if an
345 address could be batched with a previous (undelivered) one; under
346 certain circumstances this could lead to multiple deliveries to the same
349 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
350 in its include files, and this causes problems building Exim.
352 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
353 header_syntax) but Exim was just ignoring anything given after a slash.
354 In particular, this caused confusion with an attempt to use "verify =
355 reverse_host_lookup/defer_ok". An error is now given when options are
356 supplied for verify items that do not have them. (Maybe reverse_host_
357 lookup should have a defer_ok option, but that's a different point.)
359 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
360 defined by RFC 821) to 2048, because there were problems with some AUTH
361 commands, and RFC 1869 says the size should be increased for extended
362 SMTP commands that take arguments.
364 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
367 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
368 "unknown" error; now it says that the functionality isn't in the binary.
370 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
371 an address' error message when a string expansion fails (syntax or
372 whatever). Otherwise the password may appear in the log. Following change
373 PH/42 below, there is no longer a chance of it appearing in a bounce
376 PH/11 Installed exipick version 20050225.0 from John Jetmore.
378 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
379 of its addresses was ever tried. (Bugzilla bug #2.)
381 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
382 the result incorrectly in the debug output. (It correctly added a newline
383 to what was transported.)
385 TF/01 Added $received_time.
387 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
388 commented out examples of how to interface to a virus scanner and to
389 SpamAssassin. Also added commented examples of av_scanner and
390 spamd_address settings.
392 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
393 and controls are allowed in which ACLs. There were a couple of minor
394 errors. Some of the entries in the conditions table (which is a table of
395 where they are NOT allowed) were getting very unwieldy; rewrote them as a
396 negation of where the condition IS allowed.
398 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
400 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
401 header file does not have a version number, so I've had to invent a new
402 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
403 API. The code is untested by me (my Linux distribution still has 0.3.2 of
404 radiusclient), but it was contributed by a Radius user.
406 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
407 files or directories, for OpenSSL.
409 PH/19 When an Exim process that is running as root has to create an Exim log
410 file, it does so in a subprocess that runs as exim:exim so as to get the
411 ownership right at creation (otherwise, other Exim processes might see
412 the file with the wrong ownership). There was no test for failure of this
413 fork() call, which would lead to the process getting stuck as it waited
414 for a non-existent subprocess. Forks do occasionally fail when resources
415 run out. I reviewed all the other calls to fork(); they all seem to check
418 PH/20 When checking for unexpected SMTP input at connect time (before writing
419 the banner), Exim was not dealing correctly with a non-positive return
420 from the read() function. If the client had disconnected by this time,
421 the result was a log entry for a synchronization error with an empty
422 string after "input=" when read() returned zero. If read() returned -1
423 (an event I could not check), uninitialized data bytes were printed.
424 There were reports of junk text (parts of files, etc) appearing after
427 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
429 PH/22 Added support for macro redefinition, and (re)definition in between
430 driver and ACL definitions.
432 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
433 forgetting to use the resulting value; it was using the unexpanded value.
435 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
436 hadn't been configured. The fix is from Juergen Kreileder, who
437 understands it better than I do:
439 "Here's what I see happening with three configured cyrus_sasl
440 authenticators configured (plain, login, cram-md5):
442 On startup auth_cyrus_sasl_init() gets called for each of these.
443 This means three calls to sasl_listmech() without a specified mech_list.
444 => SASL tests which mechs of all available mechs actually work
445 => three warnings about OTP not working
446 => the returned list contains: plain, login, cram-md5, digest-md5, ...
448 With the patch, sasl_listmech() also gets called three times. But now
449 SASL's mech_list option is set to the server_mech specified in the the
450 authenticator. Or in other words, the answer from sasl_listmech()
451 gets limited to just the mech you're testing for (which is different
453 => the return list contains just 'plain' or 'login', 'cram-md5' or
454 nothing depending on the value of ob->server_mech.
456 I've just tested the patch: Authentication still works fine,
457 unavailable mechs specified in the exim configuration are still
458 caught, and the auth.log warnings about OTP are gone."
460 PH/25 When debugging is enabled, the contents of the command line are added
461 to the debugging output, even when log_selector=+arguments is not
464 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
465 answer is "GNU", and only if the return is "GNU/something" is the answer
468 PH/27 $acl_verify_message is now set immediately after the failure of a
469 verification in an ACL, and so is available in subsequent modifiers. In
470 particular, the message can be preserved by coding like this:
472 warn !verify = sender
473 set acl_m0 = $acl_verify_message
475 Previously, $acl_verify_message was set only while expanding "message"
476 and "log_message" when a very denied access.
478 PH/28 Modified OS/os.c-Linux with
480 -#ifndef OS_LOAD_AVERAGE
481 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
483 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
484 nomenclature these days.)
486 PH/29 Installed patch from the Sieve maintainer that adds the options
487 sieve_useraddress and sieve_subaddress to the redirect router.
489 PH/30 In these circumstances:
490 . Two addresses routed to the same list of hosts;
491 . First host does not offer TLS;
492 . First host accepts first address;
493 . First host gives temporary error to second address;
494 . Second host offers TLS and a TLS session is established;
495 . Second host accepts second address.
496 Exim incorrectly logged both deliveries with the TLS parameters (cipher
497 and peerdn, if requested) that were in fact used only for the second
500 PH/31 When doing a callout as part of verifying an address, Exim was not paying
501 attention to any local part prefix or suffix that was matched by the
502 router that accepted the address. It now behaves in the same way as it
503 does for delivery: the affixes are removed from the local part unless
504 rcpt_include_affixes is set on the transport.
506 PH/32 Add the sender address, as F=<...>, to the log line when logging a
507 timeout during the DATA phase of an incoming message.
509 PH/33 Sieve envelope tests were broken for match types other than :is. I have
510 applied a patch sanctioned by the Sieve maintainer.
512 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
513 the uid or gid is negative. A case of a negative gid caused this to be
514 noticed. The fix allows for either to be negative.
516 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
517 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
520 PH/36 The change PH/12 above was broken. Fixed it.
522 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
523 the grounds that routing the same address twice would always produce the
524 same answer. This might have been true once, but it is certainly no
525 longer true now. Routing a child address may depend on the previous
526 routing that produced that child. Some complicated redirection strategies
527 went wrong when messages had multiple recipients, and made Exim's
528 behaviour dependent on the order in which the addresses were given.
530 I have moved the duplicate checking until after the routing is complete.
531 Exim scans the addresses that are assigned to local and remote
532 transports, and removes any duplicates. This means that more work will be
533 done, as duplicates will always all be routed, but duplicates are
534 presumably rare, so I don't expect this is of any significance.
536 For deliveries to pipes, files, and autoreplies, the duplicate checking
537 still happens during the routing process, since they are not going to be
540 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
541 It corrects a timeout issue with spamd. This is Ian's comment: "The
542 background is that sometimes spamd either never reads data from a
543 connection it has accepted, or it never writes response data. The exiscan
544 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
545 blindly assumes that writes won't block so it may never time out."
547 PH/39 Allow G after quota size as well as K and M.
549 PH/40 The value set for $authenticated_id in an authenticator may not contain
550 binary zeroes or newlines because the value is written to log lines and
551 to spool files. There was no check on this. Now the value is run through
552 the string_printing() function so that such characters are converted to
553 printable escape sequences.
555 PH/41 $message_linecount is a new variable that contains the total number of
556 lines in the message. Compare $body_linecount, which is the count for the
559 PH/42 Exim no longer gives details of delivery errors for specific addresses in
560 bounce and delay warning messages, except in certain special cases, which
563 (a) An SMTP error message from a remote host;
564 (b) A message specified in a :fail: redirection;
565 (c) A message specified in a "fail" command in a system filter;
566 (d) A message specified in a FAIL return from the queryprogram router;
567 (e) A message specified by the cannot_route_message router option.
569 In these cases only, Exim does include the error details in bounce and
570 warning messages. There are also a few cases where bland messages such
571 as "unrouteable address" or "local delivery error" are given.
573 PH/43 $value is now also set for the "else" part of a ${run expansion.
575 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
576 being worked on, but at least Exim now implements the latest version to
579 PH/45 In a pipe transport, although a timeout while waiting for the pipe
580 process to complete was treated as a delivery failure, a timeout while
581 writing the message to the pipe was logged, but erroneously treated as a
582 successful delivery. Such timeouts include transport filter timeouts. For
583 consistency with the overall process timeout, these timeouts are now
584 treated as errors, giving rise to delivery failures by default. However,
585 there is now a new Boolean option for the pipe transport called
586 timeout_defer, which, if set TRUE, converts the failures into defers for
587 both kinds of timeout. A transport filter timeout is now identified in
590 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
591 systems where "make" and "gmake" are different, calling "gmake" at top
592 level broke things. I've arranged for the value of $(MAKE) to be passed
593 from the Makefile to this script so that it can call the same version of
597 A note about Exim versions 4.44 and 4.50
598 ----------------------------------------
600 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
601 changes of various kinds. As a consequence, a big documentation update was
602 needed. This delayed the release for rather longer than seemed good, especially
603 in the light of a couple of (minor) security issues. Therefore, the changes
604 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
605 release. So 4.44 and 4.50 are in effect two different branches that both start
608 I have left the 4.50 change log unchanged; it contains all the changes since
609 4.43. The change log for 4.44 is below; many of its items are identical to
610 those for 4.50. This seems to be the most sensible way to preserve the
611 historical information.
617 1. Minor wording change to the doc/README.SIEVE file.
619 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
620 computation of the current number of files was incorrect.
622 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
623 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
626 4. Give more explanation in the error message when the command for a transport
627 filter fails to execute.
629 5. There are several places where Exim runs a non-Exim command in a
630 subprocess. The SIGUSR1 signal should be disabled for these processes. This
631 was being done only for the command run by the queryprogram router. It is
632 now done for all such subprocesses. The other cases are: ${run, transport
633 filters, and the commands run by the lmtp and pipe transports.
635 6. Added CONFIGURE_GROUP build-time option.
637 7. Some older OS have a limit of 256 on the maximum number of file
638 descriptors. Exim was using setrlimit() to set 1000 as a large value
639 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
640 systems. I've change it so that if it can't get 1000, it tries for 256.
642 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
643 was an oversight, and furthermore, ever since the addition of extra
644 controls (e.g. 4.43/32), the checks on when to allow different forms of
645 "control" were broken. There should now be diagnostics for all cases when a
646 control that does not make sense is encountered.
648 9. Added the /retain_sender option to "control=submission".
650 10. $recipients is now available in the predata ACL (oversight).
652 11. Tidy the search cache before the fork to do a delivery from a message
653 received from the command line. Otherwise the child will trigger a lookup
654 failure and thereby defer the delivery if it tries to use (for example) a
655 cached ldap connection that the parent has called unbind on.
657 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
658 of $address_data from the recipient verification was clobbered by the
661 13. The value of address_data from a sender verification is now available in
662 $sender_address_data in subsequent conditions in the ACL statement.
664 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
666 15. Added a new option "connect=<time>" to callout options, to set a different
669 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
670 was its contents. (It was OK if the option was not defined at all.)
672 17. A "Completed" log line is now written for messages that are removed from
673 the spool by the -Mrm option.
675 18. New variables $sender_verify_failure and $recipient_verify_failure contain
676 information about exactly what failed.
678 19. Added -dd to debug only the daemon process.
680 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
681 handles timeouts, both on the server side and network timeouts. Renamed the
682 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
684 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
685 It is now set to "smtps".
687 22. $host_address is now set to the target address during the checking of
690 23. When checking ignore_target_hosts for an ipliteral router, no host name was
691 being passed; this would have caused $sender_host_name to have been used if
692 matching the list had actually called for a host name (not very likely,
693 since this list is usually IP addresses). A host name is now passed as
696 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
697 code that specifies a non-restarting handler (typically sigaction() in
698 modern systems) in an attempt to fix a rare and obscure crash bug.
700 25. Narrowed the window for a race in the daemon that could cause it to ignore
701 SIGCHLD signals. This is not a major problem, because they are used only to
702 wake it up if nothing else does.
704 26. A malformed maildirsize file could cause Exim to calculate negative values
705 for the mailbox size or file count. Odd effects could occur as a result.
706 The maildirsize information is now recalculated if the size or filecount
709 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
710 support for a long time. Removed HAVE_SYS_VFS_H.
712 28. Installed the latest version of exipick from John Jetmore.
714 29. In an address list, if the pattern was not a regular expression, an empty
715 subject address (from a bounce message) matched only if the pattern was an
716 empty string. Non-empty patterns were not even tested. This was the wrong
717 because it is perfectly reasonable to use an empty address as part of a
718 database query. An empty address is now tested by patterns that are
719 lookups. However, all the other forms of pattern expect the subject to
720 contain a local part and a domain, and therefore, for them, an empty
721 address still always fails if the pattern is not itself empty.
723 30. Exim went into a mad DNS loop when attempting to do a callout where the
724 host was specified on an smtp transport, and looking it up yielded more
727 31. Re-factored the code for checking spool and log partition space into a
728 function that finds that data and another that does the check. The former
729 is then used to implement four new variables: $spool_space, $log_space,
730 $spool_inodes, and $log_inodes.
732 32. The RFC2047 encoding function was originally intended for short strings
733 such as real names; it was not keeping to the 75-character limit for
734 encoded words that the RFC imposes. It now respects the limit, and
735 generates multiple encoded words if necessary. To be on the safe side, I
736 have increased the buffer size for the ${rfc2047: expansion operator from
739 33. It is now permitted to omit both strings after an "if" condition; if the
740 condition is true, the result is "true". As before, when the second string
741 is omitted, a false condition yields an empty string. This makes it less
742 cumbersome to write custom ACL and router conditions.
744 34. Failure to deliver a bounce message always caused it to be frozen, even if
745 there was an errors_to setting on the router. The errors_to setting is now
748 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
749 canonical form (fully expanded) before being placed in
750 $sender_host_address.
752 36. The table in the code that translates DNS record types into text (T_A to
753 "A" for instance) was missing entries for NS and CNAME. It is just possible
754 that this could have caused confusion if both these types were looked up
755 for the same domain, because the text type is used as part of Exim's
756 per-process caching. But the chance of anyone hitting this buglet seems
759 37. The dnsdb lookup has been extended in a number of ways.
761 (1) There is a new type, "zns", which walks up the domain tree until it
762 finds some nameserver records. It should be used with care.
764 (2) There is a new type, "mxh", which is like "mx" except that it returns
765 just the host names, not the priorities.
767 (3) It is now possible to give a list of domains (or IP addresses) to be
768 looked up. The behaviour when one of the lookups defers can be
769 controlled by a keyword.
771 (4) It is now possible to specify the separator character for use when
772 multiple records are returned.
774 38. The dnslists ACL condition has been extended: it is now possible to supply
775 a list of IP addresses and/or domains to be looked up in a particular DNS
778 39. Added log_selector=+queue_time_overall.
780 40. When running the queue in the test harness, wait just a tad after forking a
781 delivery process, to get repeatability of debugging output.
783 41. Include certificate and key file names in error message when GnuTLS fails
784 to set them up, because the GnuTLS error message doesn't include the name
785 of the failing file when there is a problem reading it.
787 42. Allow both -bf and -bF in the same test run.
789 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
791 44. The "Exiscan patch" is now merged into the mainline Exim source.
793 45. Sometimes the final signoff response after QUIT could fail to get
794 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
795 < 0 before doing a fflush(). This bug looks as though it goes back to the
796 introduction of TLS in release 3.20, but "sometimes" must have been rare
797 because the tests only now provoked it.
799 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
800 (this can affect the format of dates).
802 47. exim_tidydb, when checking for the continued existence of a message for
803 which it has found a message-specific retry record, was not finding
804 messages that were in split spool directories. Consequently, it was
805 deleting retry records that should have stayed in existence.
807 48. Steve fixed some bugs in eximstats.
809 49. The SPA authentication driver was not abandoning authentication and moving
810 on to the next authenticator when an expansion was forced to fail,
811 contradicting the general specification for all authenticators. Instead it
812 was generating a temporary error. It now behaves as specified.
814 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
815 (the order specifies the preference for clients). The order is now AES256,
816 AES128, 3DES, ARCFOUR128.
818 51. Small patch to Sieve code - explicitly set From: when generating an
821 52. Exim crashed if a remote delivery caused a very long error message to be
822 recorded - for instance if somebody sent an entire SpamAssassin report back
823 as a large number of 550 error lines. This bug was coincidentally fixed by
824 increasing the size of one of Exim's internal buffers (big_buffer) that
825 happened as part of the Exiscan merge. However, to be on the safe side, I
826 have made the code more robust (and fixed the comments that describe what
829 53. Now that there can be additional text after "Completed" in log lines (if
830 the queue_time_overall log selector is set), a one-byte patch to exigrep
831 was needed to allow it to recognize "Completed" as not the last thing in
834 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
835 patch that reportedly fixes this has been added. I am not expert enough to
836 create a test for it. This is what the patch creator wrote:
838 "I found a little strange behaviour of ldap code when working with
839 Windows 2003 AD Domain, where users was placed in more than one
840 Organization Units. When I tried to give exim partial DN, the exit code
841 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
842 But simultaneously result of request was absolutely normal ldap result,
843 so I produce this patch..."
845 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
846 REFERENCE, so I have modified the code to exclude the patch when that macro
849 55. Some experimental protocols are using DNS PTR records for new purposes. The
850 keys for these records are domain names, not reversed IP addresses. The
851 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
852 leaves it alone. Component reversal etc. now happens only for IP addresses.
855 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
857 57. Double the size of the debug message buffer (to 2048) so that more of very
858 long debug lines gets shown.
860 58. The exicyclog utility now does better if the number of log files to keep
861 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
863 59. Two changes related to the smtp_active_hostname option:
865 (1) $smtp_active_hostname is now available as a variable.
866 (2) The default for smtp_banner uses $smtp_active_hostname instead
867 of $primary_hostname.
869 60. The host_aton() function is supposed to be passed a string that is known
870 to be a valid IP address. However, in the case of IPv6 addresses, it was
871 not checking this. This is a hostage to fortune. Exim now panics and dies
872 if the condition is not met. A case was found where this could be provoked
873 from a dnsdb PTR lookup with an IPv6 address that had more than 8
874 components; fortuitously, this particular loophole had already been fixed
875 by change 4.50/55 above.
877 If there are any other similar loopholes, the new check in host_aton()
878 itself should stop them being exploited. The report I received stated that
879 data on the command line could provoke the exploit when Exim was running as
880 exim, but did not say which command line option was involved. All I could
881 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
885 61. There was a buffer overflow vulnerability in the SPA authentication code
886 (which came originally from the Samba project). I have added a test to the
887 spa_base64_to_bits() function which I hope fixes it.
890 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
891 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
893 63. The daemon start-up calls getloadavg() while still root for those OS that
894 need the first call to be done as root, but it missed one case: when
895 deliver_queue_load_max is set with deliver_drop_privilege. This is
896 necessary for the benefit of the queue runner, because there is no re-exec
897 when deliver_drop_privilege is set.
899 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
902 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
903 was in use, was not putting the data itself into the right store pool;
904 consequently, it could be overwritten for a subsequent message in the same
905 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
908 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
910 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
911 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
912 the function were treating the return as a boolean value, which happened to
913 work because 0=false and not-0=true, but is not correct code.
915 68. The host_aton() function was not handling scoped IPv6 addresses (those
916 with, for example, "%eth0" on the end) correctly.
918 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
919 negated items (that is, ~something) in unsigned ints. Some compilers
920 apparently mutter when there is no cast.
922 70. If an address verification called from an ACL failed, and did not produce a
923 user-specific message (i.e. there was only a "system" message), nothing was
924 put in $acl_verify_message. In this situation, it now puts the system
927 71. Change 4.23/11 added synchronization checking at the start of an SMTP
928 session; change 4.31/43 added the unwanted input to the log line - except
929 that it did not do this in the start of session case. It now does.
931 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
932 This is wrong and can cause the other end to generate a synchronization
933 error if it is another Exim or anything else that does the synchronization
934 check. A QUIT command is no longer sent after a timeout.
936 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
939 74. The defer_ok option of callout verification was not working if it was used
940 when verifying addresses in header lines, that is, for this case:
942 verify = header_sender/callout=defer_ok
944 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
945 those file descriptors could be used for SMTP connections. If anything
946 wrote to stderr (the example that came up was "warn" in embedded Perl), it
947 could be sent to the SMTP client, causing chaos. The daemon now opens
948 stdin, stdout, and stderr to /dev/null when it puts itself into the
951 76. Arrange for output from Perl's "warn" command to be written to Exim's main
952 log by default. The user can override this with suitable Perl magic.
954 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
955 the log message when discard triggers, was not working for the DATA ACL or
956 for the non-SMTP ACL.
958 78. Error message wording change in sieve.c.
960 79. If smtp_accept_max_per_host was set, the number of connections could be
961 restricted to fewer than expected, because the daemon was trying to set up
962 a new connection before checking whether the processes handling previous
963 connections had finished. The check for completed processes is now done
964 earlier. On busy systems, this bug wouldn't be noticed because something
965 else would have woken the daemon, and it would have reaped the completed
968 80. If a message was submitted locally by a user whose login name contained one
969 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
970 It caused a spool format error. I have fixed the spool reading code. A
971 related problem was that the "from" clause in the Received: line became
972 illegal because of the space(s). It is now covered by ${quote_local_part.
974 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
977 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
979 83. Patch from Sieve maintainer to fix unterminated string problem in
982 84. Some minor changes to the Linux configuration files to help with other
983 OS variants using glibc.
985 85. One more patch for Sieve to update vacation handling to latest spec.
988 ----------------------------------------------------
989 See the note above about the 4.44 and 4.50 releases.
990 ----------------------------------------------------
996 1. Change 4.43/35 introduced a bug that caused file counts to be
997 incorrectly computed when quota_filecount was set in an appendfile
1000 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1001 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1004 3. Give more explanation in the error message when the command for a transport
1005 filter fails to execute.
1007 4. There are several places where Exim runs a non-Exim command in a
1008 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1009 was being done only for the command run by the queryprogram router. It is
1010 now done for all such subprocesses. The other cases are: ${run, transport
1011 filters, and the commands run by the lmtp and pipe transports.
1013 5. Some older OS have a limit of 256 on the maximum number of file
1014 descriptors. Exim was using setrlimit() to set 1000 as a large value
1015 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1016 systems. I've change it so that if it can't get 1000, it tries for 256.
1018 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1019 was an oversight, and furthermore, ever since the addition of extra
1020 controls (e.g. 4.43/32), the checks on when to allow different forms of
1021 "control" were broken. There should now be diagnostics for all cases when a
1022 control that does not make sense is encountered.
1024 7. $recipients is now available in the predata ACL (oversight).
1026 8. Tidy the search cache before the fork to do a delivery from a message
1027 received from the command line. Otherwise the child will trigger a lookup
1028 failure and thereby defer the delivery if it tries to use (for example) a
1029 cached ldap connection that the parent has called unbind on.
1031 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1032 of $address_data from the recipient verification was clobbered by the
1033 sender verification.
1035 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1036 was its contents. (It was OK if the option was not defined at all.)
1038 11. A "Completed" log line is now written for messages that are removed from
1039 the spool by the -Mrm option.
1041 12. $host_address is now set to the target address during the checking of
1042 ignore_target_hosts.
1044 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1045 being passed; this would have caused $sender_host_name to have been used if
1046 matching the list had actually called for a host name (not very likely,
1047 since this list is usually IP addresses). A host name is now passed as
1050 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1051 code that specifies a non-restarting handler (typically sigaction() in
1052 modern systems) in an attempt to fix a rare and obscure crash bug.
1054 15. Narrowed the window for a race in the daemon that could cause it to ignore
1055 SIGCHLD signals. This is not a major problem, because they are used only to
1056 wake it up if nothing else does.
1058 16. A malformed maildirsize file could cause Exim to calculate negative values
1059 for the mailbox size or file count. Odd effects could occur as a result.
1060 The maildirsize information is now recalculated if the size or filecount
1063 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1064 support for a long time. Removed HAVE_SYS_VFS_H.
1066 18. Updated exipick to current release from John Jetmore.
1068 19. Allow an empty sender to be matched against a lookup in an address list.
1069 Previously the only cases considered were a regular expression, or an
1072 20. Exim went into a mad DNS lookup loop when doing a callout where the
1073 host was specified on the transport, if the DNS lookup yielded more than
1076 21. The RFC2047 encoding function was originally intended for short strings
1077 such as real names; it was not keeping to the 75-character limit for
1078 encoded words that the RFC imposes. It now respects the limit, and
1079 generates multiple encoded words if necessary. To be on the safe side, I
1080 have increased the buffer size for the ${rfc2047: expansion operator from
1083 22. Failure to deliver a bounce message always caused it to be frozen, even if
1084 there was an errors_to setting on the router. The errors_to setting is now
1087 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1088 canonical form (fully expanded) before being placed in
1089 $sender_host_address.
1091 24. Updated eximstats to version 1.33
1093 25. Include certificate and key file names in error message when GnuTLS fails
1094 to set them up, because the GnuTLS error message doesn't include the name
1095 of the failing file when there is a problem reading it.
1097 26. Expand error message when OpenSSL has problems setting up cert/key files.
1100 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1101 (this can affect the format of dates).
1103 28. exim_tidydb, when checking for the continued existence of a message for
1104 which it has found a message-specific retry record, was not finding
1105 messages that were in split spool directories. Consequently, it was
1106 deleting retry records that should have stayed in existence.
1108 29. eximstats updated to version 1.35
1109 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1110 1.35 - bugfix such that pie charts by volume are generated correctly
1112 30. The SPA authentication driver was not abandoning authentication and moving
1113 on to the next authenticator when an expansion was forced to fail,
1114 contradicting the general specification for all authenticators. Instead it
1115 was generating a temporary error. It now behaves as specified.
1117 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1118 (the order specifies the preference for clients). The order is now AES256,
1119 AES128, 3DES, ARCFOUR128.
1121 31. Small patch to Sieve code - explicitly set From: when generating an
1124 32. Exim crashed if a remote delivery caused a very long error message to be
1125 recorded - for instance if somebody sent an entire SpamAssassin report back
1126 as a large number of 550 error lines. This bug was coincidentally fixed by
1127 increasing the size of one of Exim's internal buffers (big_buffer) that
1128 happened as part of the Exiscan merge. However, to be on the safe side, I
1129 have made the code more robust (and fixed the comments that describe what
1132 33. Some experimental protocols are using DNS PTR records for new purposes. The
1133 keys for these records are domain names, not reversed IP addresses. The
1134 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1135 leaves it alone. Component reversal etc. now happens only for IP addresses.
1138 34. The host_aton() function is supposed to be passed a string that is known
1139 to be a valid IP address. However, in the case of IPv6 addresses, it was
1140 not checking this. This is a hostage to fortune. Exim now panics and dies
1141 if the condition is not met. A case was found where this could be provoked
1142 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1143 components; fortuitously, this particular loophole had already been fixed
1144 by change 4.50/55 or 4.44/33 above.
1146 If there are any other similar loopholes, the new check in host_aton()
1147 itself should stop them being exploited. The report I received stated that
1148 data on the command line could provoke the exploit when Exim was running as
1149 exim, but did not say which command line option was involved. All I could
1150 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1151 running as the user.
1154 35. There was a buffer overflow vulnerability in the SPA authentication code
1155 (which came originally from the Samba project). I have added a test to the
1156 spa_base64_to_bits() function which I hope fixes it.
1159 36. The daemon start-up calls getloadavg() while still root for those OS that
1160 need the first call to be done as root, but it missed one case: when
1161 deliver_queue_load_max is set with deliver_drop_privilege. This is
1162 necessary for the benefit of the queue runner, because there is no re-exec
1163 when deliver_drop_privilege is set.
1165 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1166 was in use, was not putting the data itself into the right store pool;
1167 consequently, it could be overwritten for a subsequent message in the same
1168 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1171 38. Sometimes the final signoff response after QUIT could fail to get
1172 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1173 < 0 before doing a fflush(). This bug looks as though it goes back to the
1174 introduction of TLS in release 3.20, but "sometimes" must have been rare
1175 because the tests only now provoked it.
1181 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
1182 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
1183 Now it returns an integer. A number of calls were still expecting a T/F
1184 return. Fortuitously, in all cases, the tests worked in OK situations,
1185 which is the norm. However, things would have gone wrong on any write
1186 failures on the smtp file descriptor. This function is used when sending
1187 messages over SMTP and also when doing verify callouts.
1189 2. When Exim is called to do synchronous delivery of a locally submitted
1190 message (the -odf or -odi options), it no longer closes stderr before doing
1193 3. Implemented the mua_wrapper option.
1195 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
1197 5. Implemented the functions header_remove(), header_testname(),
1198 header_add_at_position(), and receive_remove_recipient(), and exported them
1201 6. If an ACL "warn" statement specified the addition of headers, Exim already
1202 inserted X-ACL-Warn: at the start if there was no header name. However, it
1203 was not making this test for the second and subsequent header lines if
1204 there were newlines in the string. This meant that an invalid header could
1205 be inserted if Exim was badly configured.
1207 7. Allow an ACL "warn" statement to add header lines at the start or after all
1208 the Received: headers, as well as at the end.
1210 8. Added the rcpt_4xx retry error code.
1212 9. Added postmaster_mailfrom=xxx to callout verification option.
1214 10. Added mailfrom=xxxx to the callout verification option, for verify=
1217 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
1218 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
1220 12. Inserted some casts to stop certain compilers warning when using pointer
1221 differences as field lengths or precisions in printf-type calls (mostly
1222 affecting debugging statements).
1224 13. Added optional readline() support for -be (dynamically loaded).
1226 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
1227 same clock tick as a message's arrival, so that its received time was the
1228 same as the "first fail" time on the retry record, and that message
1229 remained on the queue past the ultimate address timeout, every queue runner
1230 would try a delivery (because it was past the ultimate address timeout) but
1231 after another failure, the ultimate address timeout, which should have then
1232 bounced the address, did not kick in. This was a "< instead of <=" error;
1233 in most cases the first failure would have been in the next clock tick
1234 after the received time, and all would be well.
1236 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
1237 being recognized when the domain list was tested by the match_domain
1238 condition in an expansion string.
1240 16. Added the ${str2b64: operator.
1242 17. Exim was always calling setrlimit() to set a large limit for the number of
1243 processes, without checking whether the existing limit was already
1244 adequate. (It did check for the limit on file descriptors.) Furthermore,
1245 errors from getrlimit() and setrlimit() were being ignored. Now they are
1246 logged to the main and panic logs, but Exim does carry on, to try to do its
1247 job under whatever limits there are.
1249 18. Imported PCRE 5.0.
1251 19. Trivial typo in log message " temporarily refused connection" (the leading
1254 20. If the log selector return_path_on_delivery was set and an address was
1255 redirected to /dev/null, the delivery process crashed because it assumed
1256 that a return path would always be set for a "successful" delivery. In this
1257 case, the whole delivery is bypassed as an optimization, and therefore no
1260 21. Internal re-arrangement: the function for sending a challenge and reading
1261 a response while authentication was assuming a zero-terminated challenge
1262 string. It's now changed to take a pointer and a length, to allow for
1263 binary data in such strings.
1265 22. Added the cyrus_sasl authenticator (code supplied by MBM).
1267 23. Exim was not respecting finduser_retries when seeking the login of the
1268 uid under which it was called; it was always trying 10 times. (The default
1269 setting of finduser_retries is zero.) Also, it was sleeping after the final
1270 failure, which is pointless.
1272 24. Implemented tls_on_connect_ports.
1274 25. Implemented acl_smtp_predata.
1276 26. If the domain in control=submission is set empty, Exim assumes that the
1277 authenticated id is a complete email address when it generates From: or
1278 Sender: header lines.
1280 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
1281 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
1282 chown and chgrp in /bin and hostname in /usr/bin.
1284 28. Exim was keeping the "process log" file open after each use, just as it
1285 does for the main log. This opens the possibility of it remaining open for
1286 long periods when the USR1 signal hits a daemon. Occasional processlog
1287 errors were reported, that could have been caused by this. Anyway, it seems
1288 much more sensible not to leave this file open at all, so that is what now
1291 29. The long-running daemon process does not normally write to the log once it
1292 has entered its main loop, and it closes the log before doing so. This is
1293 so that log files can straightforwardly be renamed and moved. However,
1294 there are a couple of unusual error situations where the daemon does write
1295 log entries, and I had neglected to close the log afterwards.
1297 30. The text of an SMTP error response that was received during a remote
1298 delivery was being truncated at 512 bytes. This is too short for some of
1299 the long messages that one sometimes sees. I've increased the limit to
1302 31. It is now possible to make retry rules that apply only when a message has a
1303 specific sender, in particular, an empty sender.
1305 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
1306 it possible to be selective about when SMTP synchronization is enforced.
1308 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
1310 32. Implemented hosts_connection_nolog.
1312 33. Added an ACL for QUIT.
1314 34. Setting "delay_warning=" to disable warnings was not working; it gave a
1317 35. Added mailbox_size and mailbox_filecount to appendfile.
1319 36. Added control = no_multiline_responses to ACLs.
1321 37. There was a bug in the logic of the code that waits for the clock to tick
1322 in the case where the clock went backwards by a substantial amount such
1323 that the microsecond fraction of "now" was more than the microsecond
1324 fraction of "then" (but the whole seconds number was less).
1326 38. Added support for the libradius Radius client library this is found on
1327 FreeBSD (previously only the radiusclient library was supported).
1333 1. When certain lookups returned multiple values in the form name=value, the
1334 quoting of the values was not always being done properly. Specifically:
1335 (a) If the value started with a double quote, but contained no whitespace,
1337 (b) If the value contained whitespace other than a space character (i.e.
1338 tabs or newlines or carriage returns) it was not quoted.
1339 This fix has been applied to the mysql and pgsql lookups by writing a
1340 separate quoting function and calling it from the lookup code. The fix
1341 should probably also be applied to nisplus, ibase and oracle lookups, but
1342 since I cannot test any of those, I have not disturbed their existing code.
1344 2. A hit in the callout cache for a specific address caused a log line with no
1345 reason for rejecting RCPT. Now it says "Previous (cached) callout
1346 verification failure".
1348 3. There was an off-by-one bug in the queryprogram router. An over-long
1349 return line was truncated at 256 instead of 255 characters, thereby
1350 overflowing its buffer with the terminating zero. As well as fixing this, I
1351 have increased the buffer size to 1024 (and made a note to document this).
1353 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
1354 when Exim is waiting for an SMTP response from a remote server, Exim
1355 restarts its select() call on the socket, thereby resetting its timeout.
1356 This is not a problem when such interrupts are rare. Somebody set up a cron
1357 job to run exiwhat every 2 minutes, which is less than the normal select()
1358 timeout (5 or 10 minutes). This meant that the select() timeout never
1359 kicked in because it was always reset. I have fixed this by comparing the
1360 time when an interrupt arrives with the time at the start of the first call
1361 to select(). If more time than the timeout has elapsed, the interrupt is
1362 treated as a timeout.
1364 5. Some internal re-factoring in preparation for the addition of Sieve
1365 extensions (by MH). In particular, the "personal" test is moved to a
1366 separate function, and given an option for scanning Cc: and Bcc: (which is
1367 not set for Exim filters).
1369 6. When Exim created an email address using the login of the caller as the
1370 local part (e.g. when creating a From: or Sender: header line), it was not
1371 quoting the local part when it contained special characters such as @.
1373 7. Installed new OpenBSD configuration files.
1375 8. Reworded some messages for syntax errors in "and" and "or" conditions to
1376 try to make them clearer.
1378 9. Callout options, other than the timeout value, were being ignored when
1379 verifying sender addresses in header lines. For example, when using
1381 verify = header_sender/callout=no_cache
1383 the cache was (incorrectly) being used.
1385 10. Added a missing instance of ${EXE} to the exim_install script; this affects
1386 only the Cygwin environment.
1388 11. When return_path_on_delivery was set as a log selector, if different remote
1389 addresses in the same message used different return paths and parallel
1390 remote delivery occurred, the wrong values would sometimes be logged.
1391 (Whenever a remote delivery process finished, the return path value from
1392 the most recently started remote delivery process was logged.)
1394 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
1395 lines when AUTH and/or TLS are in use. This is the "received protocol"
1396 field. Exim used to use "asmtp" for authenticated SMTP, without any
1397 indication (in the protocol name) for TLS use. Now it follows the RFC and
1398 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
1399 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
1400 names appear in log lines as well as in Received: header lines.
1402 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
1403 extensions, and comparison tests, and to fix some bugs.
1405 14. Changes to the "personal" filter test:
1407 (1) The test was buggy in that it was just doing the equivalent of
1408 "contains" tests on header lines. For example, if a user's address was
1409 anne@some.where, the "personal" test would incorrectly be true for
1411 To: susanne@some.where
1413 This test is now done by extracting each address from the header in turn,
1414 and checking the entire address. Other tests that are part of "personal"
1415 are now done using regular expressions (for example, to check local parts
1416 of addresses in From: header lines).
1418 (2) The list of non-personal local parts in From: addresses has been
1419 extended to include "listserv", "majordomo", "*-request", and "owner-*",
1420 taken from the Sieve specification recommendations.
1422 (3) If the message contains any header line starting with "List-" it is
1423 treated as non-personal.
1425 (4) The test for "circular" in the Subject: header line has been removed
1426 because it now seems ill-conceived.
1428 15. Minor typos in src/EDITME comments corrected.
1430 16. Installed latest exipick from John Jetmore.
1432 17. If headers_add on a router specified a text string that was too long for
1433 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
1434 of string_sprintf() is now avoided.
1436 18. $message_body_size was not set (it was always zero) when running the DATA
1437 ACL and the local_scan() function.
1439 19. For the "mail" command in an Exim filter, no default was being set for
1440 the once_repeat time, causing a random time value to be used if "once" was
1441 specified. (If the value happened to be <= 0, no repeat happened.) The
1442 default is now 0s, meaning "never repeat". The "vacation" command was OK
1443 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
1444 (I found it when inspecting the code).
1446 20. There is now an overall timeout for performing a callout verification. It
1447 defaults to 4 times the callout timeout, which applies to individual SMTP
1448 commands during the callout. The overall timeout applies when there is more
1449 than one host that can be tried. The timeout is checked before trying the
1450 next host. This prevents very long delays if there are a large number of
1451 hosts and all are timing out (e.g. when the network connections are timing
1452 out). The value of the overall timeout can be changed by specifying an
1453 additional sub-option for "callout", called "maxwait". For example:
1455 verify = sender/callout=5s,maxwait=20s
1457 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
1458 to the end before writing, but this should make it even safer).
1460 22. Exim was forgetting that it had advertised PIPELINING for the second and
1461 subsequent messages on an SMTP connection. It was also not resetting its
1462 memory on STARTTLS and an internal HELO.
1464 23. When Exim logs an SMTP synchronization error within a session, it now
1465 records whether PIPELINING has been advertised or not.
1467 24. Added 3 instances of "(long int)" casts to time_t variables that were being
1468 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
1469 rather than long int.
1471 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
1473 26. Added the never_mail option to autoreply.
1479 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
1480 crash if the getsockname() call failed; this can happen if a connection is
1481 closed very soon after it is established. The problem was simply in the
1482 order in which certain operations were done, causing Exim to try to write
1483 to the SMTP stream before it had set up the file descriptor. The bug has
1484 been fixed by making things happen in the correct order.
1490 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
1491 before the connection was closed, thus losing the rejection response.
1493 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
1494 some early Solaris releases, but causes trouble in current releases where
1495 socklen_t is defined.
1497 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
1500 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
1503 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
1504 that releasing the top of it at the end releases what was used for sub-
1505 expansions (unless the block got too big). However, discard this block if
1506 the first thing is a variable or header, so that we can use its block when
1507 it is dynamic (useful for very large $message_headers, for example).
1509 6. Lookups now cache *every* query, not just the most recent. A new, separate
1510 store pool is used for this. It can be recovered when all lookup caches are
1511 flushed. Lookups now release memory at the end of their result strings.
1512 This has involved some general refactoring of the lookup sources.
1514 7. Some code has been added to the store_xxx() functions to reduce the amount
1515 of flapping under certain conditions.
1517 8. log_incoming_interface used to affect only the <= reception log lines. Now
1518 it causes the local interface and port to be added to several more SMTP log
1519 lines, for example "SMTP connection from", and rejection lines.
1521 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
1523 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
1525 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
1526 could be overwritten at the end of the current message (or the start of a
1527 new message if it was set in a HELO ACL). The value is now preserved for
1528 the duration of the SMTP connection.
1530 12. If a transport had a headers_rewrite setting, and a matching header line
1531 contained an unqualified address, that address was qualified, even if it
1532 did not match any rewriting rules. The underlying bug was that the values
1533 of the flags that permit the existence of unqualified sender and recipient
1534 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
1535 non-local messages, and by -bnq for local messages) were not being
1536 preserved with the message after it was received.
1538 13. When Exim was logging an SMTP synchronization error, it could sometimes log
1539 "next input=" as part of the text comprising the host identity instead of
1540 the correct text. The code was using the same buffer for two different
1541 strings. However, depending on which order the printing function evaluated
1542 its arguments, the bug did not always show up. Under Linux, for example, my
1543 test suite worked just fine.
1545 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
1546 doesn't work with some older versions of Perl. It has been changed to "my",
1547 which in any case is probably the better facility to use.
1549 15. A really picky compiler found some instances of statements for creating
1550 error messages that either had too many or two few arguments for the format
1553 16. The size of the buffer for calls to the DNS resolver has been increased
1554 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
1555 for addresses that have a lot of PTR records. This alleviates a problem; it
1556 does not fully solve it.
1558 17. A dnsdb lookup for PTR records that receives more data than will fit in the
1559 buffer now truncates the list and logs the incident, which is the same
1560 action as happens when Exim is looking up a host name and its aliases.
1561 Previously in this situation something unpredictable would happen;
1562 sometimes it was "internal error: store_reset failed".
1564 18. If a server dropped the connection unexpectedly when an Exim client was
1565 using GnuTLS and trying to read a response, the client delivery process
1566 crashed while trying to generate an error log message.
1568 19. If a "warn" verb in an ACL added multiple headers to a message in a single
1569 string, for example:
1571 warn message = H1: something\nH2: something
1573 the text was added as a single header line from Exim's point of view
1574 though it ended up OK in the delivered message. However, searching for the
1575 second and subsequent header lines using $h_h2: did not work. This has been
1576 fixed. Similarly, if a system filter added multiple headers in this way,
1577 the routers could not see them.
1579 20. Expanded the error message when iplsearch is called with an invalid key to
1580 suggest using net-iplsearch in a host list.
1582 21. When running tests using -bh, any delays imposed by "delay" modifiers in
1583 ACLs are no longer actually imposed (and a message to that effect is
1586 22. If a "gecos" field in a passwd entry contained escaped characters, in
1587 particular, if it contained a \" sequence, Exim got it wrong when building
1588 a From: or a Sender: header from that name. A second bug also caused
1589 incorrect handling when an unquoted " was present following a character
1590 that needed quoting.
1592 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
1593 was not being matched caselessly.
1595 24. Arranged for all hyphens in the exim.8 source to be escaped with
1598 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
1599 properly. Recipient callout cache records were still being keyed to include
1600 the sender, even when use_sender was set false. This led to far more
1601 callouts that were necessary. The sender is no longer included in the key
1602 when use_sender is false.
1604 26. Added "control = submission" modifier to ACLs.
1606 27. Added the ${base62d: operator to decode base 62 numbers.
1608 28. dnsdb lookups can now access SRV records.
1610 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
1611 the configuration file.
1613 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
1614 (-v) mode. This makes the output for a verbose queue run more intelligible.
1616 31. Added a use_postmaster feature to recipient callouts.
1618 32. Added the $body_zerocount variable, containing the number of binary zero
1619 bytes in the message body.
1621 33. The time of last modification of the "new" subdirectory is now used as the
1622 "mailbox time last read" when there is a quota error for a maildir
1625 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
1627 35. Added +ignore_unknown as a special item in host lists.
1629 36. Code for decoding IPv6 addresses in host lists is now included, even if
1630 IPv6 support is not being compiled. This fixes a bug in which an IPv6
1631 address was recognized as an IP address, but was then not correctly decoded
1632 into binary, causing unexpected and incorrect effects when compared with
1639 1. Very minor rewording of debugging text in manualroute to say "list of
1640 hosts" instead of "hostlist".
1642 2. If verify=header_syntax was set, and a header line with an unqualified
1643 address (no domain) and a large number of spaces between the end of the
1644 name and the colon was received, the reception process suffered a buffer
1645 overflow, and (when I tested it) crashed. This was caused by some obsolete
1646 code that should have been removed. The fix is to remove it!
1648 3. When running in the test harness, delay a bit after writing a bounce
1649 message to get a bit more predictability in the log output.
1651 4. Added a call to search_tidyup() just before forking a reception process. In
1652 theory, someone could use a lookup in the expansion of smtp_accept_max_
1653 per_host which, without the tidyup, could leave open a database connection.
1655 5. Added the variables $recipient_data and $sender_data which get set from a
1656 lookup success in an ACL "recipients" or "senders" condition, or a router
1657 "senders" option, similar to $domain_data and $local_part_data.
1659 6. Moved the writing of debug_print from before to after the "senders" test
1662 7. Change 4.31/66 (moving the time when the Received: is generated) caused
1663 problems for message scanning, either using a data ACL, or using
1664 local_scan() because the Received: header was not generated till after they
1665 were called (in order to set the time as the time of reception completion).
1666 I have revised the way this works. The header is now generated after the
1667 body is received, but before the ACL or local_scan() are called. After they
1668 are run, the timestamp in the header is updated.
1674 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
1675 before starting a queue runner without re-exec. This happened only when
1676 deliver_drop_privilege was set or when the Exim user was set to root. The
1677 effect of the bug was that timeouts during subsequent deliveries caused
1678 crashes instead of being properly handled. The handler is now left at its
1679 default (and expected) setting.
1681 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
1682 message, again when deliver_drop_privilege is set or Exim is run as root.
1683 The bug described in (1) was not present in this case, but the tidying up
1684 of the other signals was missing. I have made the two cases consistent.
1686 3. The ignore_target_hosts setting on a manualroute router was being ignored
1687 for hosts that were looked up using the /MX notation.
1689 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
1692 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
1693 operated on the sender address. After changing the $sender_address to <>
1694 for the sender address verify, Exim was re-instated it as the original
1695 (before rewriting) address, but remembering that it had rewritten it, so it
1696 wasn't rewriting it again. This bug also had the effect of breaking the
1697 sender address verification caching when the sender address was rewritten.
1699 6. The ignore_target_hosts option was being ignored by the ipliteral router.
1700 This has been changed so that if the ip literal address matches
1701 ignore_target_hosts, the router declines.
1703 7. Added expansion conditions match_domain, match_address, and match_local_
1704 part (NOT match_host).
1706 8. The placeholder for the Received: header didn't have a length field set.
1708 9. Added code to Exim itself and to exim_lock to test for a specific race
1709 condition that could lead to file corruption when using MBX delivery. The
1710 issue is with the lockfile that is created in /tmp. If this file is removed
1711 after a process has opened it but before that process has acquired a lock,
1712 there is the potential for a second process to recreate the file and also
1713 acquire a lock. This could lead to two Exim processes writing to the file
1714 at the same time. The added code performs the same test as UW imapd; it
1715 checks after acquiring the lock that its file descriptor still refers to
1716 the same named file.
1718 10. The buffer for building added header lines was of fixed size, 8192 bytes.
1719 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
1722 11. Added the smtp_active_hostname option. If used, this will typically be made
1723 to depend on the incoming interface address. Because $interface_address is
1724 not set up until the daemon has forked a reception process, error responses
1725 that can happen earlier (such as "too many connections") no longer contain
1728 12. If an expansion in a condition on a "warn" statement fails because a lookup
1729 defers, the "warn" statement is abandoned, and the next ACL statement is
1730 processed. Previously this caused the whole ACL to be aborted.
1732 13. Added the iplsearch lookup type.
1734 14. Added ident_timeout as a log selector.
1736 15. Added tls_certificate_verified as a log selector.
1738 16. Added a global option tls_require_ciphers (compare the smtp transport
1739 option of the same name). This controls incoming TLS connections.
1741 17. I finally figured out how to make tls_require_ciphers do a similar thing
1742 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
1743 before starting the TLS session.
1745 18. Tabs are now shown as \t in -bP output.
1747 19. If the log selector return_path_on_delivery was set, Exim crashed when
1748 bouncing a message because it had too many Received: header lines.
1750 20. If two routers both had headers_remove settings, and the first one included
1751 a superfluous trailing colon, the final name in the first list and the
1752 first name in the second list were incorrectly joined into one item (with a
1753 colon in the middle).
1759 1. Added -C and -D options to the exinext utility, mainly to make it easier
1760 to include in the automated testing, but these could be helpful when
1761 multiple configurations are in use.
1763 2. The exinext utility was not formatting the output nicely when there was
1764 an alternate port involved in the retry record key, nor when there was a
1765 message id as well (for retries that were specific to a specific message
1766 and a specific host). It was also confused by IPv6 addresses, because of
1767 the additional colons they contain. I have fixed the IPv4 problem, and
1768 patched it up to do a reasonable job for IPv6.
1770 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
1771 delivery, the log line now contains "pipelined" if PIPELINING was used.
1773 4. An SMTP transport process used to panic and die if the bind() call to set
1774 an explicit outgoing interface failed. This has been changed; it is now
1775 treated in the same way as a connect() failure.
1777 5. A reference to $sender_host_name in the part of a conditional expansion
1778 that was being skipped was still causing a DNS lookup. This no longer
1781 6. The def: expansion condition was not recognizing references to header lines
1782 that used bh_ and bheader_.
1784 7. Added the _cache feature to named lists.
1786 8. The code for checking quota_filecount in the appendfile transport was
1787 allowing one more file than it should have been.
1789 9. For compatibility with Sendmail, the command line option
1797 and sets the incoming protocol and host name (for trusted callers). The
1798 host name and its colon can be omitted when only the protocol is to be set.
1799 Note the Exim already has two private options, -pd and -ps, that refer to
1800 embedded Perl. It is therefore impossible to set a protocol value of "d" or
1801 "s", but I don't think that's a major issue.
1803 10. A number of refactoring changes to the code, none of which should affect
1806 (a) The number of logging options was getting close to filling up the
1807 32-bit word that was used as a bit map. I have split them into two classes:
1808 those that are passed in the argument to log_write(), and those that are
1809 only ever tested independently outside of that function. These are now in
1810 separate 32-bit words, so there is plenty of room for expansion again.
1811 There is no change in the user interface or the logging behaviour.
1813 (b) When building, for example, log lines, the code previously used a
1814 macro that called string_cat() twice, in order to add two strings. This is
1815 not really sufficiently general. Furthermore, there was one instance where
1816 it was actually wrong because one of the argument was used twice, and in
1817 one call a function was used. (As it happened, calling the function twice
1818 did not affect the overall behaviour.) The macro has been replaced by a
1819 function that can join an arbitrary number of extra strings onto a growing
1822 (c) The code for expansion conditions now uses a table and a binary chop
1823 instead of a serial search (which was left over from when there were very
1824 few conditions). Also, it now recognizes conditions like "pam" even when
1825 the relevant support is not compiled in: a suitably worded error message is
1826 given if an attempt is made to use such a condition.
1828 11. Added ${time_interval:xxxxx}.
1830 12. A bug was causing one of the ddress fields not to be passed back correctly
1831 from remote delivery subprocesses. The field in question was not being
1832 subsequently used, so this caused to problems in practice.
1834 13. Added new log selectors queue_time and deliver_time.
1836 14. Might have fixed a bug in maildirsizefile handling that threw up
1837 "unexpected character" debug warnings, and recalculated the data
1838 unnecessarily. In any case, I expanded the warning message to give more
1841 15. Added the message "Restricted characters in address" to the statements in
1842 the default ACL that block characters like @ and % in local parts.
1844 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
1845 Three changes have been made:
1847 (a) There was a serious bug; a negative response to MAIL caused the whole
1848 recipient domain to be cached as invalid, thereby blocking all messages
1849 to all local parts at the same domain, from all senders. This bug has
1850 been fixed. The domain is no longer cached after a negative response to
1851 MAIL if the sender used is not empty.
1853 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
1856 (c) A new callout option, "use_sender" has been added for people who want
1857 the modified behaviour.
1863 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
1866 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
1867 indeed breaks things for older releases.
1869 3. Added additional logging to the case where there is a problem reading data
1870 from a filter that is running in a subprocess using a pipe, in order to
1871 try to track down a specific problem.
1873 4. Testing facility fudge: when running in the test harness and attempting
1874 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
1875 getting "No route to host". Convert this to a timeout.
1877 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
1880 6. Some OS don't have socklen_t but use size_t instead. This affects the
1881 fifth argument of getsockopt() amongst other things. This is now
1882 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
1883 can be set for individual OS. I have set it for SunOS5, OSF1, and
1884 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
1885 some earlier ones do not.
1887 7. Change 4.30/15 was not doing the test caselessly.
1889 8. The standard form for an IPv6 address literal was being rejected by address
1890 parsing in, for example, MAIL and RCPT commands. An example of this kind of
1891 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
1892 this, as well as the form without the "IPv6" on the front (but only when
1893 address literals are enabled, of course).
1895 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
1897 10. Exim crashed if a message with an empty sender address specified by -f
1898 encountered a router with an errors_to setting. This could be provoked only
1899 by a command such as
1903 where an empty string was supplied; "<>" did not hit this bug.
1905 11. Installed PCRE release 4.5.
1907 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
1908 remained set. It is now erased.
1910 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
1911 times from message ids (which are base 36 rather than the normal 62).
1913 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
1914 were being counted as actual protocol errors, and logged if the log
1915 selector +smtp_protocol_error was set. One cannot be perfect in this test,
1916 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
1917 and DATA following a set of rejected RCPTs do not count as protocol errors.
1918 In other words, Exim assumes they were pipelined, though this may not
1919 actually be the case. Of course, in all cases the client gets an
1920 appropriate error code.
1922 15. If a lookup fails in an ACL condition, a message about the failure may
1923 be available; it is used if testing the ACL cannot continue, because most
1924 such messages specify what the cause of the deferral is. However, some
1925 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
1926 that caused an old message to be retained and used if a later statement
1927 caused a defer, replacing the real cause of the deferral.
1929 16. If an IP address had so many PTR records that the DNS lookup buffer
1930 was not large enough to hold them, Exim could crash while trying to process
1931 the truncated data. It now detects and logs this case.
1933 17. Further to 4.21/58, another change has been made: if (and only if) the
1934 first line of a message (the first header line) ends with CRLF, a bare LF
1935 in a subsequent header line has a space inserted after it, so as not to
1936 terminate the header.
1938 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
1939 unnecessarily, used atoi() instead of strtol(), and didn't check the
1940 termination when getting file sizes from file names by regex.
1942 19. Completely re-implemented the support for maildirsize files, in the light
1943 of a number of problems with the previous contributed implementation
1944 (4.30/29). In particular:
1946 . If the quota is zero, the maildirsize file is maintained, but no quota is
1949 . If the maildir directory does not exist, it is created before any attempt
1950 to write a maildirsize file.
1952 . The quota value in the file is just a cache; if the quota is changed in
1953 the transport, the new value overrides.
1955 . A regular expression is available for excluding directories from the
1958 20. The autoreply transport checks the characters in options that define the
1959 message's headers; it allows continued headers, but it was checking with
1960 isspace() after an embedded newline instead of explicitly looking for a
1963 21. If all the "regular" hosts to which an address was routed had passed their
1964 expiry times, and had not reached their retry times, the address was
1965 bounced, even if fallback hosts were defined. Now Exim should go on to try
1968 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
1969 equivalent code in the SMTP transport. Some hosts send humungous responses
1970 to HELO/EHLO, more than 1024 it seems.
1972 23. Refactoring: code in filter.c used (void *) for "any old type" but this
1973 gives compiler warnings in some environments. I've now done it "properly",
1976 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
1977 (because of problems with the built-in one) was declared to return uschar *
1978 instead of char *, causing compiler failure.
1980 25. Fixed a file descriptor leak when processing alias/forward files.
1982 26. Fixed a minor format string issue in dbfn.c.
1984 27. Typo in exim.c: ("dmbnz" for "dbmnz").
1986 28. If a filter file refered to $h_xxx or $message_headers, and the headers
1987 contained RFC 2047 "words", Exim's memory could, under certain conditions,
1990 29. When a sender address is verified, it is cached, to save repeating the test
1991 when there is more than one recipient in a message. However, when the
1992 verification involves a callout, it is possible for different callout
1993 options to be set for different recipients. It is too complicated to keep
1994 track of this in the cache, so now Exim always runs a verification when a
1995 callout is required, relying on the callout cache for the optimization.
1996 The overhead is duplication of the address routing, but this should not be
1999 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2000 to be verified with callout=postmaster, and the main callout worked but the
2001 postmaster check failed, the verification correctly failed. However, if a
2002 subsequent RCPT command asked for sender verification *without* the
2003 postmaster check, incorrect caching caused this verification also to fail,
2006 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2007 it was not caching the DNS options (qualify_single, search_parents) that
2008 were used when the lookup failed. A subsequent lookup with different
2009 options therefore always gave the same answer, though there were cases
2010 where it should not have. (Example: a "domains = !$mx_any" option on a
2011 dnslookup router: the "domains" option is always processed without any
2012 widening, but the router might have qualify_single set.) Now Exim uses the
2013 cached value only when the same options are set.
2015 32. Added John Jetmore's "exipick" utility to the distribution.
2017 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2018 than a timeout (e.g. a certificate is required, and is not provided), an
2019 Exim server now closes the connection immediately. Previously it waited for
2020 the client to close - but if the client is SSL, it seems that they each
2021 wait for each other, leading to a delay before one of them times out.
2023 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2024 maintained 0.8.x compatibility because I don't think many are using it, and
2025 it is clearly obsolete.
2027 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2030 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2031 client certificate was expired. A simple patch fixes this, though I don't
2032 understand the full logic of why the verify callback is called multiple
2035 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2036 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2037 which causes problems with some clients (such as the Certicom SSL Plus
2038 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2039 disables the coutermeasure allowing Eudora to connect."
2041 38. Exim was not checking that a write() to a log file succeeded. This could
2042 lead to Bad Things if a log got too big, in particular if it hit a file
2043 size limit. Exim now panics and dies if it cannot write to a log file, just
2044 as it does if it cannot open a log file.
2046 39. Modified OS/Makefile-Linux so that it now contains
2048 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2050 The two -D definitions ensure that Exim is compiled with large file
2051 support, which makes it possible to handle log files that are bigger than
2054 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2055 instance) a domain was checked against a named list that involved a lookup,
2056 causing $domain_data to be set, then another domain was checked against the
2057 same list, then the first domain was re-checked, the value of $domain_data
2058 after the final check could be wrong. In particular, if the second check
2059 failed, it could be set empty. This bug probably also applied to
2062 41. The strip_trailing_dot option was not being applied to the address given
2063 with the -f command-line option.
2065 42. The code for reading a message's header from the spool was incrementing
2066 $received_count, but never initializing it. This meant that the value was
2067 incorrect (doubled) while delivering a message in the same process in which
2068 it was received. In the most common configuration of Exim, this never
2069 happens - a fresh exec is done - but it can happen when
2070 deliver_drop_privilege is set.
2072 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2073 it now includes up to 150 characters of the unexpected data in the log
2076 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2077 and building data strings. The size of both of these buffers was 10 000
2078 bytes - far larger than anybody would *ever* want, thought I. Needless to
2079 say, somebody hit the limit. I have increased the maximum line length to
2080 20 000 and the maximum data length of concatenated lines to 100 000. I have
2081 also fixed two bugs, because there was no checking on these buffers. Tsk,
2082 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2083 buffer is too small.
2085 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2086 lsearch lookups. Now it does.
2088 46. When parsing a route_list item in a manualroute router, a fixed-length
2089 buffer was used for the list of hosts. I made this 1024 bytes long,
2090 thinking that nobody would ever have a list of hosts that long. Wrong.
2091 Somebody had a whole pile of complicated expansion conditions, and the
2092 string was silently truncated, leading to an expansion error. It turns out
2093 that it is easier to change to an unlimited length (owing to other changes
2094 that have happened since this code was originally written) than to build
2095 structure for giving a limitation error. The length of the item that
2096 expands into the list of hosts is now unlimited.
2098 47. The lsearch lookup could not handle data where the length of text line was
2099 more than 4095 characters. Such lines were truncated, leading to shortened
2100 data being returned. It should now handle lines of any length.
2102 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2103 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2106 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2107 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2108 no longer happens in comment lines. A list of replacements is now placed
2109 at the head of all of the source files, except those whose only change is
2110 to replace PERL_COMMAND in the very first #! line.
2112 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2113 messages on the queue, with a bottom-up merge sort, using code contributed
2114 by Michael Haardt. This should make operations like -bp somewhat faster on
2115 large queues. It won't affect queue runners, except when queue_run_in_order
2118 51. Installed eximstats 1.31 in the distribution.
2120 52. Added support for SRV lookups to the dnslookup router.
2122 53. If an ACL referred to $message_body or $message_body_end, the value was not
2123 reset for any messages that followed in the same SMTP session.
2125 54. The store-handling optimization for building very long strings was not
2126 differentiating between the different store pools. I don't think this
2127 actually made any difference in practice, but I've tidied it.
2129 55. While running the routers to verify a sender address, $sender_address
2130 was still set to the sender address. This is wrong, because when routing to
2131 send a bounce to the sender, it would be empty. Therefore, I have changed
2132 it so that, while verifying a sender address, $sender_address is set to <>.
2133 (There is no change to what happens when verifying a recipient address.)
2135 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
2136 target A or AAAA records (if not already returned) without resetting the
2137 qualify_single or search_parents options of the DNS resolver. These are
2138 inappropriate in this case because the targets of MX and SRV records must
2139 be FQDNs. A broken DNS record could cause trouble if it happened to have a
2140 target that, when qualified, matched something in the local domain. These
2141 two options are now turned off when doing these lookups.
2143 57. It seems that at least some releases of Reiserfs (which does not have the
2144 concept of a fixed number of inodes) returns zero and not -1 for the
2145 number of available inodes. This interacted badly with check_spool_inodes,
2146 which assumed that -1 was the "no such thing" setting. What I have done is
2147 to check that the total number of inodes is greater than zero before doing
2148 the test of how many are available.
2150 58. When a "warn" ACL statement has a log_message modifier, the message is
2151 remembered, and not repeated. This is to avoid a lot of repetition when a
2152 message has many recipients that cause the same warning to be written.
2153 Howewer, Exim was preserving the list of already written lines for an
2154 entire SMTP session, which doesn't seem right. The memory is now reset if a
2155 new message is started.
2157 59. The "rewrite" debugging flag was not showing the result of rewriting in the
2158 debugging output unless log_rewrite was also set.
2160 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
2161 of (int)(handle) when we know that handle contains (void *)(-1).
2163 61. The Exim daemon panic-logs an error return when it closes the incoming
2164 connection. However "connection reset by peer" seems to be common, and
2165 isn't really an error worthy of noting specially, so that particular error
2168 62. When Exim is trying to find all the local interfaces, it used to panic and
2169 die if the ioctl to get the interface flags failed. However, it seems that
2170 on at least one OS (Solaris 9) it is possible to have an interface that is
2171 included in the list of interfaces, but for which you get a failure error
2172 for this call. This happens when the interface is not "plumbed" into a
2173 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
2174 failure of the "get flags" call assumes that the interface is down.
2176 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
2177 makes life easier for people who are doing arithmetic on fields extracted
2178 from dates, where you often get leading zeros that should not be
2179 interpreted as octal.
2181 64. Added qualify_domain to the redirect router, to override the global
2184 65. If a pathologically long header line contained very many addresses (the
2185 report of this problem mentioned 10 000) and each of them was rewritten,
2186 Exim could use up a very large amount of memory. (It kept on making new
2187 copies of the header line as it rewrote, and never released the old ones.)
2188 At the expense of a bit more processing, the header rewriting function has
2189 been changed so that it no longer eats memory in this way.
2191 66. The generation of the Received: header has been moved from the time that a
2192 message starts to be received, to the time that it finishes. The timestamp
2193 in the Received: header should now be very close to that of the <= log
2194 line. There are two side-effects of this change:
2196 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
2197 logged header lines no longer include the local Received: line, because
2198 it has not yet been created. The same applies to a copy of the message
2199 that is returned to a non-SMTP sender when a message is rejected.
2201 (b) When a filter file is tested using -bf, no additional Received: header
2202 is added to the test message. After some thought, I decided that this
2205 This change does not affect the value of $received_for. It is still set
2206 after address rewriting, but before local_scan() is called.
2208 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
2210 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
2211 gave an unhelpful panic error message, and a defer error. I have managed to
2212 change this behaviour so that it now rejects any supplied certificate,
2213 which seems right, as the list of acceptable certificates is empty.
2215 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
2216 gave an unhelpful defer error. I have not managed to make this reject any
2217 supplied certificates, but the error message it gives is "no certificate
2218 supplied", which is not helpful.
2220 70. exigrep's output now also includes lines that are not associated with any
2221 message, but which match the given pattern. Implemented by a patch from
2222 Martin Sluka, which also tidied up the Perl a bit.
2224 71. Recipient callout verification, like sender verification, was using <> in
2225 the MAIL FROM command. This isn't really the right thing, since the actual
2226 sender may affect whether the remote host accepts the recipient or not. I
2227 have changed it to use the actual sender in the callout; this means that
2228 the cache record is now keyed on a recipient/sender pair, not just the
2229 recipient address. There doesn't seem to be a real danger of callout loops,
2230 since a callout by the remote host to check the sender would use <>.
2231 [SEE ABOVE: changed after hitting problems.]
2233 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
2234 temporary errors. However, in the case of such a code being given after
2235 the end of a data transmission (i.e. after ".") Exim was failing to write
2236 a retry record for the message. (Yes, there was some broken host that was
2237 actually sending 8xx at this point.)
2239 73. An unknown lookup type in a host list could cause Exim to panic-die when
2240 the list was checked. (An example that provoked this was putting <; in the
2241 middle of a list instead of at the start.) If this happened during a DATA
2242 ACL check, a -D file could be left lying around. This kind of configuration
2243 error no longer causes Exim to die; instead it causes a defer errror. The
2244 incident is still logged to the main and panic logs.
2246 74. Buglet left over from Exim 3 conversion. The message "too many messages
2247 in one connection" was written to the rejectlog but not the mainlog, except
2248 when address rewriting (yes!) was being logged.
2250 75. Added write_rejectlog option.
2252 76. When a system filter was run not as root (that is, when system_filter_user
2253 was set), the values of the $n variables were not being returned to the
2254 main process; thus, they were not subsequently available in the $sn
2257 77. Added +return_path_on_delivery log selector.
2259 78. A connection timeout was being treated differently from recipients deferred
2260 when testing hosts_max_try with a message that was older than the host's
2261 retry timeout. (The host should not be counted, thus allowing all hosts to
2262 be tried at least once before bouncing.) This may have been the cause of an
2263 occasionally reported bug whereby a message would remain on the queue
2264 longer than the retry timeout, but would be bounced if a delivery was
2265 forced. I say "may" because I never totally pinned down the problem;
2266 setting up timeout/retry tests is difficult. See also the next item.
2268 79. The ultimate address timeout was not being applied to errors that involved
2269 a combination of host plus message (for example, a timeout on a MAIL
2270 command). When an address resolved to a number of possible hosts, and they
2271 were not all tried for each delivery (e.g. because of hosts_max_try), a
2272 message could remain on the queue longer than the retry timeout.
2274 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
2277 81. Fixed an obscure SMTP outgoing bug which required at least the following
2278 conditions: (a) there was another message waiting for the same server;
2279 (b) the server returned 5xx to all RCPT commands in the first message so
2280 that the message was not completed; (c) the server dropped the connection
2281 or gave a negative response to the RSET that Exim sends to abort the
2282 transaction. The observed case was a dropped connection after DATA that had
2283 been sent in pipelining mode. That is, the server had advertised PIPELINING
2284 but was not implementing it correctly. The effect of the bug was incorrect
2285 behaviour, such as trying another host, and this could lead to a crash.
2291 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
2292 and daemon.c were passed as pointers to ints; they should have been
2293 pointers to socklen_t variables (which are typically unsigned ints).
2295 2. Some signed/unsigned type warnings in the os.c file for Linux have been
2298 3. Fixed a really odd bug that affected only the testing scheme; patching a
2299 certain fixed string in the binary changed the value of another string that
2300 happened to be identical to the end of the original first string.
2302 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
2303 name", it returns that address as the IP address. On some operating
2304 systems (e.g. Solaris), it also passes back the IP address string as the
2305 "host name". However, on others (e.g. Linux), it passes back an empty
2306 string. Exim wasn't checking for this, and was changing the host name to an
2307 empty string, assuming it had been canonicized.
2309 5. Although rare, it is permitted to have more than one PTR record for a given
2310 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
2311 all the names associated with an address, because they do in Solaris.
2312 However, it seems that they do not in Linux for data that comes from the
2313 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
2314 I found this out when I moved to a new Linux workstation and tried to run
2315 the Exim test suite.
2317 To get round this problem I have changed the code so that it now does its
2318 own call to the DNS to look up PTR records when searching for a host name.
2319 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
2320 addresses that are only in /etc/hosts are still found.
2322 This behaviour is, however, controlled by an option called host_lookup_
2323 order, which defaults to "bydns:byaddr". If people want to use the other
2324 order, or indeed, just use one or the other means of lookup, they can
2325 specify it in this variable.
2327 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
2328 some operating systems, this comes back from gethostbyaddr() as an empty
2329 string, and this is what Exim used to test for. However, it seems that in
2330 other systems, "." is yielded. Exim now tests for this case too.
2332 7. The values of check_spool_space and check_log_space are now held internally
2333 as a number of kilobytes instead of an absolute number of bytes. If a
2334 numbers is specified without 'K' or 'M', it is rounded up to the nearest
2335 kilobyte. This means that much larger values can be stored.
2337 8. Exim monitor: an attempt to get the action menu when not actually pointing
2338 at a message produces an empty menu entitled "No message selected". This
2339 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
2340 no entries in it ("Shell widget menu has zero width and/or height"). So I
2341 have added a single, blank menu entry in this case.
2343 9. Added ${quote_local_part.
2345 10. MIME decoding is now applied to the contents of Subject: header lines when
2348 11. Now that a reference to $sender_host_address automatically causes a reverse
2349 lookup to occur if necessary (4.13/18), there is no need to arrange for a
2350 host lookup before query-style lookups in lists that might use this
2351 variable. This has therefore been abolished, and the "net-" prefix is no
2352 longer necessary for query-style lookups.
2354 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
2355 have been a typo for LFLAGS, so it has been changed.
2357 13. The install script calls Exim with "-C /dev/null" in order to find the
2358 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
2359 to be output. Howeve, since Exim outputs its version number before the
2360 error, it didn't break the script. It just looked ugly. I fixed this by
2361 always allowing "-C /dev/null" if the caller is root.
2363 14. Ignore overlarge ACL variable number when reading spool file - insurance
2364 against a later release with more variables having written the file.
2366 15. The standard form for an IPv6 address literal was being rejected by EHLO.
2367 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2368 this, as well as the form without the "IPv6" on the front.
2370 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
2371 OS/Makefile-Darwin file.
2373 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
2374 only to LDAP libraries that do not have LDAP_OPT_DEREF.
2376 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
2377 variable. However, some OS use ints for inodes. I've added cast to long int
2378 to get rid of the compiler warning.
2380 19. I had forgotten to lock out "/../" in configuration file names when
2381 ALT_CONFIG_PREFIX was set.
2383 20. Routers used for verification do not need to specify transports. However,
2384 if such a router generated a host list, and callout was configured, Exim
2385 crashed, because it could not find a port number from the (non-existent)
2386 transport. It now assumes port 25 in this circumstance.
2388 21. Added the -t option to exigrep.
2390 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
2391 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
2392 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
2393 them is set, LOOKUP_LSEARCH is forced.
2395 23. "exim -bV" now outputs a list of lookups that are included in the binary.
2397 24. Added sender and host information to the "rejected by local_scan()" log
2398 line; previously there was no indication of these.
2400 25. Added .include_if_exists.
2402 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
2403 Linux. It turns out that not all file systems support this. Apparently some
2404 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
2405 do it.) To cope with this, the error EINVAL, which means that sync-ing is
2406 not supported on the file descriptor, is now ignored when Exim is trying to
2407 sync a directory. This applies only to Linux.
2409 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
2411 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
2412 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
2415 29. Added support for maildirsize files from supplied patch (modified a bit).
2417 30. The use of :fail: followed by an empty string could lead Exim to respond to
2418 sender verification failures with (e.g.):
2420 550 Verification failed for <xxx>
2421 550 Sender verify failed
2423 where the first response line was missing the '-' that indicates it is not
2424 the final line of the response.
2426 31. The loop for finding the name of the user that called Exim had a hardwired
2427 limit of 10; it now uses the value of finduser_retries, which is used for
2428 all other user lookups.
2430 32. Added $received_count variable, available in data and not_smtp ACLs, and at
2433 33. Exim was neglecting to zero errno before one call of strtol() when
2434 expanding a string and expecting an integer value. On some systems this
2435 resulted in spurious "integer overflow" errors. Also, it was casting the
2436 result into an int without checking.
2438 34. Testing for a connection timeout using "timeout_connect" in the retry rules
2439 did not work. The code looks as if it has *never* worked, though it appears
2440 to have been documented since at least releast 1.62. I have made it work.
2442 35. The "timeout_DNS" error in retry rules, also documented since at least
2443 1.62, also never worked. As it isn't clear exactly what this means, and
2444 clearly it isn't a major issue, I have abolished the feature by treating it
2445 as "timeout", and writing a warning to the main and panic logs.
2447 36. The display of retry rules for -brt wasn't always showing the error code
2450 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
2451 timeout_connect_A, timeout_connect_MX.
2453 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
2454 to the empty sender.
2456 39. The daemon was not analysing the content of -oX till after it had closed
2457 stderr and disconnected from the controlling terminal. This meant that any
2458 syntax errors were only noted on the panic log, and the return code from
2459 the command was 0. By re-arranging the code a little, I've made the
2460 decoding happen first, so such errors now appear on stderr, and the return
2461 code is 1. However, the actual setting up of the sockets still happens in
2462 the disconnected process, so errors there are still only recorded on the
2465 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
2466 connections (as happens on some IP stacks) was logged at start up time as
2467 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
2468 it from "IPv6 and IPv4", which means that two separate sockets are being
2471 41. The debug output for gethostbyname2() or getipnodebyname() failures now
2472 says whether AF_INET or AF_INET6 was passed as an argument.
2474 42. Exiwhat output was messed up when time zones were included in log
2477 43. Exiwhat now gives more information about the daemon's listening ports,
2478 and whether -tls-on-connect was used.
2480 44. The "port" option of the smtp transport is now expanded.
2482 45. A "message" modifier in a "warn" statement in a non-message ACL was being
2483 silently ignored. Now an error message is written to the main and panic
2486 46. There's a new ACL modifier called "logwrite" which writes to a log file
2487 as soon as it is encountered.
2489 47. Added $local_user_uid and $local_user_gid at routing time.
2491 48. Exim crashed when trying to verify a sender address that was being
2494 49. Exim was recognizing only a space character after ".include". It now also
2495 recognizes a tab character.
2497 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
2498 extracting the relevant information from the specification. The man page no
2499 longer contains scrambled data for the -d option, and I've added a section
2500 at the front about calling Exim under different names.
2502 51. Added "extra_headers" argument to the "mail" command in filter files.
2504 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
2507 53. Installed eximstats 1.29.
2509 54. Added transport_filter_timeout as a generic transport option.
2511 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
2512 Cc: header lines. This was required by RFC 822, but it not required by RFC
2515 56. Exim used to add From:, Date:, and Message-Id: header lines to any
2516 incoming messages that did not have them. Now it does so only if the
2517 message originates locally, that is, if there is no associated remote host
2518 address. When Resent- header lines are present, this applies to the Resent-
2519 lines rather than the non-Resent- lines.
2521 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
2522 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
2524 58. Messages for configuration errors now include the name of the main
2525 configuration file - useful now that there may be more than one file in a
2526 list (.included file names were always shown).
2528 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
2529 for those rare installations that do not start the daemon as root or run it
2530 setuid root. I've cut out the call to initgroups() if the daemon is not
2533 60. The Exim user and group can now be bound into the binary as text strings
2534 that are looked up at the start of Exim's processing.
2536 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
2538 62. Added $mailstore_basename variable.
2540 63. Installed patch to sieve.c from Michael Haardt.
2542 64. When Exim failed to open the panic log after failing to open the main log,
2543 the original message it was trying to log was written to stderr and debug
2544 output, but if they were not available (the usual case in production), it
2545 was lost. Now it is written to syslog before the two lines that record the
2546 failures to open the logs.
2548 65. Users' Exim filters run in subprocesses under the user's uid. It is
2549 possible for a "deliver" command or an alias in a "personal" command to
2550 provoke an address rewrite. If logging of address rewriting is configured,
2551 this fails because the process is not running as root or exim. There may be
2552 a better way of dealing with this, but for the moment (because 4.30 needs
2553 to be released), I have disabled address rewrite logging when running a
2554 filter in a non-root, non-exim process.
2560 1. The buildconfig auxiliary program wasn't quoting the value set for
2561 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
2562 not defined. This bug was masked in 4.22 by the effect that was fixed in
2565 2. Some messages that were rejected after a message id was allocated were
2566 shown as "incomplete" by exigrep. It no longer does this for messages that
2567 are rejected by local_scan() or the DATA or non-SMTP ACLs.
2569 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
2570 have allow_domain_literals set, the ID did not get logged in the <= line.
2571 Domain literals are now always recognized in Message-ID: header lines.
2573 4. The first argument for a ${extract expansion item is the key name or field
2574 number. Leading and trailing spaces in this item were not being ignored,
2575 causing some misleading effects.
2577 5. When deliver_drop_privilege was set, single queue runner processes started
2578 manually (i.e. by the command "exim -q") or by the daemon (which uses the
2579 same command in the process it spins off) were not dropping privilege.
2581 6. When the daemon running as "exim" started a queue runner, it always
2582 re-executed Exim in the spun-off process. This is a waste of effort when
2583 deliver_drop_privilege is set. The new process now just calls the
2584 queue-runner function directly.
2590 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
2593 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
2594 ignored. Though the use of -oP was forcing the writing of a pid file, it
2595 was always written to the default place.
2597 3. If the message "no IP address found for host xxxx" is generated during
2598 incoming verification, it is now followed by identification of the incoming
2599 connection (so you can more easily find what provoked it).
2601 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
2603 5. Added some features to "harden" Exim a bit more against certain attacks:
2605 (a) There is now a build-time option called FIXED_NEVER_USERS that can
2606 be put in Local/Makefile. This is like the never_users runtime option,
2607 but it cannot be overridden. The default setting is "root".
2609 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
2610 prefix string with which any file named in a -C command line option
2613 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
2614 is retained for -C and -D only if the caller of Exim is root. Without
2615 it, the exim user may also use -C and -D and retain privilege.
2617 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
2618 command line option is disabled.
2620 6. Macro names set by the -D option must start with an upper case letter, just
2621 like macro names defined in the configuration file.
2623 7. Added "dereference=" facility to LDAP.
2625 8. Two instances of the typo "uknown" in the source files are fixed.
2627 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
2628 the Configure-Makefile script screwed up while processing it.
2630 10. Incorporated PCRE 4.4.
2632 11. The SMTP synchronization check was not operating right at the start of an
2633 SMTP session. For example, it could not catch a HELO sent before the client
2634 waited for the greeting. There is now a check for outstanding input at the
2635 point when the greeting is written. Because of the duplex, asynchronous
2636 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
2637 way, but not yet received, when the check is performed.
2639 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
2640 on TCP/IP sockets, because this apparently causes some broken clients to
2643 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
2644 unchanged) from the Cygwin maintainer.
2646 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
2647 was supported instead of testing for mbx. Effectively a typo.
2649 15. The spa authenticator server code was not checking that the input it
2650 received was valid base64.
2652 16. The debug output line for the "set" modifier in ACLs was not showing the
2653 name of the variable that was being set.
2655 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
2657 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
2658 lookup on its own. Something else was needed to trigger the lookup. For
2659 example, a match in host_lookup or the need for a host name in a host list.
2660 Now, if $sender_host_name is referenced and the host name has not yet been
2661 looked up, a lookup is performed. If the lookup fails, the variable remains
2662 empty, and $host_lookup_failed is set to "1".
2664 19. Added "eqi" as a case-independent comparison operator.
2666 20. The saslauthd authentication condition could segfault if neither service
2667 nor realm was specified.
2669 21. If an overflowing value such as "2048M" was set for message_size_limit, the
2670 error message that was logged was misleading, and incoming SMTP
2671 connections were dropped. The message is now more accurate, and temporary
2672 errors are given to SMTP connections.
2674 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
2675 (except RSET) with a 421 error, until QUIT is received. However, it was
2676 failing to send a response to QUIT.
2678 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
2679 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
2680 ACL is now run after the helo_try_verify_hosts code.
2682 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
2683 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
2684 case-independent, so other case variants are also recognized). Apparently
2685 some systems use these upper case variants.
2687 25. If more than two messages were waiting for the same host, and a transport
2688 filter was specified for the transport, Exim sent two messages over the
2689 same TCP/IP connection, and then failed with "socket operation on non-
2690 socket" when it tried to send the third.
2692 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
2694 27. The extern definition of crypt16() in expand.c was not being excluded when
2695 the OS had its own crypt16() function.
2697 28. Added bounce_return_body as a new option, and bounce_return_size_limit
2698 as a preferred synonym for return_size_limit, both as an option and as an
2701 29. Added LIBS=-liconv to OS/Makefile-OSF1.
2703 30. Changed the default configuration ACL to relax the local part checking rule
2704 for addresses that are not in any local domains. For these addresses,
2705 slashes and pipe symbols are allowed within local parts, but the sequence
2706 /../ is explicitly forbidden.
2708 31. SPA server authentication was not clearing the challenge buffer before
2711 32. log_message in a "warn" ACL statement was writing to the reject log as
2712 well as to the main log, which contradicts the documentation and doesn't
2713 seem right (because no rejection is happening). So I have stopped it.
2715 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
2716 However, I am unable to do any testing of this.
2718 34. Fixed an infelicity in the appendfile transport. When checking directories
2719 for a mailbox, to see if any needed to be created, it was accidentally
2720 using path names with one or more superfluous leading slashes; tracing
2721 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
2723 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
2724 contents are added to the log line that is written for every discarded
2725 recipient. (Previously a log_message setting was ignored.)
2727 36. The ${quote: operator now quotes the string if it is empty.
2729 37. The install script runs exim in order to find its version number. If for
2730 some reason other than non-existence or emptiness, which it checks, it
2731 could not run './exim', it was installing it with an empty version number,
2732 i.e. as "exim-". This error state is now caught, and the installation is
2735 38. An argument was missing from the function that creates an error message
2736 when Exim fails to connect to the socket for saslauthd authentication.
2737 This could cause Exim to crash, or give a corrupted message.
2739 39. Added isip, isip4, and isip6 to ${if conditions.
2741 40. The ACL variables $acl_xx are now saved with the message, and can be
2742 accessed later in routers, transports, and filters.
2744 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
2745 strings in the file are not string-expanded.
2747 42. If a MAIL command specified a SIZE value that was too large to fit into an
2748 int variable, the check against message_size_limit failed. Such values are
2749 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
2750 day this will have to be increased, but I don't think I want to be around
2751 when emails are that large.
2758 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
2759 iconv() is not standard in FreeBSD.
2761 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
2762 IPv6 enabled. The observed symptom was a segmentation fault on return from
2763 the function os_common_find_running_interfaces() in src/os.c.
2765 3. In the check_special_case() function in daemon.c I had used "errno" as an
2766 argument name, which causes warnings on some systems. This was basically a
2767 typo, since it was named "eno" in the comments!
2769 4. The code that waits for the clock to tick (at a resolution of some fraction
2770 of a second) so as to ensure message-id uniqueness was always waiting for
2771 at least one whole tick, when it could have waited for less. [This is
2772 almost certainly not relevant at current processor speeds, where it is
2773 unlikely to ever wait at all. But we try to future-proof.]
2775 5. The function that sleeps for a time interval that includes fractions of a
2776 second contained a race. It did not block SIGALRM between setting the
2777 timer, and suspending (a couple of lines later). If the interval was short
2778 and the sigsuspend() was delayed until after it had expired, the suspension
2779 never ended. On busy systems this could lead to processes getting stuck for
2782 6. Some uncommon configurations may cause a lookup to happen in a queue runner
2783 process, before it forks any delivery processes. The open lookup caching
2784 mechanism meant that the open file or database connection was passed into
2785 the delivery process. The problem was that delivery processes always tidy
2786 up cached lookup data. This could cause a problem for the next delivery
2787 process started by the queue runner, because the external queue runner
2788 process does not know about the closure. So the next delivery process
2789 still has data in the lookup cache. In the case of a file lookup, there was
2790 no problem because closing a file descriptor in a subprocess doesn't affect
2791 the parent. However, if the lookup was caching a connection to a database,
2792 the connection was closed, and the second delivery process was likely to
2793 see errors such as "PGSQL: query failed: server closed the connection
2794 unexpectedly". The problem has been fixed by closing all cached lookups
2795 in a queue runner before running a delivery process.
2797 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
2798 seem to have the "const" qualifier which it has on other OS. I've
2801 8. Change 4.21/2 was too strict. It is only if there are two authenticators
2802 *of the same type* (client or server) with the same public name that an
2803 error should be diagnosed.
2805 9. When Exim looked up a host name for an IP address, but failed to find the
2806 original IP address when looking up the host name (a safety check), it
2807 output the message "<ip address> does not match any IP for NULL", which was
2808 confusing, to say the least. The bug was that the host name should have
2809 appeared instead of "NULL".
2811 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
2812 user that is built into the binary, and the -C or -D options is used, root
2813 privilege is dropped before the configuration file is read. In addition,
2814 logging is switched to stderr instead of the normal log files. If the
2815 configuration then re-defines the Exim user, the unprivileged environment
2816 is probably not what is expected, so Exim logs a panic warning message (but
2819 However, if deliver_drop_privilege is set, the unprivileged state may well
2820 be exactly what is intended, so the warning has been cut out in that case,
2821 and Exim is allowed to try to write to its normal log files.
2827 1. smtp_return_error_details was not giving details for temporary sender
2828 or receiver verification errors.
2830 2. Diagnose a configuration error if two authenticators have the same public
2833 3. Exim used not to create the message log file for a message until the first
2834 delivery attempt. This could be confusing when incoming messages were held
2835 for policy or load reasons. The message log file is now created at the time
2836 the message is received, and an initial "Received" line is written to it.
2838 4. The automatically generated man page for command line options had a minor
2839 bug that caused no ill effects; however, a more serious problem was that
2840 the procedure for building the man page automatically didn't always
2841 operate. Consequently, release 4.20 contains an out-of-date version. This
2842 shouldn't happen again.
2844 5. When building Exim with embedded Perl support, the script that builds the
2845 Makefile was calling 'perl' to find its compile-time parameters, ignoring
2846 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
2848 6. The freeze_tell option was not being used for messages that were frozen on
2849 arrival, either by an ACL or by local_scan().
2851 7. Added the smtp_incomplete_transaction log selector.
2853 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
2854 was accepting AUTH without a new EHLO.
2856 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
2857 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
2860 10. Logging of TCP/IP connections (when configured) now happens in the main
2861 daemon process instead of the child process, so that the TCP/IP connection
2862 count is more accurate (but it can never be perfect).
2864 11. The use of "drop" in a nested ACL was not being handled correctly in the
2865 outer ACL. Now, if condition failure induced by the nested "drop" causes
2866 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
2867 or "require"), the connection is dropped.
2869 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
2870 that yield "discard" can now be used with an "accept" or a "discard" verb,
2871 but an error is generated for any others (because I can't see a useful way
2872 to define what should happen).
2874 13. When an ACL is read dynamically from a file (or anywhere else), the lines
2875 are now processed in the same way as lines in the Exim configuration file.
2876 In particular, continuation lines are supported.
2878 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
2880 15. Added -ti meaning -t -i.
2882 16. Check for letters, digits, hyphens, and dots in the names of dnslist
2883 domains, and warn by logging if others are found.
2885 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
2886 returned from GIFCONF when Exim is trying to find the list of interfaces on
2887 a host. The code in os.c has been modified to copy each ifreq to an aligned
2888 structure in all cases.
2890 Also, in some cases, the returned ifreq's were being copied to a 'struct
2891 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
2892 means the last couple of bytes of an IPv6 address could be chopped if the
2893 ifreq contained only a normal sockaddr (14 bytes storage).
2895 18. Named domain lists were not supported in the hosts_treat_as_local option.
2896 An entry such as +xxxx was not recognized, and was treated as a literal
2899 19. Ensure that header lines added by a DATA ACL are included in the reject log
2900 if the ACL subsequently rejects the message.
2902 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
2903 MD5 (which is deprecated).
2905 21. When testing a filter file using -bf, Exim was writing a message when it
2906 took the sender from a "From " line in the message, but it was not doing so
2907 when it took $return_path from a Return-Path: header line. It now does.
2909 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
2910 with a valid header line field name (a series of printing characters
2911 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
2913 23. Changed "disc" in the source to "disk" to conform to the documentation and
2914 the book and for uniformity.
2916 24. Ignore Sendmail's -Ooption=value command line item.
2918 25. When execve() failed while trying to run a command in a pipe transport,
2919 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
2920 could be confused with a return value of 69 from the command itself. This
2921 has been changed to 127, the value the shell returns if it is asked to run
2922 a non-existent command. The wording for the related log line suggests a
2923 non-existent command as the problem.
2925 26. If received_header_text expands to an empty string, do not add a Received:
2926 header line to the message. (Well, it adds a token one on the spool, but
2927 marks it "old" so that it doesn't get used or transmitted.)
2929 27. Installed eximstats 1.28 (addition of -nt option).
2931 28. There was no check for failure on the call to getsockname() in the daemon
2932 code. This can fail if there is a shortage of resources on the system, with
2933 ENOMEM, for example. A temporary error is now given on failure.
2935 29. Contrary to the C standard, it seems that in some environments, the
2936 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
2937 program. Exim now does this explicitly; it affects the formatting of
2938 timestamps using strftime().
2940 30. If exiqsumm was given junk data, it threw up some uninitialized variable
2941 complaints. I've now initialized all the variables, to avoid this.
2943 32. Header lines added by a system filter were not being "seen" during
2944 transport-time rewrites.
2946 33. The info_callback() function passed to OpenSSL is set up with type void
2947 (*)(SSL *, int, int), as described somewhere. However, when calling the
2948 function (actually a macro) that sets it up, the type void(*)() is
2949 expected. I've put in a cast to prevent warnings from picky compilers.
2951 34. If a DNS black list lookup found a CNAME record, but there were no A
2952 records associated with the domain it pointed at, Exim crashed.
2954 35. If a DNS black list lookup returned more than one A record, Exim ignored
2955 all but the first. It now scans all returned addresses if a particular IP
2956 value is being sought. In this situation, the contents of the
2957 $dnslist_value variable are a list of all the addresses, separated by a
2960 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
2961 to accept a host name and all its aliases if the forward lookup for any of
2962 them yielded the IP address of the incoming connection. Now it accepts only
2963 those names whose forward lookup yields the correct IP address. Any other
2964 names are discarded. This closes a loophole whereby a rogue DNS
2965 administrator could create reverse DNS records to break through a
2966 wildcarded host restriction in an ACL.
2968 37. If a user filter or a system filter that ran in a subprocess used any of
2969 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
2970 the wrong values were passed to the pipe command ($thisaddress had the
2971 value of $0, $0 had the value of $1, etc). This bug was introduced by
2972 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
2974 38. Improved the line breaking for long SMTP error messages from ACLs.
2975 Previously, if there was no break point between 40 and 75 characters, Exim
2976 left the rest of the message alone. Two changes have been made: (a) I've
2977 reduced the minimum length to 35 characters; (b) if it can't find a break
2978 point between 35 and 75 characters, it looks ahead and uses the first one
2979 that it finds. This may give the occasional overlong line, but at least the
2980 remaining text gets split now.
2982 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
2983 file descriptors might be low, and that setting 1000 would always raise it.
2984 It turns out that in some environments, the limit is already over 1000 and
2985 that lowering it causes trouble. So now Exim takes care not to decrease it.
2987 40. When delivering a message, the value of $return_path is set to $sender_
2988 address at the start of routing (routers may change the value). By an
2989 oversight, this default was not being set up when an address was tested by
2990 -bt or -bv, which affected the outcome if any router or filter referred to
2993 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
2994 log only when "message" or "log_message" are set. However, if one of the
2995 conditions was an address verification, or a call to a nested ACL, the
2996 messages generated by the underlying test were being passed through. This
2997 no longer happens. The underlying message is available in $acl_verify_
2998 message for both "message" and "log_message" expansions, so it can be
2999 passed through if needed.
3001 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3002 new expansion $bh_ to give the encoded byte string without charset
3003 translation. Translation happens only if iconv() is available; HAVE_ICONV
3004 indicates this at build time. HEADERS_CHARSET gives the charset to
3005 translate to; headers_charset can change it in the configuration, and
3006 "headers charset" can change it in an individual filter file.
3008 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3009 that creates RFC 2047 encoded "words" labels them as that charset instead
3010 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3011 expansion operator; (ii) when Exim creates a From: line for a local
3012 message; (iii) when a header line is rewritten to include a "phrase" part.
3014 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3015 buggy, causing it to skip the first lines of messages whose message ID
3016 ended in 'D'. This would not have bitten before Exim release 4.14, because
3017 message IDs were unlikely to end in 'D' before then. The effect was to have
3018 incorrect size information for certain domains.
3020 45. #include "config.h" was missing at the start of the crypt16.c module. This
3021 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3024 46. If there was a timeout during a "random" callout check, Exim treated it as
3025 a failure of the random address, and carried on sending RSET and the real
3026 address. If the delay was just some slowness somewhere, the response to the
3027 original RCPT would be taken as a response to RSET and so on, causing
3028 mayhem of various kinds.
3030 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3031 when I implemented it. It didn't allow for the fact that some option values
3032 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3033 the right test for positive values.
3035 48. Domain names in DNS records are case-independent. Exim always looks them up
3036 in lower case. Some resolvers return domain names in exactly the case they
3037 appear in the zone file, that is, they may contain uppercase letters. Not
3038 all resolvers do this - some return always lower case. Exim was treating a
3039 change of case by a resolver as a change of domain, similar to a widening
3040 of a domain abbreviation. This triggered its re-routing code and so it was
3041 trying to route what was effectively the same domain again. This normally
3042 caused routing to fail (because the router wouldn't handle the domain
3043 twice). Now Exim checks for this case specially, and just changes the
3044 casing of the domain that it ultimately uses when it transmits the message
3047 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3050 50. If a filter generated a file delivery with a non-absolute name (possible if
3051 no home directory exists for the router), the forbid_file option was not
3054 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3055 the existing equality matching.
3057 52. Exim was using ints instead of ino_t variables in some places where it was
3058 dealing with inode numbers.
3060 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3061 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3062 called TMPDIR, and if it finds it is different, it changes its value.
3064 54. The smtp_printf() function is now made available to local_scan() so
3065 additional output lines can be written before returning. There is also an
3066 smtp_fflush() function to enable the detection of a dropped connection.
3067 The variables smtp_input and smtp_batched_input are exported to
3070 55. Changed the default runtime configuration: the message "Unknown user"
3071 has been removed from the ACL, and instead placed on the localuser router,
3072 using the cannot_route_message feature. This means that any verification
3073 failures that generate their own messages won't get overridden. Similarly,
3074 the "Unrouteable address" message that was in the ACL for unverifiable
3075 relay addresses has also been removed.
3077 56. Added hosts_avoid_esmtp to the smtp transport.
3079 57. The exicyclog script was not checking for the esoteric option
3080 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3081 will work only if exicyclog is run under the appropriate euid.
3083 58. Following a discussion on the list, the rules by which Exim recognises line
3084 endings on incoming messages have been changed. The -dropcr and drop_cr
3085 options are now no-ops, retained only for backwards compatibility. The
3086 following line terminators are recognized: LF CRLF CR. However, special
3087 processing applies to CR:
3089 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3090 nor a local message in the state where . is a terminator.
3092 (ii) If a bare CR is encountered in a header line, an extra space is added
3093 after the line terminator so as not to end the header. The reasoning
3094 behind this is that bare CRs in header lines are most likely either
3095 to be mistakes, or people trying to play silly games.
3097 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3098 was being incorrectly given as 18 bytes larger than it should have been.
3099 This is a VOB (very old bug).
3101 60. This may never have affected anything current, but just in case it has:
3102 When the local host is found other than at the start of a list of hosts,
3103 the local host, those with the same MX, and any that follow, are discarded.
3104 When the list in question was part of a longer list of hosts, the following
3105 hosts (not currently being processed) were also being discarded. This no
3106 longer happens. I'm not sure if this situation could ever has previously
3109 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3112 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3113 header. This is something that is recommended in a new Internet Draft, and
3114 is something that is documented as being done by Sendmail. There are two
3115 possible values. For messages generated by the autoreply transport, Exim
3118 Auto-Submitted: auto-replied
3120 whereas for all other generated messages (e.g. bounces) it adds
3122 Auto-Submitted: auto-generated
3124 63. The "personal" condition in filters now includes a test for the
3125 Auto-Submitted: header. If it contains the string "auto-" the message it
3126 not considered personal.
3128 64. Added rcpt_include_affixes as a generic transport option.
3130 65. Added queue_only_override (default true).
3132 66. Added the syslog_duplication option.
3134 67. If what should have been the first header line of a message consisted of
3135 a space followed by a colon, Exim was mis-interpreting it as a header line.
3136 It isn't of course - it is syntactically invalid and should therefore be
3137 treated as the start of the message body. The misbehaviour could have
3138 caused a number of strange effects, including loss of data in subsequent
3139 header lines, and spool format errors.
3141 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
3142 client host had authenticated. This control can now be exercised by an ACL
3143 for more flexibility.
3145 69. By default, callouts do not happen when testing with -bh. There is now a
3146 variant, -bhc, which does actually run the callout code, including
3147 consulting and updating the callout cache.
3149 70. Added support for saslauthd authentication, courtesy of Alexander
3152 71. If statvfs() failed on the spool or log directories while checking their
3153 size for availability, Exim confusingly gave the error "space shortage".
3154 Furthermore, in debugging mode it crashed with a floating point exception.
3155 These checks are done if check_{spool,log}_{space,inodes} are set, and when
3156 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
3157 serious problem, Exim now writes to the main and panic logs when this
3158 happens, with details of the failure. It then refuses to accept the
3159 incoming message, giving the message "spool directory problem" or "log
3160 directory problem" with a 421 code for SMTP messages.
3162 72. When Exim is about to re-exec itself, it ensures that the file descriptors
3163 0, 1, and 2 exist, because some OS complain for execs without them (see
3164 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
3165 descriptors. However, the code omitted to check that the open succeeded,
3166 causing mysterious errors if for some reason the permissions on /dev/null
3167 got screwed. Now Exim writes a message to the main and panic logs, and
3168 bombs out if it can't open /dev/null.
3170 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
3171 interact so that it is all more flexible. It is supposed to remain
3172 backwards compatible. Also added extra_local_interfaces.
3174 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
3175 to bomb out with an assertion failure - to the client this appears as a
3176 connection drop. This problem occurs in the part of the code that was taken
3177 from the Samba project. Fortunately, the assertion is in a very simple
3178 function, so I have fixed this by reproducing the function inline in the
3179 one place where it is called, and arranging for authentication to fail
3180 instead of killing the process with assert().
3182 75. The SPA client code was not working when the server requested OEM rather
3183 than Unicode encoding.
3185 76. Added code to make require_files with a specific uid setting more usable in
3186 the case where statting the file as root fails - usually a non-root-mounted
3187 NFS file system. When this happens and the failure is EACCES, Exim now
3188 forks a subprocess and does the per-uid checking as the relevant uid.
3190 77. Added process_log_path.
3192 78. If log_file_path was not explicitly set, a setting of check_log_space or
3193 check_log_inodes was ignored.
3195 79. If a space check for the spool or log partitions fails, the incident is now
3196 logged. Of course, in the latter case the data may get lost...
3198 80. Added the %p formatting code to string_format() so that it can be used to
3199 print addresses in debug_print(). Adjusted all the address printing in the
3200 debugging in store.c to use %p rather than %d.
3202 81. There was a concern that a line of code in smtp_in.c could overflow a
3203 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
3204 initially expressed, the concern was not well-founded, because trailing
3205 spaces are removed early. However, if the trailing spaces were followed by
3206 a NULL, they did not get removed, so the overflow was possible. Two fixes
3209 (a) I re-wrote the offending code in a cleaner fashion.
3210 (b) If an incoming SMTP command contains a NULL character, it is rejected
3213 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
3214 runs initgroups(), so that if the Exim user is in any additional groups,
3215 they will be used during message reception.
3221 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob