1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.377 2006/07/14 14:32:08 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 SC/01 Use a glob alias rather than an array ref in eximstats generated
10 parser. This improves both readability and performance.
12 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
13 Don't display local sender or destination tables in eximstats unless
14 there is data to show.
15 Added average volumes into the eximstats top table text output.
17 SC/03 Collect data on the number of addresses (recipients) as well
18 as the number of messages in eximstats.
20 TF/01 Correct an error in the documentation for the redirect router. Exim
21 does (usually) call initgroups() when daemonizing.
23 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
24 with consistent privilege compared to when running as a daemon.
26 TF/03 Note in the spec that $authenticated_id is not set for local
27 submissions from trusted users.
29 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
30 Thanks to Dean Brooks <dean@iglou.com> for the patch.
32 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
33 by adding some example configuration directives to the default
34 configuration file. A little bit of work is required to uncomment the
35 directives and define how usernames and passwords are checked, but
36 there is now a framework to start from.
38 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
39 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
40 without this. I don't know how relevant this is to other LDAP libraries.
42 PH/02 Add the verb name to the "unknown ACL verb" error.
44 PH/03 Magnus Holmgren's patch for filter_prepend_home.
46 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
48 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
49 directory not expanded when it should be if an expanded home directory
50 was set for the address (which is overridden by the transport).
52 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
55 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
56 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
57 because it is too late at that time, and has no effect.
59 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
92 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
93 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
96 PH/17 Added an extra condition to the default value of delay_warning_condition
100 { match{$h_precedence:}{(?i)bulk|list|junk} }
101 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} }
104 The Auto-Submitted: header is standardised, whereas I don't think
105 Precedence: ever was.
107 PH/18 Refactored debugging code in route_finduser() to show more information,
108 in particular, the error code if getpwnam() issues one.
110 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
111 This is apparently needed in addition to the PH/07 change above to avoid
112 any possible encoding problems.
114 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
115 but not after initializing Perl.
121 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
122 other effects) broke the use of negated acl sub-conditions.
124 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
127 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
128 "Deny" causes Exim to reject the incoming connection with a 554 error.
129 Unfortunately, if there is a major crisis, such as a disk failure,
130 tcp-wrappers gives "deny", whereas what one would like would be some
131 kind of temporary error. A kludge has been added to help with this.
132 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
133 554 error is used if errno is still zero or contains ENOENT (which occurs
134 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
137 PH/03 Add -lutil to the default FreeBSD LIBS setting.
139 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
140 errors. Otherwise a message that provokes a temporary error (when other
141 messages do not) can cause a whole host to time out.
143 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
144 addresses were routed directly to files or pipes from a redirect router.
145 File deliveries just didn't batch; pipe deliveries might have suffered
148 PH/06 A failure to get a lock for a hints database would erroneously always say
149 "Failed to get write lock", even when it was really a read lock.
151 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
152 of 0600. This has been changed to use the value of the lockfile_mode
153 option (which defaults to 0600).
155 PH/08 Applied small patch from the Sieve maintainer.
157 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
158 folder from quota calculations, a direct delivery into this folder messed
159 up the contents of the maildirsize file. This was because the regex was
160 used only to exclude .Trash (or whatever) when the size of the mailbox
161 was calculated. There was no check that a delivery was happening into an
162 excluded directory. This bug has been fixed by ignoring all quota
163 processing for deliveries into excluded directories.
165 PH/10 Added the maildirfolder_create_regex option to appendfile.
171 PH/01 The code for finding all the local interface addresses on a FreeBSD
172 system running IPv6 was broken. This may well have applied to all BSD
173 systems, as well as to others that have similar system calls. The broken
174 code found IPv4 interfaces correctly, but gave incorrect values for the
175 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
176 that it would not match correctly against @[] and not recognize the IPv6
179 PH/02 The ipliteral router was not recognizing addresses of the form user@
180 [ipv6:....] because it didn't know about the "ipv6:" prefix.
182 PH/03 Added disable_ipv6.
184 PH/04 Changed $reply_address to use the raw form of the headers instead of the
185 decoded form, because it is most often used to construct To: headers
186 lines in autoreplies, and the decoded form may well be syntactically
187 invalid. However, $reply_address has leading white space removed, and all
188 newlines turned into spaces so that the autoreply transport does not
191 PH/05 If group was specified without a user on a router, and no group or user
192 was specified on a transport, the group from the router was ignored.
194 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
195 for visible compile-time settings that can be used to change these
196 numbers, for those that want even more. Backwards compatibility with old
197 spool files has been maintained. However, going back to a previous Exim
198 release will lost any variables that are in spool files.
200 PH/07 Two small changes when running in the test harness: increase delay when
201 passing a TCP/IP connection to a new process, in case the original
202 process has to generate a bounce, and remove special handling of
203 127.0.0.2 (sic), which is no longer necessary.
205 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
206 be the same on different OS.
208 PH/09 Moved a debug statement in filter processing to avoid a race problem when
211 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
212 whether --show-vars was specified or not
214 JJ/02 exipick: Added support for new ACL variable spool format introduced
217 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
218 syntactically invalid From: or Reply-to: line, and a filter used this to
219 generate an autoreply, and therefore failed to obtain an address for the
220 autoreply, Exim could try to deliver to a non-existent relative file
221 name, causing unrelated and misleading errors. What now happens is that
222 it logs this as a hard delivery error, but does not attempt to create a
225 PH/11 The exinext utility has a -C option for testing purposes, but although
226 the given file was scanned by exinext itself; it wasn't being passed on
227 when Exim was called.
229 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
230 an end-of-file indication when reading a command response.
232 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
233 compiled. In many other places in Exim, IPv6 addresses are always
234 recognized, so I have changed this. It also means that IPv4 domain
235 literals of the form [IPV4:n.n.n.n] are now always recognized.
237 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
238 used if the router is not running as root, for example, when verifying at
239 ACL time, or when using -bh. The debugging output from this situation was
240 non-existent - all you got was a failure to exec. I have made two
243 (a) Failures to set uid/gid, the current directory, or a process leader
244 in a subprocess such as that created by queryprogram now generate
245 suitable debugging ouput when -d is set.
247 (b) The queryprogram router detects when it is not running as root,
248 outputs suitable debugging information if -d is set, and then runs
249 the subprocess without attempting to change uid/gid.
251 PH/15 Minor change to Makefile for building test_host (undocumented testing
254 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
255 additional section of a DNS packet that returns MX or SRV records.
256 Instead, it always explicitly searches for A/AAAA records. This avoids
257 major problems that occur when a DNS server includes only records of one
258 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
259 fixed another bug: if SRV records were looked up and the corresponding
260 address records were *not* found in the additional section, the port
261 values from the SRV records were lost.
263 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
264 using the correct key (the original address) when searching the retry
265 rules in order to find which one to use for generating the retry hint.
267 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
268 adding the default one. Similarly, if it contains a Reply-To: header, the
269 errors_reply_to option, if set, is not used.
271 PH/19 When calculating a retry time, Exim used to measure the "time since
272 failure" by looking at the "first failed" field in the retry record. Now
273 it does not use this if it is later than than the arrival time of the
274 message. Instead it uses the arrival time. This makes for better
275 behaviour in cases where some deliveries succeed, thus re-setting the
276 "first failed" field. An example is a quota failure for a huge message
277 when small messages continue to be delivered. Without this change, the
278 "time since failure" will always be short, possible causing more frequent
279 delivery attempts for the huge message than are intended.
280 [Note: This change was subsequently modified - see PH/04 for 4.62.]
282 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
283 $1, $2, $3) because the numerical variables can be reset during some
284 expansion items (e.g. "match"), thereby losing the authentication data.
286 PH/21 Make -bV show the size of off_t variables so that the test suite can
287 decide whether to run tests for quotas > 2G.
289 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
290 mailbox_size, and mailbox_filecount in the appendfile transport. If a
291 filecount value is greater than 2G or if a quota value is greater than 2G
292 on a system where the size of off_t is not greater than 4, a panic error
295 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
296 never match. The debug and -bh output now contains an explicit error
297 message indicating a malformed IPv4 address or mask.
299 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
300 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
303 PH/25 Do not write to syslog when running in the test harness. The only
304 occasion when this arises is a failure to open the main or panic logs
305 (for which there is an explicit test).
307 PH/26 Added the /no_tell option to "control=freeze".
309 PH/27 If a host name lookup failed very early in a connection, for example, if
310 the IP address matched host_lookup and the reverse lookup yielded a name
311 that did not have a forward lookup, an error message of the form "no IP
312 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
313 could be logged. Now it outputs the IP address instead of "NULL".
315 PH/28 An enabling patch from MH: add new function child_open_exim2() which
316 allows the sender and the authenticated sender to be set when
317 submitting a message from within Exim. Since child_open_exim() is
318 documented for local_scan(), the new function should be too.
320 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
321 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
322 results in an empty string is now treated as unset.
324 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
326 PH/31 Added sender_verify_fail logging option.
328 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
329 needed by RFC 822 but not by RFC 2822 was commented out. I have now
330 tidied the source and removed it altogether.
332 PH/33 When a queue run was abandoned because the load average was too high, a
333 log line was always written; now it is written only if the queue_run log
334 selector is set. In addition, the log line for abandonment now contains
335 information about the queue run such as the pid. This is always present
336 in "start" and "stop" lines but was omitted from the "abandon" line.
338 PH/34 Omit spaces between a header name and the colon in the error message that
339 is given when verify = headers_syntax fails (if there are lots of them,
340 the message gets confusing).
342 PH/35 Change the default for dns_check_names_pattern to allow slashes within
343 names, as there are now some PTR records that contain slashes. This check
344 is only to protect against broken name servers that fall over on strange
345 characters, so the fact that it applies to all lookups doesn't matter.
347 PH/36 Now that the new test suite is complete, we can remove some of the
348 special code in Exim that was needed for the old test suite. For example,
349 sorting DNS records because real resolvers return them in an arbitrary
350 order. The new test suite's fake resolver always returns records in the
353 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
354 bounces) except when queue_only is set, to avoid logging races between
355 the different processes.
357 PH/38 Panic-die if .include specifies a non-absolute path.
359 PH/39 A tweak to the "H" retry rule from its user.
361 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
362 a label. They prevented compilation on older perls.
364 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
365 a warning to be raised on newish perls.
367 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
368 on queue. Changes to match documented behaviour of showing count of
369 messages matching specified criteria.
371 PH/40 Changed the default ident timeout from 30s to 5s.
373 PH/41 Added support for the use of login_cap features, on those BSD systems
374 that have them, for controlling the resources used by pipe deliveries.
376 PH/42 The content-scanning code uses fopen() to create files in which to put
377 message data. Previously it was not paying any attention to the mode of
378 the files. Exim runs with umask(0) because the rest of the code creates
379 files with open(), and sets the required mode explicitly. Thus, these
380 files were ending up world-writeable. This was not a big issue, because,
381 being within the spool directory, they were not world-accessible. I have
382 created a function called modefopen, which takes an additional mode
383 argument. It sets umask(777), creates the file, chmods it to the required
384 mode, then resets the umask. All the relevant calls to fopen() in the
385 content scanning code have been changed to use this function.
387 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
388 to 24 hours. This avoids potential overflow problems when processing G
389 and H retry rules. I suspect nobody ever tinkers with this value.
391 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
393 PH/45 When the plaintext authenticator is running as a client, the server's
394 challenges are checked to ensure they are valid base64 strings. By
395 default, the authentication attempt is cancelled if an invalid string is
396 received. Setting client_ignore_invalid_base64 true ignores these errors.
397 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
398 they are received. Thus, the responses can be made to depend on the
399 challenges. If an invalid string is ignored, an empty string is placed in
402 PH/46 Messages that are created by the autoreply transport now contains a
403 References: header, in accordance with RFCs 2822 and 3834.
405 PH/47 Added authenticated_sender_force to the smtp transport.
407 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
409 PH/49 Installed latest patch from the Sieve maintainer.
411 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
412 was also set, the appendfile transport was unnecessarily scanning a
413 directory of message files (e.g. for maildir delivery) to find the count
414 of files (along with the size), even though it did not need this
415 information. It now does the scan only if it needs to find either the
416 size of the count of files.
418 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
420 PH/52 Two bugs concerned with error handling when the smtp transport is
423 (i) Exim was not creating retry information for temporary errors given
424 for individual recipients after the DATA command when the smtp transport
425 was used in LMTP mode. This meant that they could be retried too
426 frequently, and not timed out correctly.
428 (ii) Exim was setting the flag that allows error details to be returned
429 for LMTP errors on RCPT commands, but not for LMTP errors for individual
430 recipients that were returned after the DATA command.
432 PH/53 This is related to PH/52, but is more general: for any failing address,
433 when detailed error information was permitted to be returned to the
434 sender, but the error was temporary, then after the final timeout, only
435 "retry timeout exceeded" was returned. Now it returns the full error as
436 well as "retry timeout exceeded".
438 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
439 do this, and (what is worse) MTAs that accept it.
441 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
442 will now be deprecated.
444 PH/56 New os.c-cygwin from the Cygwin maintainer.
446 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
447 formats (previously only available in exim formats via -bpr, -bpru,
448 and -bpra. Now also available in native and exiqgrep formats)
450 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
451 with very large, slow to parse queues
453 JJ/08 exipick: added ! as generic prefix to negate any criteria format
455 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
457 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
458 responses to authentication challenges, though it was showing the
459 challenges; (ii) I've removed the CR characters from the debug output for
462 PH/58 Allow for the insertion of a newline as well as a space when a string
463 is turned into more than one encoded-word during RFC 2047 encoding. The
464 Sieve code now uses this.
466 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
467 data_4xx, lost_connection, tls_required.
469 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
470 message was being sent as an SMTP response.
472 PH/61 Add -l and -k options to exicyclog.
474 PH/62 When verifying, if an address was redirected to one new address, so that
475 verification continued, and the new address failed or deferred after
476 having set something in $address_data, the value of $address_data was not
477 passed back to the ACL. This was different to the case when no
478 redirection occurred. The value is now passed back in both cases.
480 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
481 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
482 use login_cap.h, so on its own it isn't the distinguishing feature. The
483 new name refers directly to the setclassresources() function.
485 PH/65 Added configuration files for NetBSD3.
487 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
489 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
490 is preferred over IPv4.
492 PH/68 The bounce_return_message and bounce_return_body options were not being
493 honoured for bounces generated during the reception of non-SMTP messages.
494 In particular, this applied to messages rejected by the ACL. This bug has
495 been fixed. However, if bounce_return_message is true and bounce_return_
496 body is false, the headers that are returned for a non-SMTP message
497 include only those that have been read before the error was detected.
498 (In the case of an ACL rejection, they have all been read.)
500 PH/69 The HTML version of the specification is now built in a directory called
501 spec_html instead of spec.html, because the latter looks like a path with
502 a MIME-type, and this confuses some software.
504 PH/70 Catch two compiler warnings in sieve.c.
506 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
507 function verify_get_ident() calls ip_connect() to connect a socket, but
508 if the "connect()" function timed out, ip_connect() used to close the
509 socket. However, verify_get_ident() also closes the socket later, and in
510 between Exim writes to the log, which may get opened at this point. When
511 the socket was closed in ip_connect(), the log could get the same file
512 descriptor number as the socket. This naturally causes chaos. The fix is
513 not to close the socket in ip_connect(); the socket should be closed by
514 the function that creates it. There was only one place in the code where
515 this was missing, in the iplookup router, which I don't think anybody now
516 uses, but I've fixed it anyway.
518 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
519 well as to direct DNS lookups. Otherwise the handling of names in host
520 lists is inconsistent and therefore confusing.
526 PH/01 Two changes to the default runtime configuration:
528 (1) Move the checks for relay_from_hosts and authenticated clients from
529 after to before the (commented out) DNS black list checks.
531 (2) Add control=submission to the relay_from_hosts and authenticated
532 clients checks, on the grounds that messages accepted by these
533 statements are most likely to be submissions.
535 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
537 (1) Generate an error if the third argument for the ${prvs expansion is
540 (2) Treat a missing third argument of ${prvscheck as if it were an empty
543 (3) Reset the variables that are obtained from the first argument of
544 ${prvscheck and used in the second argument before leaving the code,
545 because their memory is reclaimed, so using them afterwards may do
548 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
549 one (it's much easier than Tom thought :-).
551 (5) Because of (4), we can now allow for the use of $prvscheck_result
552 inside the third argument.
554 PH/03 For some reason, the default setting of PATH when running a command from
555 a pipe transport was just "/usr/bin". I have changed it to
558 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
559 anything to be listed in the output from -bV.
561 PH/05 When a filter generated an autoreply, the entire To: header line was
562 quoted in the delivery log line, like this:
564 => >A.N.Other <ano@some.domain> <original@ddress> ...
566 This has been changed so that it extracts the operative address. There
567 may be more than one such address. If so, they are comma-separated, like
570 => >ano@some.domain,ona@other.domain <original@ddress> ...
572 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
573 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
574 not being looked up in the rDNS to get a host name, Exim was showing the
575 IP address twice in Received: lines, even though the IP addresses were
576 identical. For example:
578 Received: from [1.2.3.4] (helo=[1.2.3.4])
580 However, if the real host name was known, it was omitting the HELO data
581 if it matched the actual IP address. This has been tidied up so that it
582 doesn't show the same IP address twice.
584 PH/07 When both +timestamp and +memory debugging was on, the value given by
585 $tod_xxx expansions could be wrong, because the tod_stamp() function was
586 called by the debug printing, thereby overwriting the timestamp buffer.
587 Debugging no longer uses the tod_stamp() function when +timestamp is set.
589 PH/08 When the original message was included in an autoreply transport, it
590 always said "this is a copy of the message, including all the headers",
591 even if body_only or headers_only was set. It now gives an appropriate
594 PH/09 Applied a patch from the Sieve maintainer which:
596 o fixes some comments
597 o adds the (disabled) notify extension core
598 o adds some debug output for the result of if/elsif tests
599 o points to the current vacation draft in the documentation
600 and documents the missing references header update
604 o fixes a bug in processing the envelope test (when testing
605 multiple envelope elements, the last element determinted the
608 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
609 Electronic Mail") by including:
611 Auto-submitted: auto-generated
613 in the messages that it generates (bounce messages and others, such as
614 warnings). In the case of bounce messages for non-SMTP mesages, there was
615 also a typo: it was using "Auto_submitted" (underscore instead of
616 hyphen). Since every message generated by Exim is necessarily in response
617 to another message, thes have all been changed to:
619 Auto-Submitted: auto-replied
621 in accordance with these statements in the RFC:
623 The auto-replied keyword:
625 - SHOULD be used on messages sent in direct response to another
626 message by an automatic process,
628 - MUST NOT be used on manually-generated messages,
630 - MAY be used on Delivery Status Notifications (DSNs) and Message
631 Disposition Notifications (MDNs),
633 - MUST NOT be used on messages generated by automatic or periodic
634 processes, except for messages which are automatic responses to
637 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
638 to the default Received: header definition.
640 PH/12 Added log selector acl_warn_skipped (default on).
642 PH/13 After a successful wildlsearch lookup, discard the values of numeric
643 variables because (a) they are in the wrong storage pool and (b) even if
644 they were copied, it wouldn't work properly because of the caching.
646 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
647 checking when decoding. Apparently there are clients that generate
648 overlong encoded strings. Why am I not surprised?
650 PH/15 If the first argument of "${if match_address" was not empty, but did not
651 contain an "@" character, Exim crashed. Now it writes a panic log message
652 and treats the condition as false.
654 PH/16 In autoreply, treat an empty string for "once" the same as unset.
656 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
657 extension "envelope-auth". The code is finished and in agreement with
658 other implementations, but there is no documentation so far and in fact,
659 nobody wrote the draft yet. This extension is currently #undef'ed, thus
660 not changing the active code.
662 Print executed "if" and "elsif" statements when debugging is used. This
663 helps a great deal to understand what a filter does.
665 Document more things not specified clearly in RFC3028. I had all this
666 sorted out, when out of a sudden new issues came to my mind. Oops."
668 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
671 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
672 canonical form (as documented). However, after a host name lookup from
673 the IP address, check_host() was doing a simple string comparison with
674 addresses acquired from the DNS when checking that the found name did
675 have the original IP as one of its addresses. Since any found IPv6
676 addresses are likely to be in abbreviated form, the comparison could
677 fail. Luckily, there already exists a function for doing the comparison
678 by converting both addresses to binary, so now that is used instead of
681 PH/20 There was another similar case to PH/19, when a complete host name was
682 given in a host list; looking up its IP address could give an abbreviated
683 form, whereas the current host's name might or might not be abbreviated.
684 The same fix has been applied.
690 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
691 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
694 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
695 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
697 PH/03 Typo: missing ".o" in src/pcre/Makefile.
699 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
700 header line, restrict the check to what is listed in RFCs 2369 and 2929.
701 Also, for "Auto-Submitted", treat anything other than "no" as
702 non-personal, in accordance with RFC 3834. (Previously it treated
703 anything starting "auto-" as non-personal.)
705 TF/01 The control=submission/name=... option had a problem with syntax
706 errors if the name included a slash character. The /name= option
707 now slurps the rest of the string, so it can include any characters
708 but it must come last in the list of options (after /sender_retain
711 PH/05 Some modifications to the interface to the fake nameserver for the new
719 TK/01 Added the "success_on_redirect" address verification option. See
720 NewStuff for rationale and an example.
722 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
724 PH/02 Patch to exigrep to allow it to work on syslog lines.
726 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
727 fread() to skip over the body file's header line, because in Cygwin the
728 header line is locked and is inaccessible.
730 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
731 co-exist for some time) to make it clear that it is the Exim ID that is
732 referenced, not the Message-ID: header line.
734 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
735 string_format() function, because snprintf() does not exist on all
738 PH/06 The use of forbid_filter_existstest now also locks out the use of the
739 ${stat: expansion item.
741 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
742 protocol synchronization error", to keep the pedants happy.
744 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
745 well as for IRIX systems, when gcc is being used. See the host.c source
748 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
750 PH/10 Named domain lists were not working if used in a queue_smtp_domains
753 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
754 transport and to the smtp transport in LMTP mode.
756 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
758 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
759 run a filter in a subprocess. This could lead to confusion in subsequent
760 lookups in the parent process. There should also be a search_tidyup() at
761 the end of the subprocess.
763 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
764 only if the host matched helo_try_verify_hosts, which caused the
765 verification to occur when the EHLO/HELO command was issued. The ACL just
766 tested the remembered result. Now, if a previous verification attempt has
767 not happened, "verify = helo" does it there and then.
769 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
771 TK/03 Fix log output including CR from clamd.
773 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
774 exist provoked a memory error which could cause a segfault.
776 PH/15 Installed PCRE 6.2
778 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
780 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
781 of the problem. Specifically, suggested +O2 rather than +O1 for the
784 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
786 PH/20 If a delivery was routed to a non-standard port by means of an SRV
787 record, the port was not correctly logged when the outgoing_port log
788 selector was set (it logged the transort's default port).
790 PH/21 Added support for host-specific ports to manualroute, queryprogram,
791 fallback_hosts, and "hosts" in the smtp transport.
793 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
794 host errors such as "Connection refused".
796 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
797 authentication with radiusclient 0.4.9:
799 - Error returned from rc_read_config was caught wrongly
800 - Username/password not passed on to radius server due to wrong length.
802 The presumption is that some radiusclient API changes for 4.51/PH/17
803 were not taken care of correctly. The code is still untested by me (my
804 Linux distribution still has 0.3.2 of radiusclient), but it was
805 contributed by a Radius user.
807 PH/24 When doing a callout, the value of $domain wasn't set correctly when
808 expanding the "port" option of the smtp transport.
810 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
811 while reading a MIME header. Thanks to Tom Hughes for a patch.
813 PH/24 Include config.h inside local_scan.h so that configuration settings are
816 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
817 that in an ACL for RCPT (for example), you can examine exactly what was
820 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
821 commands, but it was not correctly comparing the address with the actual
822 client host address. Thus, it would show the EHLO address in Received:
823 header lines when this was not necessary.
825 PH/27 Added the % operator to ${eval:}.
827 PH/28 Exim tries to create and chdir to its spool directory when it starts;
828 it should be ignoring failures (because with -C, for example, it has lost
829 privilege). It wasn't ignoring creation failures other than "already
832 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
835 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
836 because an input line was too long, either on its own, or by virtue of
837 too many continuations, the temporary file was not being removed, and the
838 return code was incorrect.
840 PH/31 Missing "BOOL" in function definition in filtertest.c.
842 PH/32 Applied Sieve patches from the maintainer.
844 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
846 PH/33 Added "verify = not_blind".
848 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
849 Local/Makefile (with some defaults set). These are used in built scripts
850 such as exicyclog, but they have never been used in the exim_install
851 script (though there are many overriding facilities there). I have
852 arranged that the exim_install script now takes note of these two
855 PH/35 Installed configuration files for Dragonfly.
857 PH/36 When a locally submitted message by a trusted user did not contain a
858 From: header, and the sender address was obtained from -f or from an SMTP
859 MAIL command, and the trusted user did not use -F to supply a sender
860 name, $originator_name was incorrectly used when constructing a From:
861 header. Furthermore, $originator_name was used for submission mode
862 messages from external hosts without From: headers in a similar way,
863 which is clearly wrong.
865 PH/37 Added control=suppress_local_fixups.
867 PH/38 When log_selector = +received_sender was set, and the addition of the
868 sender made the log line's construction buffer exactly full, or one byte
869 less than full, an overflow happened when the terminating "\n" was
872 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
873 when the result of a list match is failure because a DNS lookup failed.
875 PH/40 RM_COMMAND is now used in the building process.
877 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
878 the "build-* directories that it finds.
880 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
881 address in a domain literal was a prefix of an interface address.
883 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
884 when verifying a sender address, unless rewrite_headers is false.
886 PH/44 Wrote a long comment about why errors_to addresses are verified as
887 recipients, not senders.
889 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
890 the ratelimit ACL was added.
892 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
894 PH/46 Added extra information about PostgreSQL errors to the error string.
896 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
897 avoiding the need to install special zones in a real server. This is
898 backwards compatible; if it can't find the fake resolver, it drops back.
899 Thus, both old and new test suites can be run.
901 TF/02 Added util/ratelimit.pl
903 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
904 clock is set back in time.
906 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
907 Candler <B.Candler@pobox.com>.
909 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
910 OK for addresses that are the result of redirections.
912 PH/48 A number of further additions for the benefit of the new test suite,
913 including a fake gethostbyname() that interfaces to the fake DNS resolver
916 TF/06 The fix for widen_domains has also been applied to qualify_single and
917 search_parents which are the other dnslookup options that can cause
920 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
923 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
925 TF/07 Exim produced the error message "an SRV record indicated no SMTP
926 service" if it encountered an MX record with an empty target hostname.
927 The message is now "an MX or SRV record indicated no SMTP service".
929 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
930 if the DNS of the sending site is misconfigured. This is quite a
931 common situation. This change restores the behaviour of treating a
932 helo verification defer as a failure.
934 PH/51 If self=fail was set on a router, the bounce message did not include the
935 actual error message.
941 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
943 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
944 command itself ended in error, the underlying message about the transport
945 filter timeout was being overwritten with the pipe command error. Now the
946 underlying error message should be appended to the second error message.
948 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
950 PH/02 Reduce the amount of output that "make" produces by default. Full output
951 can still be requested.
953 PH/03 The warning log line about a condition test deferring for a "warn" verb
954 was being output only once per connection, rather than after each
955 occurrence (because it was using the same function as for successful
956 "warn" verbs). This seems wrong, so I have changed it.
958 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
959 it should not have, which might have caused a crash in the right
960 circumstances, but probably never did.
962 PH/04 Installed a modified version of Tony Finch's patch to make submission
963 mode fix the return path as well as the Sender: header line, and to
964 add a /name= option so that you can make the user's friendly name appear
967 TF/03 Added the control = fakedefer ACL modifier.
969 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
970 Mark Lowes for thorough testing.
972 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
974 TK/03 Merged latest SRS patch from Miles Wilton.
976 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
977 with the definition in sysexits.h (which is #included earlier).
978 Fortunately, Exim does not actually use EX_OK. The code used to try to
979 preserve the sysexits.h value, by assumimg that macro definitions were
980 scanned for macro replacements. I have been disabused of this notion,
981 so now the code just undefines EX_OK before #including unistd.h.
983 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
984 in the smtp transport. When a block could not be written in a single
985 write() function, the timeout was being re-applied to each part-write.
986 This seems wrong - if the receiver was accepting one byte at a time it
987 would take for ever. The timeout is now adjusted when this happens. It
988 doesn't have to be particularly precise.
990 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
991 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
993 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
994 without a domain if the check to <postmaster@domain> fails.
996 SC/01 Eximstats: added -xls and the ability to specify output files
997 (patch written by Frank Heydlauf).
999 SC/02 Eximstats: use FileHandles for outputing results.
1001 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1003 SC/04 Eximstats: fixed display of large numbers with -nvr option
1005 SC/05 Eximstats: fixed merging of reports with empty tables.
1007 SC/06 Eximstats: added the -include_original_destination flag
1009 SC/07 Eximstats: removed tabs and trailing whitespace.
1011 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1013 TK/06 MBOX spool code: Add real "From " MBOX separator line
1014 so the .eml file is really in mbox format (even though
1015 most programs do not really care). Patch from Alex Miller.
1017 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1018 The latter is generated from $received_to and is only set if the
1019 message has one envelope recipient. SA can use these headers,
1020 obviously out-of-the-box. Patch from Alex Miller.
1022 PH/08 The ${def test on a variable was returning false if the variable's
1023 value was "0", contrary to what the specification has always said!
1024 The result should be true unless the variable is empty.
1026 PH/09 The syntax error of a character other than { following "${if
1027 def:variable_name" (after optional whitespace) was not being diagnosed.
1028 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1029 accidental colon was present, for example, could give incorrect results.
1031 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1032 result is used (not including appendfile, where other changes are about
1035 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1036 This involved changing a lot of size variables from int to off_t. It
1037 should work with maildirs and everything.
1039 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1040 spamd dying while we are connected to it.
1042 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1045 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1046 function for BDB 4.3.
1048 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1050 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1051 the sources for PCRE have been re-arranged and more files are now
1054 PH/15 The code I had for printing potentially long long variables in PH/11
1055 above was not the best (it lost precision). The length of off_t variables
1056 is now inspected at build time, and an appropriate printing format (%ld
1057 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1058 to be "long long int" or "long int". This is needed for the internal
1059 formatting function string_vformat().
1061 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1062 the configuration file to be ":syslog", then the script "guesses" where
1063 the logs files are, rather than using the compiled in default. In our
1064 case the guess is not the same as the compiled default, so the script
1065 suddenly stopped working when I started to use syslog. The patch checks
1066 to see if log_file_path is "". If so, it attempts to read it from exim
1067 with no configuration file to get the compiled in version, before it
1068 falls back to the previous guessing code."
1070 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1071 implementing BATV in an Exim configuration. See NewStuff for the gory
1074 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1075 Makefile that are specific to HP-UX.
1077 PH/18 If the "use_postmaster" option was set for a recipient callout together
1078 with the "random" option, the postmaster address was used as the MAIL
1079 FROM address for the random test, but not for the subsequent recipient
1080 test. It is now used for both.
1082 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1083 patch removes a few documentation additions to RFC 3028, because the
1084 latest draft now contains them. It adds the new en;ascii-case comparator
1085 and a new error check for 8bit text in MIME parts. Comparator and
1086 require names are now matched exactly. I enabled the subaddress
1087 extension, but it is not well tested yet (read: it works for me)."
1089 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1090 rework some of the code of TK/09 above to avoid the hardwired use of
1091 "%lld" and "long long". Replaced the call to snprintf() with a call to
1094 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1095 records point to non-existent hosts", "retry timeout exceeded", and
1096 "retry time not reached for any host after a long failure period".
1098 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1099 experimental DomainKeys support:
1101 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1102 (2) On an error such as an illegally used "control", the wrong name for
1103 the control was given.
1105 These problems did NOT occur unless DomainKeys support was compiled.
1107 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1109 PH/24 Added ${if match_ip condition.
1111 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1112 databases so that it will be absolutely obvious if a crash occurs in the
1113 DB library. This is a regular occurrence (often caused by mis-matched
1116 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1117 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1118 picked up on a user's system that detects such things. There doesn't seem
1119 to be a gcc warning option for this - only an attribute that has to be
1120 put on the function's prototype. It seems that in Fedora Core 4 they have
1121 set this on a number of new functions. No doubt there will be more in due
1124 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1125 specify a transport. However, if an address that was verified by such a
1126 router was the subject of a callout, Exim crashed because it tried to
1127 read the rcpt_include_affixes from the non-existent transport. Now it
1128 just assumes that the setting of that option is false. This bug was
1129 introduced by 4.51/PH/31.
1131 PH/28 Changed -d+all to exclude +memory, because that information is very
1132 rarely of interest, but it makes the output a lot bigger. People tend to
1133 do -d+all out of habit.
1135 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1136 code in os-type was giving problems when libc.so lives in lib64, like on
1139 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1140 aren't the modern standard, and it seems that some systems' include files
1141 don't always have them. Exim was already checking for some of the newer
1142 ones like T_AAAA, and defining it itself. I've added checks for all the
1143 record types that Exim uses.
1145 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1146 not automatically generating a new one, as it is supposed to. This
1147 prevented TLS from working. If the file did exist, but contained invalid
1148 data, a new version was generated, as expected. It was only the case of a
1149 non-existent file that was broken.
1151 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1152 with a change in libdomainkeys > 0.64.
1154 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1155 from DNS. If the selector record carries the flag, it now has
1156 precedence over the domain-wide flag.
1158 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1160 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1161 the use of a port name that isn't defined in /etc/services) occurred, the
1162 message was deferred as in a normal delivery, and thus remained on the
1163 spool, instead of being failed because of the mua_wrapper setting. This
1164 is now fixed, and I tidied up some of the mua_wrapper messages at the
1167 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1168 the messages in a hash of arrays rather than using individual hashes.
1169 This is a bit cleaner and results in dramatic memory savings, albeit
1170 at a slight CPU cost.
1172 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1173 as requested by Marc Sherman.
1175 SC/10 Eximstats: added histograms for user specified patterns as requested
1178 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1180 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1181 fopen() in the content-scanning modules that did not already have it.
1187 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1188 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1190 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1192 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1194 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1195 to test Sieve filters that use "vacation".
1197 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1198 that changes the way the GnuTLS parameters are stored in the cache file.
1199 The new format can be generated externally. For backward compatibility,
1200 if the data in the cache doesn't make sense, Exim assumes it has read an
1201 old-format file, and it generates new data and writes a new file. This
1202 means that you can't go back to an older release without removing the
1205 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1206 work if there are any delivery delays because "one_time" forces the
1207 parent to be marked "delivered", so its unseen clone is never tried
1208 again. For this reason, Exim now forbids the simultaneous setting of
1211 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1212 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1213 entry if you want to know the details. The fix, however, neglected to
1214 consider the case when local delivery batching is involved. The test for
1215 "previously delivered" was not happening when checking to see if an
1216 address could be batched with a previous (undelivered) one; under
1217 certain circumstances this could lead to multiple deliveries to the same
1220 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1221 in its include files, and this causes problems building Exim.
1223 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1224 header_syntax) but Exim was just ignoring anything given after a slash.
1225 In particular, this caused confusion with an attempt to use "verify =
1226 reverse_host_lookup/defer_ok". An error is now given when options are
1227 supplied for verify items that do not have them. (Maybe reverse_host_
1228 lookup should have a defer_ok option, but that's a different point.)
1230 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1231 defined by RFC 821) to 2048, because there were problems with some AUTH
1232 commands, and RFC 1869 says the size should be increased for extended
1233 SMTP commands that take arguments.
1235 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1238 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1239 "unknown" error; now it says that the functionality isn't in the binary.
1241 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1242 an address' error message when a string expansion fails (syntax or
1243 whatever). Otherwise the password may appear in the log. Following change
1244 PH/42 below, there is no longer a chance of it appearing in a bounce
1247 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1249 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1250 of its addresses was ever tried. (Bugzilla bug #2.)
1252 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1253 the result incorrectly in the debug output. (It correctly added a newline
1254 to what was transported.)
1256 TF/01 Added $received_time.
1258 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1259 commented out examples of how to interface to a virus scanner and to
1260 SpamAssassin. Also added commented examples of av_scanner and
1261 spamd_address settings.
1263 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1264 and controls are allowed in which ACLs. There were a couple of minor
1265 errors. Some of the entries in the conditions table (which is a table of
1266 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1267 negation of where the condition IS allowed.
1269 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1271 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1272 header file does not have a version number, so I've had to invent a new
1273 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1274 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1275 radiusclient), but it was contributed by a Radius user.
1277 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1278 files or directories, for OpenSSL.
1280 PH/19 When an Exim process that is running as root has to create an Exim log
1281 file, it does so in a subprocess that runs as exim:exim so as to get the
1282 ownership right at creation (otherwise, other Exim processes might see
1283 the file with the wrong ownership). There was no test for failure of this
1284 fork() call, which would lead to the process getting stuck as it waited
1285 for a non-existent subprocess. Forks do occasionally fail when resources
1286 run out. I reviewed all the other calls to fork(); they all seem to check
1289 PH/20 When checking for unexpected SMTP input at connect time (before writing
1290 the banner), Exim was not dealing correctly with a non-positive return
1291 from the read() function. If the client had disconnected by this time,
1292 the result was a log entry for a synchronization error with an empty
1293 string after "input=" when read() returned zero. If read() returned -1
1294 (an event I could not check), uninitialized data bytes were printed.
1295 There were reports of junk text (parts of files, etc) appearing after
1298 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1300 PH/22 Added support for macro redefinition, and (re)definition in between
1301 driver and ACL definitions.
1303 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1304 forgetting to use the resulting value; it was using the unexpanded value.
1306 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1307 hadn't been configured. The fix is from Juergen Kreileder, who
1308 understands it better than I do:
1310 "Here's what I see happening with three configured cyrus_sasl
1311 authenticators configured (plain, login, cram-md5):
1313 On startup auth_cyrus_sasl_init() gets called for each of these.
1314 This means three calls to sasl_listmech() without a specified mech_list.
1315 => SASL tests which mechs of all available mechs actually work
1316 => three warnings about OTP not working
1317 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1319 With the patch, sasl_listmech() also gets called three times. But now
1320 SASL's mech_list option is set to the server_mech specified in the the
1321 authenticator. Or in other words, the answer from sasl_listmech()
1322 gets limited to just the mech you're testing for (which is different
1324 => the return list contains just 'plain' or 'login', 'cram-md5' or
1325 nothing depending on the value of ob->server_mech.
1327 I've just tested the patch: Authentication still works fine,
1328 unavailable mechs specified in the exim configuration are still
1329 caught, and the auth.log warnings about OTP are gone."
1331 PH/25 When debugging is enabled, the contents of the command line are added
1332 to the debugging output, even when log_selector=+arguments is not
1335 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1336 answer is "GNU", and only if the return is "GNU/something" is the answer
1339 PH/27 $acl_verify_message is now set immediately after the failure of a
1340 verification in an ACL, and so is available in subsequent modifiers. In
1341 particular, the message can be preserved by coding like this:
1343 warn !verify = sender
1344 set acl_m0 = $acl_verify_message
1346 Previously, $acl_verify_message was set only while expanding "message"
1347 and "log_message" when a very denied access.
1349 PH/28 Modified OS/os.c-Linux with
1351 -#ifndef OS_LOAD_AVERAGE
1352 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1354 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1355 nomenclature these days.)
1357 PH/29 Installed patch from the Sieve maintainer that adds the options
1358 sieve_useraddress and sieve_subaddress to the redirect router.
1360 PH/30 In these circumstances:
1361 . Two addresses routed to the same list of hosts;
1362 . First host does not offer TLS;
1363 . First host accepts first address;
1364 . First host gives temporary error to second address;
1365 . Second host offers TLS and a TLS session is established;
1366 . Second host accepts second address.
1367 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1368 and peerdn, if requested) that were in fact used only for the second
1371 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1372 attention to any local part prefix or suffix that was matched by the
1373 router that accepted the address. It now behaves in the same way as it
1374 does for delivery: the affixes are removed from the local part unless
1375 rcpt_include_affixes is set on the transport.
1377 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1378 timeout during the DATA phase of an incoming message.
1380 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1381 applied a patch sanctioned by the Sieve maintainer.
1383 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1384 the uid or gid is negative. A case of a negative gid caused this to be
1385 noticed. The fix allows for either to be negative.
1387 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1388 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1391 PH/36 The change PH/12 above was broken. Fixed it.
1393 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1394 the grounds that routing the same address twice would always produce the
1395 same answer. This might have been true once, but it is certainly no
1396 longer true now. Routing a child address may depend on the previous
1397 routing that produced that child. Some complicated redirection strategies
1398 went wrong when messages had multiple recipients, and made Exim's
1399 behaviour dependent on the order in which the addresses were given.
1401 I have moved the duplicate checking until after the routing is complete.
1402 Exim scans the addresses that are assigned to local and remote
1403 transports, and removes any duplicates. This means that more work will be
1404 done, as duplicates will always all be routed, but duplicates are
1405 presumably rare, so I don't expect this is of any significance.
1407 For deliveries to pipes, files, and autoreplies, the duplicate checking
1408 still happens during the routing process, since they are not going to be
1411 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1412 It corrects a timeout issue with spamd. This is Ian's comment: "The
1413 background is that sometimes spamd either never reads data from a
1414 connection it has accepted, or it never writes response data. The exiscan
1415 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1416 blindly assumes that writes won't block so it may never time out."
1418 PH/39 Allow G after quota size as well as K and M.
1420 PH/40 The value set for $authenticated_id in an authenticator may not contain
1421 binary zeroes or newlines because the value is written to log lines and
1422 to spool files. There was no check on this. Now the value is run through
1423 the string_printing() function so that such characters are converted to
1424 printable escape sequences.
1426 PH/41 $message_linecount is a new variable that contains the total number of
1427 lines in the message. Compare $body_linecount, which is the count for the
1430 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1431 bounce and delay warning messages, except in certain special cases, which
1434 (a) An SMTP error message from a remote host;
1435 (b) A message specified in a :fail: redirection;
1436 (c) A message specified in a "fail" command in a system filter;
1437 (d) A message specified in a FAIL return from the queryprogram router;
1438 (e) A message specified by the cannot_route_message router option.
1440 In these cases only, Exim does include the error details in bounce and
1441 warning messages. There are also a few cases where bland messages such
1442 as "unrouteable address" or "local delivery error" are given.
1444 PH/43 $value is now also set for the "else" part of a ${run expansion.
1446 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1447 being worked on, but at least Exim now implements the latest version to
1450 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1451 process to complete was treated as a delivery failure, a timeout while
1452 writing the message to the pipe was logged, but erroneously treated as a
1453 successful delivery. Such timeouts include transport filter timeouts. For
1454 consistency with the overall process timeout, these timeouts are now
1455 treated as errors, giving rise to delivery failures by default. However,
1456 there is now a new Boolean option for the pipe transport called
1457 timeout_defer, which, if set TRUE, converts the failures into defers for
1458 both kinds of timeout. A transport filter timeout is now identified in
1461 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1462 systems where "make" and "gmake" are different, calling "gmake" at top
1463 level broke things. I've arranged for the value of $(MAKE) to be passed
1464 from the Makefile to this script so that it can call the same version of
1468 A note about Exim versions 4.44 and 4.50
1469 ----------------------------------------
1471 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1472 changes of various kinds. As a consequence, a big documentation update was
1473 needed. This delayed the release for rather longer than seemed good, especially
1474 in the light of a couple of (minor) security issues. Therefore, the changes
1475 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1476 release. So 4.44 and 4.50 are in effect two different branches that both start
1479 I have left the 4.50 change log unchanged; it contains all the changes since
1480 4.43. The change log for 4.44 is below; many of its items are identical to
1481 those for 4.50. This seems to be the most sensible way to preserve the
1482 historical information.
1488 1. Minor wording change to the doc/README.SIEVE file.
1490 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1491 computation of the current number of files was incorrect.
1493 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1494 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1497 4. Give more explanation in the error message when the command for a transport
1498 filter fails to execute.
1500 5. There are several places where Exim runs a non-Exim command in a
1501 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1502 was being done only for the command run by the queryprogram router. It is
1503 now done for all such subprocesses. The other cases are: ${run, transport
1504 filters, and the commands run by the lmtp and pipe transports.
1506 6. Added CONFIGURE_GROUP build-time option.
1508 7. Some older OS have a limit of 256 on the maximum number of file
1509 descriptors. Exim was using setrlimit() to set 1000 as a large value
1510 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1511 systems. I've change it so that if it can't get 1000, it tries for 256.
1513 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1514 was an oversight, and furthermore, ever since the addition of extra
1515 controls (e.g. 4.43/32), the checks on when to allow different forms of
1516 "control" were broken. There should now be diagnostics for all cases when a
1517 control that does not make sense is encountered.
1519 9. Added the /retain_sender option to "control=submission".
1521 10. $recipients is now available in the predata ACL (oversight).
1523 11. Tidy the search cache before the fork to do a delivery from a message
1524 received from the command line. Otherwise the child will trigger a lookup
1525 failure and thereby defer the delivery if it tries to use (for example) a
1526 cached ldap connection that the parent has called unbind on.
1528 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1529 of $address_data from the recipient verification was clobbered by the
1530 sender verification.
1532 13. The value of address_data from a sender verification is now available in
1533 $sender_address_data in subsequent conditions in the ACL statement.
1535 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1537 15. Added a new option "connect=<time>" to callout options, to set a different
1540 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1541 was its contents. (It was OK if the option was not defined at all.)
1543 17. A "Completed" log line is now written for messages that are removed from
1544 the spool by the -Mrm option.
1546 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1547 information about exactly what failed.
1549 19. Added -dd to debug only the daemon process.
1551 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1552 handles timeouts, both on the server side and network timeouts. Renamed the
1553 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1555 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1556 It is now set to "smtps".
1558 22. $host_address is now set to the target address during the checking of
1559 ignore_target_hosts.
1561 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1562 being passed; this would have caused $sender_host_name to have been used if
1563 matching the list had actually called for a host name (not very likely,
1564 since this list is usually IP addresses). A host name is now passed as
1567 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1568 code that specifies a non-restarting handler (typically sigaction() in
1569 modern systems) in an attempt to fix a rare and obscure crash bug.
1571 25. Narrowed the window for a race in the daemon that could cause it to ignore
1572 SIGCHLD signals. This is not a major problem, because they are used only to
1573 wake it up if nothing else does.
1575 26. A malformed maildirsize file could cause Exim to calculate negative values
1576 for the mailbox size or file count. Odd effects could occur as a result.
1577 The maildirsize information is now recalculated if the size or filecount
1580 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1581 support for a long time. Removed HAVE_SYS_VFS_H.
1583 28. Installed the latest version of exipick from John Jetmore.
1585 29. In an address list, if the pattern was not a regular expression, an empty
1586 subject address (from a bounce message) matched only if the pattern was an
1587 empty string. Non-empty patterns were not even tested. This was the wrong
1588 because it is perfectly reasonable to use an empty address as part of a
1589 database query. An empty address is now tested by patterns that are
1590 lookups. However, all the other forms of pattern expect the subject to
1591 contain a local part and a domain, and therefore, for them, an empty
1592 address still always fails if the pattern is not itself empty.
1594 30. Exim went into a mad DNS loop when attempting to do a callout where the
1595 host was specified on an smtp transport, and looking it up yielded more
1596 than one IP address.
1598 31. Re-factored the code for checking spool and log partition space into a
1599 function that finds that data and another that does the check. The former
1600 is then used to implement four new variables: $spool_space, $log_space,
1601 $spool_inodes, and $log_inodes.
1603 32. The RFC2047 encoding function was originally intended for short strings
1604 such as real names; it was not keeping to the 75-character limit for
1605 encoded words that the RFC imposes. It now respects the limit, and
1606 generates multiple encoded words if necessary. To be on the safe side, I
1607 have increased the buffer size for the ${rfc2047: expansion operator from
1610 33. It is now permitted to omit both strings after an "if" condition; if the
1611 condition is true, the result is "true". As before, when the second string
1612 is omitted, a false condition yields an empty string. This makes it less
1613 cumbersome to write custom ACL and router conditions.
1615 34. Failure to deliver a bounce message always caused it to be frozen, even if
1616 there was an errors_to setting on the router. The errors_to setting is now
1619 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1620 canonical form (fully expanded) before being placed in
1621 $sender_host_address.
1623 36. The table in the code that translates DNS record types into text (T_A to
1624 "A" for instance) was missing entries for NS and CNAME. It is just possible
1625 that this could have caused confusion if both these types were looked up
1626 for the same domain, because the text type is used as part of Exim's
1627 per-process caching. But the chance of anyone hitting this buglet seems
1630 37. The dnsdb lookup has been extended in a number of ways.
1632 (1) There is a new type, "zns", which walks up the domain tree until it
1633 finds some nameserver records. It should be used with care.
1635 (2) There is a new type, "mxh", which is like "mx" except that it returns
1636 just the host names, not the priorities.
1638 (3) It is now possible to give a list of domains (or IP addresses) to be
1639 looked up. The behaviour when one of the lookups defers can be
1640 controlled by a keyword.
1642 (4) It is now possible to specify the separator character for use when
1643 multiple records are returned.
1645 38. The dnslists ACL condition has been extended: it is now possible to supply
1646 a list of IP addresses and/or domains to be looked up in a particular DNS
1649 39. Added log_selector=+queue_time_overall.
1651 40. When running the queue in the test harness, wait just a tad after forking a
1652 delivery process, to get repeatability of debugging output.
1654 41. Include certificate and key file names in error message when GnuTLS fails
1655 to set them up, because the GnuTLS error message doesn't include the name
1656 of the failing file when there is a problem reading it.
1658 42. Allow both -bf and -bF in the same test run.
1660 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1662 44. The "Exiscan patch" is now merged into the mainline Exim source.
1664 45. Sometimes the final signoff response after QUIT could fail to get
1665 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1666 < 0 before doing a fflush(). This bug looks as though it goes back to the
1667 introduction of TLS in release 3.20, but "sometimes" must have been rare
1668 because the tests only now provoked it.
1670 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1671 (this can affect the format of dates).
1673 47. exim_tidydb, when checking for the continued existence of a message for
1674 which it has found a message-specific retry record, was not finding
1675 messages that were in split spool directories. Consequently, it was
1676 deleting retry records that should have stayed in existence.
1678 48. Steve fixed some bugs in eximstats.
1680 49. The SPA authentication driver was not abandoning authentication and moving
1681 on to the next authenticator when an expansion was forced to fail,
1682 contradicting the general specification for all authenticators. Instead it
1683 was generating a temporary error. It now behaves as specified.
1685 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1686 (the order specifies the preference for clients). The order is now AES256,
1687 AES128, 3DES, ARCFOUR128.
1689 51. Small patch to Sieve code - explicitly set From: when generating an
1692 52. Exim crashed if a remote delivery caused a very long error message to be
1693 recorded - for instance if somebody sent an entire SpamAssassin report back
1694 as a large number of 550 error lines. This bug was coincidentally fixed by
1695 increasing the size of one of Exim's internal buffers (big_buffer) that
1696 happened as part of the Exiscan merge. However, to be on the safe side, I
1697 have made the code more robust (and fixed the comments that describe what
1700 53. Now that there can be additional text after "Completed" in log lines (if
1701 the queue_time_overall log selector is set), a one-byte patch to exigrep
1702 was needed to allow it to recognize "Completed" as not the last thing in
1705 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1706 patch that reportedly fixes this has been added. I am not expert enough to
1707 create a test for it. This is what the patch creator wrote:
1709 "I found a little strange behaviour of ldap code when working with
1710 Windows 2003 AD Domain, where users was placed in more than one
1711 Organization Units. When I tried to give exim partial DN, the exit code
1712 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1713 But simultaneously result of request was absolutely normal ldap result,
1714 so I produce this patch..."
1716 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1717 REFERENCE, so I have modified the code to exclude the patch when that macro
1720 55. Some experimental protocols are using DNS PTR records for new purposes. The
1721 keys for these records are domain names, not reversed IP addresses. The
1722 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1723 leaves it alone. Component reversal etc. now happens only for IP addresses.
1726 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1728 57. Double the size of the debug message buffer (to 2048) so that more of very
1729 long debug lines gets shown.
1731 58. The exicyclog utility now does better if the number of log files to keep
1732 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1734 59. Two changes related to the smtp_active_hostname option:
1736 (1) $smtp_active_hostname is now available as a variable.
1737 (2) The default for smtp_banner uses $smtp_active_hostname instead
1738 of $primary_hostname.
1740 60. The host_aton() function is supposed to be passed a string that is known
1741 to be a valid IP address. However, in the case of IPv6 addresses, it was
1742 not checking this. This is a hostage to fortune. Exim now panics and dies
1743 if the condition is not met. A case was found where this could be provoked
1744 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1745 components; fortuitously, this particular loophole had already been fixed
1746 by change 4.50/55 above.
1748 If there are any other similar loopholes, the new check in host_aton()
1749 itself should stop them being exploited. The report I received stated that
1750 data on the command line could provoke the exploit when Exim was running as
1751 exim, but did not say which command line option was involved. All I could
1752 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1753 running as the user.
1756 61. There was a buffer overflow vulnerability in the SPA authentication code
1757 (which came originally from the Samba project). I have added a test to the
1758 spa_base64_to_bits() function which I hope fixes it.
1761 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1762 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1764 63. The daemon start-up calls getloadavg() while still root for those OS that
1765 need the first call to be done as root, but it missed one case: when
1766 deliver_queue_load_max is set with deliver_drop_privilege. This is
1767 necessary for the benefit of the queue runner, because there is no re-exec
1768 when deliver_drop_privilege is set.
1770 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1771 This has been fixed.
1773 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1774 was in use, was not putting the data itself into the right store pool;
1775 consequently, it could be overwritten for a subsequent message in the same
1776 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1779 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1781 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1782 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1783 the function were treating the return as a boolean value, which happened to
1784 work because 0=false and not-0=true, but is not correct code.
1786 68. The host_aton() function was not handling scoped IPv6 addresses (those
1787 with, for example, "%eth0" on the end) correctly.
1789 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1790 negated items (that is, ~something) in unsigned ints. Some compilers
1791 apparently mutter when there is no cast.
1793 70. If an address verification called from an ACL failed, and did not produce a
1794 user-specific message (i.e. there was only a "system" message), nothing was
1795 put in $acl_verify_message. In this situation, it now puts the system
1798 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1799 session; change 4.31/43 added the unwanted input to the log line - except
1800 that it did not do this in the start of session case. It now does.
1802 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1803 This is wrong and can cause the other end to generate a synchronization
1804 error if it is another Exim or anything else that does the synchronization
1805 check. A QUIT command is no longer sent after a timeout.
1807 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1808 during host lookups.
1810 74. The defer_ok option of callout verification was not working if it was used
1811 when verifying addresses in header lines, that is, for this case:
1813 verify = header_sender/callout=defer_ok
1815 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1816 those file descriptors could be used for SMTP connections. If anything
1817 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1818 could be sent to the SMTP client, causing chaos. The daemon now opens
1819 stdin, stdout, and stderr to /dev/null when it puts itself into the
1822 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1823 log by default. The user can override this with suitable Perl magic.
1825 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1826 the log message when discard triggers, was not working for the DATA ACL or
1827 for the non-SMTP ACL.
1829 78. Error message wording change in sieve.c.
1831 79. If smtp_accept_max_per_host was set, the number of connections could be
1832 restricted to fewer than expected, because the daemon was trying to set up
1833 a new connection before checking whether the processes handling previous
1834 connections had finished. The check for completed processes is now done
1835 earlier. On busy systems, this bug wouldn't be noticed because something
1836 else would have woken the daemon, and it would have reaped the completed
1839 80. If a message was submitted locally by a user whose login name contained one
1840 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1841 It caused a spool format error. I have fixed the spool reading code. A
1842 related problem was that the "from" clause in the Received: line became
1843 illegal because of the space(s). It is now covered by ${quote_local_part.
1845 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1848 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1850 83. Patch from Sieve maintainer to fix unterminated string problem in
1851 "vacation" handling.
1853 84. Some minor changes to the Linux configuration files to help with other
1854 OS variants using glibc.
1856 85. One more patch for Sieve to update vacation handling to latest spec.
1859 ----------------------------------------------------
1860 See the note above about the 4.44 and 4.50 releases.
1861 ----------------------------------------------------
1867 1. Change 4.43/35 introduced a bug that caused file counts to be
1868 incorrectly computed when quota_filecount was set in an appendfile
1871 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1872 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1875 3. Give more explanation in the error message when the command for a transport
1876 filter fails to execute.
1878 4. There are several places where Exim runs a non-Exim command in a
1879 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1880 was being done only for the command run by the queryprogram router. It is
1881 now done for all such subprocesses. The other cases are: ${run, transport
1882 filters, and the commands run by the lmtp and pipe transports.
1884 5. Some older OS have a limit of 256 on the maximum number of file
1885 descriptors. Exim was using setrlimit() to set 1000 as a large value
1886 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1887 systems. I've change it so that if it can't get 1000, it tries for 256.
1889 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1890 was an oversight, and furthermore, ever since the addition of extra
1891 controls (e.g. 4.43/32), the checks on when to allow different forms of
1892 "control" were broken. There should now be diagnostics for all cases when a
1893 control that does not make sense is encountered.
1895 7. $recipients is now available in the predata ACL (oversight).
1897 8. Tidy the search cache before the fork to do a delivery from a message
1898 received from the command line. Otherwise the child will trigger a lookup
1899 failure and thereby defer the delivery if it tries to use (for example) a
1900 cached ldap connection that the parent has called unbind on.
1902 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1903 of $address_data from the recipient verification was clobbered by the
1904 sender verification.
1906 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1907 was its contents. (It was OK if the option was not defined at all.)
1909 11. A "Completed" log line is now written for messages that are removed from
1910 the spool by the -Mrm option.
1912 12. $host_address is now set to the target address during the checking of
1913 ignore_target_hosts.
1915 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1916 being passed; this would have caused $sender_host_name to have been used if
1917 matching the list had actually called for a host name (not very likely,
1918 since this list is usually IP addresses). A host name is now passed as
1921 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1922 code that specifies a non-restarting handler (typically sigaction() in
1923 modern systems) in an attempt to fix a rare and obscure crash bug.
1925 15. Narrowed the window for a race in the daemon that could cause it to ignore
1926 SIGCHLD signals. This is not a major problem, because they are used only to
1927 wake it up if nothing else does.
1929 16. A malformed maildirsize file could cause Exim to calculate negative values
1930 for the mailbox size or file count. Odd effects could occur as a result.
1931 The maildirsize information is now recalculated if the size or filecount
1934 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1935 support for a long time. Removed HAVE_SYS_VFS_H.
1937 18. Updated exipick to current release from John Jetmore.
1939 19. Allow an empty sender to be matched against a lookup in an address list.
1940 Previously the only cases considered were a regular expression, or an
1943 20. Exim went into a mad DNS lookup loop when doing a callout where the
1944 host was specified on the transport, if the DNS lookup yielded more than
1947 21. The RFC2047 encoding function was originally intended for short strings
1948 such as real names; it was not keeping to the 75-character limit for
1949 encoded words that the RFC imposes. It now respects the limit, and
1950 generates multiple encoded words if necessary. To be on the safe side, I
1951 have increased the buffer size for the ${rfc2047: expansion operator from
1954 22. Failure to deliver a bounce message always caused it to be frozen, even if
1955 there was an errors_to setting on the router. The errors_to setting is now
1958 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1959 canonical form (fully expanded) before being placed in
1960 $sender_host_address.
1962 24. Updated eximstats to version 1.33
1964 25. Include certificate and key file names in error message when GnuTLS fails
1965 to set them up, because the GnuTLS error message doesn't include the name
1966 of the failing file when there is a problem reading it.
1968 26. Expand error message when OpenSSL has problems setting up cert/key files.
1971 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
1972 (this can affect the format of dates).
1974 28. exim_tidydb, when checking for the continued existence of a message for
1975 which it has found a message-specific retry record, was not finding
1976 messages that were in split spool directories. Consequently, it was
1977 deleting retry records that should have stayed in existence.
1979 29. eximstats updated to version 1.35
1980 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
1981 1.35 - bugfix such that pie charts by volume are generated correctly
1983 30. The SPA authentication driver was not abandoning authentication and moving
1984 on to the next authenticator when an expansion was forced to fail,
1985 contradicting the general specification for all authenticators. Instead it
1986 was generating a temporary error. It now behaves as specified.
1988 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
1989 (the order specifies the preference for clients). The order is now AES256,
1990 AES128, 3DES, ARCFOUR128.
1992 31. Small patch to Sieve code - explicitly set From: when generating an
1995 32. Exim crashed if a remote delivery caused a very long error message to be
1996 recorded - for instance if somebody sent an entire SpamAssassin report back
1997 as a large number of 550 error lines. This bug was coincidentally fixed by
1998 increasing the size of one of Exim's internal buffers (big_buffer) that
1999 happened as part of the Exiscan merge. However, to be on the safe side, I
2000 have made the code more robust (and fixed the comments that describe what
2003 33. Some experimental protocols are using DNS PTR records for new purposes. The
2004 keys for these records are domain names, not reversed IP addresses. The
2005 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2006 leaves it alone. Component reversal etc. now happens only for IP addresses.
2009 34. The host_aton() function is supposed to be passed a string that is known
2010 to be a valid IP address. However, in the case of IPv6 addresses, it was
2011 not checking this. This is a hostage to fortune. Exim now panics and dies
2012 if the condition is not met. A case was found where this could be provoked
2013 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2014 components; fortuitously, this particular loophole had already been fixed
2015 by change 4.50/55 or 4.44/33 above.
2017 If there are any other similar loopholes, the new check in host_aton()
2018 itself should stop them being exploited. The report I received stated that
2019 data on the command line could provoke the exploit when Exim was running as
2020 exim, but did not say which command line option was involved. All I could
2021 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2022 running as the user.
2025 35. There was a buffer overflow vulnerability in the SPA authentication code
2026 (which came originally from the Samba project). I have added a test to the
2027 spa_base64_to_bits() function which I hope fixes it.
2030 36. The daemon start-up calls getloadavg() while still root for those OS that
2031 need the first call to be done as root, but it missed one case: when
2032 deliver_queue_load_max is set with deliver_drop_privilege. This is
2033 necessary for the benefit of the queue runner, because there is no re-exec
2034 when deliver_drop_privilege is set.
2036 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2037 was in use, was not putting the data itself into the right store pool;
2038 consequently, it could be overwritten for a subsequent message in the same
2039 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2042 38. Sometimes the final signoff response after QUIT could fail to get
2043 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2044 < 0 before doing a fflush(). This bug looks as though it goes back to the
2045 introduction of TLS in release 3.20, but "sometimes" must have been rare
2046 because the tests only now provoked it.
2052 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2053 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2054 Now it returns an integer. A number of calls were still expecting a T/F
2055 return. Fortuitously, in all cases, the tests worked in OK situations,
2056 which is the norm. However, things would have gone wrong on any write
2057 failures on the smtp file descriptor. This function is used when sending
2058 messages over SMTP and also when doing verify callouts.
2060 2. When Exim is called to do synchronous delivery of a locally submitted
2061 message (the -odf or -odi options), it no longer closes stderr before doing
2064 3. Implemented the mua_wrapper option.
2066 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2068 5. Implemented the functions header_remove(), header_testname(),
2069 header_add_at_position(), and receive_remove_recipient(), and exported them
2072 6. If an ACL "warn" statement specified the addition of headers, Exim already
2073 inserted X-ACL-Warn: at the start if there was no header name. However, it
2074 was not making this test for the second and subsequent header lines if
2075 there were newlines in the string. This meant that an invalid header could
2076 be inserted if Exim was badly configured.
2078 7. Allow an ACL "warn" statement to add header lines at the start or after all
2079 the Received: headers, as well as at the end.
2081 8. Added the rcpt_4xx retry error code.
2083 9. Added postmaster_mailfrom=xxx to callout verification option.
2085 10. Added mailfrom=xxxx to the callout verification option, for verify=
2088 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2089 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2091 12. Inserted some casts to stop certain compilers warning when using pointer
2092 differences as field lengths or precisions in printf-type calls (mostly
2093 affecting debugging statements).
2095 13. Added optional readline() support for -be (dynamically loaded).
2097 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2098 same clock tick as a message's arrival, so that its received time was the
2099 same as the "first fail" time on the retry record, and that message
2100 remained on the queue past the ultimate address timeout, every queue runner
2101 would try a delivery (because it was past the ultimate address timeout) but
2102 after another failure, the ultimate address timeout, which should have then
2103 bounced the address, did not kick in. This was a "< instead of <=" error;
2104 in most cases the first failure would have been in the next clock tick
2105 after the received time, and all would be well.
2107 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2108 being recognized when the domain list was tested by the match_domain
2109 condition in an expansion string.
2111 16. Added the ${str2b64: operator.
2113 17. Exim was always calling setrlimit() to set a large limit for the number of
2114 processes, without checking whether the existing limit was already
2115 adequate. (It did check for the limit on file descriptors.) Furthermore,
2116 errors from getrlimit() and setrlimit() were being ignored. Now they are
2117 logged to the main and panic logs, but Exim does carry on, to try to do its
2118 job under whatever limits there are.
2120 18. Imported PCRE 5.0.
2122 19. Trivial typo in log message " temporarily refused connection" (the leading
2125 20. If the log selector return_path_on_delivery was set and an address was
2126 redirected to /dev/null, the delivery process crashed because it assumed
2127 that a return path would always be set for a "successful" delivery. In this
2128 case, the whole delivery is bypassed as an optimization, and therefore no
2131 21. Internal re-arrangement: the function for sending a challenge and reading
2132 a response while authentication was assuming a zero-terminated challenge
2133 string. It's now changed to take a pointer and a length, to allow for
2134 binary data in such strings.
2136 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2138 23. Exim was not respecting finduser_retries when seeking the login of the
2139 uid under which it was called; it was always trying 10 times. (The default
2140 setting of finduser_retries is zero.) Also, it was sleeping after the final
2141 failure, which is pointless.
2143 24. Implemented tls_on_connect_ports.
2145 25. Implemented acl_smtp_predata.
2147 26. If the domain in control=submission is set empty, Exim assumes that the
2148 authenticated id is a complete email address when it generates From: or
2149 Sender: header lines.
2151 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2152 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2153 chown and chgrp in /bin and hostname in /usr/bin.
2155 28. Exim was keeping the "process log" file open after each use, just as it
2156 does for the main log. This opens the possibility of it remaining open for
2157 long periods when the USR1 signal hits a daemon. Occasional processlog
2158 errors were reported, that could have been caused by this. Anyway, it seems
2159 much more sensible not to leave this file open at all, so that is what now
2162 29. The long-running daemon process does not normally write to the log once it
2163 has entered its main loop, and it closes the log before doing so. This is
2164 so that log files can straightforwardly be renamed and moved. However,
2165 there are a couple of unusual error situations where the daemon does write
2166 log entries, and I had neglected to close the log afterwards.
2168 30. The text of an SMTP error response that was received during a remote
2169 delivery was being truncated at 512 bytes. This is too short for some of
2170 the long messages that one sometimes sees. I've increased the limit to
2173 31. It is now possible to make retry rules that apply only when a message has a
2174 specific sender, in particular, an empty sender.
2176 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2177 it possible to be selective about when SMTP synchronization is enforced.
2179 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2181 32. Implemented hosts_connection_nolog.
2183 33. Added an ACL for QUIT.
2185 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2188 35. Added mailbox_size and mailbox_filecount to appendfile.
2190 36. Added control = no_multiline_responses to ACLs.
2192 37. There was a bug in the logic of the code that waits for the clock to tick
2193 in the case where the clock went backwards by a substantial amount such
2194 that the microsecond fraction of "now" was more than the microsecond
2195 fraction of "then" (but the whole seconds number was less).
2197 38. Added support for the libradius Radius client library this is found on
2198 FreeBSD (previously only the radiusclient library was supported).
2204 1. When certain lookups returned multiple values in the form name=value, the
2205 quoting of the values was not always being done properly. Specifically:
2206 (a) If the value started with a double quote, but contained no whitespace,
2208 (b) If the value contained whitespace other than a space character (i.e.
2209 tabs or newlines or carriage returns) it was not quoted.
2210 This fix has been applied to the mysql and pgsql lookups by writing a
2211 separate quoting function and calling it from the lookup code. The fix
2212 should probably also be applied to nisplus, ibase and oracle lookups, but
2213 since I cannot test any of those, I have not disturbed their existing code.
2215 2. A hit in the callout cache for a specific address caused a log line with no
2216 reason for rejecting RCPT. Now it says "Previous (cached) callout
2217 verification failure".
2219 3. There was an off-by-one bug in the queryprogram router. An over-long
2220 return line was truncated at 256 instead of 255 characters, thereby
2221 overflowing its buffer with the terminating zero. As well as fixing this, I
2222 have increased the buffer size to 1024 (and made a note to document this).
2224 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2225 when Exim is waiting for an SMTP response from a remote server, Exim
2226 restarts its select() call on the socket, thereby resetting its timeout.
2227 This is not a problem when such interrupts are rare. Somebody set up a cron
2228 job to run exiwhat every 2 minutes, which is less than the normal select()
2229 timeout (5 or 10 minutes). This meant that the select() timeout never
2230 kicked in because it was always reset. I have fixed this by comparing the
2231 time when an interrupt arrives with the time at the start of the first call
2232 to select(). If more time than the timeout has elapsed, the interrupt is
2233 treated as a timeout.
2235 5. Some internal re-factoring in preparation for the addition of Sieve
2236 extensions (by MH). In particular, the "personal" test is moved to a
2237 separate function, and given an option for scanning Cc: and Bcc: (which is
2238 not set for Exim filters).
2240 6. When Exim created an email address using the login of the caller as the
2241 local part (e.g. when creating a From: or Sender: header line), it was not
2242 quoting the local part when it contained special characters such as @.
2244 7. Installed new OpenBSD configuration files.
2246 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2247 try to make them clearer.
2249 9. Callout options, other than the timeout value, were being ignored when
2250 verifying sender addresses in header lines. For example, when using
2252 verify = header_sender/callout=no_cache
2254 the cache was (incorrectly) being used.
2256 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2257 only the Cygwin environment.
2259 11. When return_path_on_delivery was set as a log selector, if different remote
2260 addresses in the same message used different return paths and parallel
2261 remote delivery occurred, the wrong values would sometimes be logged.
2262 (Whenever a remote delivery process finished, the return path value from
2263 the most recently started remote delivery process was logged.)
2265 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2266 lines when AUTH and/or TLS are in use. This is the "received protocol"
2267 field. Exim used to use "asmtp" for authenticated SMTP, without any
2268 indication (in the protocol name) for TLS use. Now it follows the RFC and
2269 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2270 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2271 names appear in log lines as well as in Received: header lines.
2273 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2274 extensions, and comparison tests, and to fix some bugs.
2276 14. Changes to the "personal" filter test:
2278 (1) The test was buggy in that it was just doing the equivalent of
2279 "contains" tests on header lines. For example, if a user's address was
2280 anne@some.where, the "personal" test would incorrectly be true for
2282 To: susanne@some.where
2284 This test is now done by extracting each address from the header in turn,
2285 and checking the entire address. Other tests that are part of "personal"
2286 are now done using regular expressions (for example, to check local parts
2287 of addresses in From: header lines).
2289 (2) The list of non-personal local parts in From: addresses has been
2290 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2291 taken from the Sieve specification recommendations.
2293 (3) If the message contains any header line starting with "List-" it is
2294 treated as non-personal.
2296 (4) The test for "circular" in the Subject: header line has been removed
2297 because it now seems ill-conceived.
2299 15. Minor typos in src/EDITME comments corrected.
2301 16. Installed latest exipick from John Jetmore.
2303 17. If headers_add on a router specified a text string that was too long for
2304 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2305 of string_sprintf() is now avoided.
2307 18. $message_body_size was not set (it was always zero) when running the DATA
2308 ACL and the local_scan() function.
2310 19. For the "mail" command in an Exim filter, no default was being set for
2311 the once_repeat time, causing a random time value to be used if "once" was
2312 specified. (If the value happened to be <= 0, no repeat happened.) The
2313 default is now 0s, meaning "never repeat". The "vacation" command was OK
2314 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2315 (I found it when inspecting the code).
2317 20. There is now an overall timeout for performing a callout verification. It
2318 defaults to 4 times the callout timeout, which applies to individual SMTP
2319 commands during the callout. The overall timeout applies when there is more
2320 than one host that can be tried. The timeout is checked before trying the
2321 next host. This prevents very long delays if there are a large number of
2322 hosts and all are timing out (e.g. when the network connections are timing
2323 out). The value of the overall timeout can be changed by specifying an
2324 additional sub-option for "callout", called "maxwait". For example:
2326 verify = sender/callout=5s,maxwait=20s
2328 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2329 to the end before writing, but this should make it even safer).
2331 22. Exim was forgetting that it had advertised PIPELINING for the second and
2332 subsequent messages on an SMTP connection. It was also not resetting its
2333 memory on STARTTLS and an internal HELO.
2335 23. When Exim logs an SMTP synchronization error within a session, it now
2336 records whether PIPELINING has been advertised or not.
2338 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2339 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2340 rather than long int.
2342 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2344 26. Added the never_mail option to autoreply.
2350 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2351 crash if the getsockname() call failed; this can happen if a connection is
2352 closed very soon after it is established. The problem was simply in the
2353 order in which certain operations were done, causing Exim to try to write
2354 to the SMTP stream before it had set up the file descriptor. The bug has
2355 been fixed by making things happen in the correct order.
2361 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2362 before the connection was closed, thus losing the rejection response.
2364 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2365 some early Solaris releases, but causes trouble in current releases where
2366 socklen_t is defined.
2368 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2371 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2374 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2375 that releasing the top of it at the end releases what was used for sub-
2376 expansions (unless the block got too big). However, discard this block if
2377 the first thing is a variable or header, so that we can use its block when
2378 it is dynamic (useful for very large $message_headers, for example).
2380 6. Lookups now cache *every* query, not just the most recent. A new, separate
2381 store pool is used for this. It can be recovered when all lookup caches are
2382 flushed. Lookups now release memory at the end of their result strings.
2383 This has involved some general refactoring of the lookup sources.
2385 7. Some code has been added to the store_xxx() functions to reduce the amount
2386 of flapping under certain conditions.
2388 8. log_incoming_interface used to affect only the <= reception log lines. Now
2389 it causes the local interface and port to be added to several more SMTP log
2390 lines, for example "SMTP connection from", and rejection lines.
2392 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2394 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2396 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2397 could be overwritten at the end of the current message (or the start of a
2398 new message if it was set in a HELO ACL). The value is now preserved for
2399 the duration of the SMTP connection.
2401 12. If a transport had a headers_rewrite setting, and a matching header line
2402 contained an unqualified address, that address was qualified, even if it
2403 did not match any rewriting rules. The underlying bug was that the values
2404 of the flags that permit the existence of unqualified sender and recipient
2405 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2406 non-local messages, and by -bnq for local messages) were not being
2407 preserved with the message after it was received.
2409 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2410 "next input=" as part of the text comprising the host identity instead of
2411 the correct text. The code was using the same buffer for two different
2412 strings. However, depending on which order the printing function evaluated
2413 its arguments, the bug did not always show up. Under Linux, for example, my
2414 test suite worked just fine.
2416 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2417 doesn't work with some older versions of Perl. It has been changed to "my",
2418 which in any case is probably the better facility to use.
2420 15. A really picky compiler found some instances of statements for creating
2421 error messages that either had too many or two few arguments for the format
2424 16. The size of the buffer for calls to the DNS resolver has been increased
2425 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2426 for addresses that have a lot of PTR records. This alleviates a problem; it
2427 does not fully solve it.
2429 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2430 buffer now truncates the list and logs the incident, which is the same
2431 action as happens when Exim is looking up a host name and its aliases.
2432 Previously in this situation something unpredictable would happen;
2433 sometimes it was "internal error: store_reset failed".
2435 18. If a server dropped the connection unexpectedly when an Exim client was
2436 using GnuTLS and trying to read a response, the client delivery process
2437 crashed while trying to generate an error log message.
2439 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2440 string, for example:
2442 warn message = H1: something\nH2: something
2444 the text was added as a single header line from Exim's point of view
2445 though it ended up OK in the delivered message. However, searching for the
2446 second and subsequent header lines using $h_h2: did not work. This has been
2447 fixed. Similarly, if a system filter added multiple headers in this way,
2448 the routers could not see them.
2450 20. Expanded the error message when iplsearch is called with an invalid key to
2451 suggest using net-iplsearch in a host list.
2453 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2454 ACLs are no longer actually imposed (and a message to that effect is
2457 22. If a "gecos" field in a passwd entry contained escaped characters, in
2458 particular, if it contained a \" sequence, Exim got it wrong when building
2459 a From: or a Sender: header from that name. A second bug also caused
2460 incorrect handling when an unquoted " was present following a character
2461 that needed quoting.
2463 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2464 was not being matched caselessly.
2466 24. Arranged for all hyphens in the exim.8 source to be escaped with
2469 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2470 properly. Recipient callout cache records were still being keyed to include
2471 the sender, even when use_sender was set false. This led to far more
2472 callouts that were necessary. The sender is no longer included in the key
2473 when use_sender is false.
2475 26. Added "control = submission" modifier to ACLs.
2477 27. Added the ${base62d: operator to decode base 62 numbers.
2479 28. dnsdb lookups can now access SRV records.
2481 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2482 the configuration file.
2484 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2485 (-v) mode. This makes the output for a verbose queue run more intelligible.
2487 31. Added a use_postmaster feature to recipient callouts.
2489 32. Added the $body_zerocount variable, containing the number of binary zero
2490 bytes in the message body.
2492 33. The time of last modification of the "new" subdirectory is now used as the
2493 "mailbox time last read" when there is a quota error for a maildir
2496 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2498 35. Added +ignore_unknown as a special item in host lists.
2500 36. Code for decoding IPv6 addresses in host lists is now included, even if
2501 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2502 address was recognized as an IP address, but was then not correctly decoded
2503 into binary, causing unexpected and incorrect effects when compared with
2510 1. Very minor rewording of debugging text in manualroute to say "list of
2511 hosts" instead of "hostlist".
2513 2. If verify=header_syntax was set, and a header line with an unqualified
2514 address (no domain) and a large number of spaces between the end of the
2515 name and the colon was received, the reception process suffered a buffer
2516 overflow, and (when I tested it) crashed. This was caused by some obsolete
2517 code that should have been removed. The fix is to remove it!
2519 3. When running in the test harness, delay a bit after writing a bounce
2520 message to get a bit more predictability in the log output.
2522 4. Added a call to search_tidyup() just before forking a reception process. In
2523 theory, someone could use a lookup in the expansion of smtp_accept_max_
2524 per_host which, without the tidyup, could leave open a database connection.
2526 5. Added the variables $recipient_data and $sender_data which get set from a
2527 lookup success in an ACL "recipients" or "senders" condition, or a router
2528 "senders" option, similar to $domain_data and $local_part_data.
2530 6. Moved the writing of debug_print from before to after the "senders" test
2533 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2534 problems for message scanning, either using a data ACL, or using
2535 local_scan() because the Received: header was not generated till after they
2536 were called (in order to set the time as the time of reception completion).
2537 I have revised the way this works. The header is now generated after the
2538 body is received, but before the ACL or local_scan() are called. After they
2539 are run, the timestamp in the header is updated.
2545 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2546 before starting a queue runner without re-exec. This happened only when
2547 deliver_drop_privilege was set or when the Exim user was set to root. The
2548 effect of the bug was that timeouts during subsequent deliveries caused
2549 crashes instead of being properly handled. The handler is now left at its
2550 default (and expected) setting.
2552 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2553 message, again when deliver_drop_privilege is set or Exim is run as root.
2554 The bug described in (1) was not present in this case, but the tidying up
2555 of the other signals was missing. I have made the two cases consistent.
2557 3. The ignore_target_hosts setting on a manualroute router was being ignored
2558 for hosts that were looked up using the /MX notation.
2560 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2563 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2564 operated on the sender address. After changing the $sender_address to <>
2565 for the sender address verify, Exim was re-instated it as the original
2566 (before rewriting) address, but remembering that it had rewritten it, so it
2567 wasn't rewriting it again. This bug also had the effect of breaking the
2568 sender address verification caching when the sender address was rewritten.
2570 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2571 This has been changed so that if the ip literal address matches
2572 ignore_target_hosts, the router declines.
2574 7. Added expansion conditions match_domain, match_address, and match_local_
2575 part (NOT match_host).
2577 8. The placeholder for the Received: header didn't have a length field set.
2579 9. Added code to Exim itself and to exim_lock to test for a specific race
2580 condition that could lead to file corruption when using MBX delivery. The
2581 issue is with the lockfile that is created in /tmp. If this file is removed
2582 after a process has opened it but before that process has acquired a lock,
2583 there is the potential for a second process to recreate the file and also
2584 acquire a lock. This could lead to two Exim processes writing to the file
2585 at the same time. The added code performs the same test as UW imapd; it
2586 checks after acquiring the lock that its file descriptor still refers to
2587 the same named file.
2589 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2590 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2593 11. Added the smtp_active_hostname option. If used, this will typically be made
2594 to depend on the incoming interface address. Because $interface_address is
2595 not set up until the daemon has forked a reception process, error responses
2596 that can happen earlier (such as "too many connections") no longer contain
2599 12. If an expansion in a condition on a "warn" statement fails because a lookup
2600 defers, the "warn" statement is abandoned, and the next ACL statement is
2601 processed. Previously this caused the whole ACL to be aborted.
2603 13. Added the iplsearch lookup type.
2605 14. Added ident_timeout as a log selector.
2607 15. Added tls_certificate_verified as a log selector.
2609 16. Added a global option tls_require_ciphers (compare the smtp transport
2610 option of the same name). This controls incoming TLS connections.
2612 17. I finally figured out how to make tls_require_ciphers do a similar thing
2613 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2614 before starting the TLS session.
2616 18. Tabs are now shown as \t in -bP output.
2618 19. If the log selector return_path_on_delivery was set, Exim crashed when
2619 bouncing a message because it had too many Received: header lines.
2621 20. If two routers both had headers_remove settings, and the first one included
2622 a superfluous trailing colon, the final name in the first list and the
2623 first name in the second list were incorrectly joined into one item (with a
2624 colon in the middle).
2630 1. Added -C and -D options to the exinext utility, mainly to make it easier
2631 to include in the automated testing, but these could be helpful when
2632 multiple configurations are in use.
2634 2. The exinext utility was not formatting the output nicely when there was
2635 an alternate port involved in the retry record key, nor when there was a
2636 message id as well (for retries that were specific to a specific message
2637 and a specific host). It was also confused by IPv6 addresses, because of
2638 the additional colons they contain. I have fixed the IPv4 problem, and
2639 patched it up to do a reasonable job for IPv6.
2641 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2642 delivery, the log line now contains "pipelined" if PIPELINING was used.
2644 4. An SMTP transport process used to panic and die if the bind() call to set
2645 an explicit outgoing interface failed. This has been changed; it is now
2646 treated in the same way as a connect() failure.
2648 5. A reference to $sender_host_name in the part of a conditional expansion
2649 that was being skipped was still causing a DNS lookup. This no longer
2652 6. The def: expansion condition was not recognizing references to header lines
2653 that used bh_ and bheader_.
2655 7. Added the _cache feature to named lists.
2657 8. The code for checking quota_filecount in the appendfile transport was
2658 allowing one more file than it should have been.
2660 9. For compatibility with Sendmail, the command line option
2668 and sets the incoming protocol and host name (for trusted callers). The
2669 host name and its colon can be omitted when only the protocol is to be set.
2670 Note the Exim already has two private options, -pd and -ps, that refer to
2671 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2672 "s", but I don't think that's a major issue.
2674 10. A number of refactoring changes to the code, none of which should affect
2677 (a) The number of logging options was getting close to filling up the
2678 32-bit word that was used as a bit map. I have split them into two classes:
2679 those that are passed in the argument to log_write(), and those that are
2680 only ever tested independently outside of that function. These are now in
2681 separate 32-bit words, so there is plenty of room for expansion again.
2682 There is no change in the user interface or the logging behaviour.
2684 (b) When building, for example, log lines, the code previously used a
2685 macro that called string_cat() twice, in order to add two strings. This is
2686 not really sufficiently general. Furthermore, there was one instance where
2687 it was actually wrong because one of the argument was used twice, and in
2688 one call a function was used. (As it happened, calling the function twice
2689 did not affect the overall behaviour.) The macro has been replaced by a
2690 function that can join an arbitrary number of extra strings onto a growing
2693 (c) The code for expansion conditions now uses a table and a binary chop
2694 instead of a serial search (which was left over from when there were very
2695 few conditions). Also, it now recognizes conditions like "pam" even when
2696 the relevant support is not compiled in: a suitably worded error message is
2697 given if an attempt is made to use such a condition.
2699 11. Added ${time_interval:xxxxx}.
2701 12. A bug was causing one of the ddress fields not to be passed back correctly
2702 from remote delivery subprocesses. The field in question was not being
2703 subsequently used, so this caused to problems in practice.
2705 13. Added new log selectors queue_time and deliver_time.
2707 14. Might have fixed a bug in maildirsizefile handling that threw up
2708 "unexpected character" debug warnings, and recalculated the data
2709 unnecessarily. In any case, I expanded the warning message to give more
2712 15. Added the message "Restricted characters in address" to the statements in
2713 the default ACL that block characters like @ and % in local parts.
2715 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2716 Three changes have been made:
2718 (a) There was a serious bug; a negative response to MAIL caused the whole
2719 recipient domain to be cached as invalid, thereby blocking all messages
2720 to all local parts at the same domain, from all senders. This bug has
2721 been fixed. The domain is no longer cached after a negative response to
2722 MAIL if the sender used is not empty.
2724 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2727 (c) A new callout option, "use_sender" has been added for people who want
2728 the modified behaviour.
2734 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2737 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2738 indeed breaks things for older releases.
2740 3. Added additional logging to the case where there is a problem reading data
2741 from a filter that is running in a subprocess using a pipe, in order to
2742 try to track down a specific problem.
2744 4. Testing facility fudge: when running in the test harness and attempting
2745 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2746 getting "No route to host". Convert this to a timeout.
2748 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2751 6. Some OS don't have socklen_t but use size_t instead. This affects the
2752 fifth argument of getsockopt() amongst other things. This is now
2753 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2754 can be set for individual OS. I have set it for SunOS5, OSF1, and
2755 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2756 some earlier ones do not.
2758 7. Change 4.30/15 was not doing the test caselessly.
2760 8. The standard form for an IPv6 address literal was being rejected by address
2761 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2762 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2763 this, as well as the form without the "IPv6" on the front (but only when
2764 address literals are enabled, of course).
2766 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2768 10. Exim crashed if a message with an empty sender address specified by -f
2769 encountered a router with an errors_to setting. This could be provoked only
2770 by a command such as
2774 where an empty string was supplied; "<>" did not hit this bug.
2776 11. Installed PCRE release 4.5.
2778 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2779 remained set. It is now erased.
2781 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2782 times from message ids (which are base 36 rather than the normal 62).
2784 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2785 were being counted as actual protocol errors, and logged if the log
2786 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2787 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2788 and DATA following a set of rejected RCPTs do not count as protocol errors.
2789 In other words, Exim assumes they were pipelined, though this may not
2790 actually be the case. Of course, in all cases the client gets an
2791 appropriate error code.
2793 15. If a lookup fails in an ACL condition, a message about the failure may
2794 be available; it is used if testing the ACL cannot continue, because most
2795 such messages specify what the cause of the deferral is. However, some
2796 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2797 that caused an old message to be retained and used if a later statement
2798 caused a defer, replacing the real cause of the deferral.
2800 16. If an IP address had so many PTR records that the DNS lookup buffer
2801 was not large enough to hold them, Exim could crash while trying to process
2802 the truncated data. It now detects and logs this case.
2804 17. Further to 4.21/58, another change has been made: if (and only if) the
2805 first line of a message (the first header line) ends with CRLF, a bare LF
2806 in a subsequent header line has a space inserted after it, so as not to
2807 terminate the header.
2809 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2810 unnecessarily, used atoi() instead of strtol(), and didn't check the
2811 termination when getting file sizes from file names by regex.
2813 19. Completely re-implemented the support for maildirsize files, in the light
2814 of a number of problems with the previous contributed implementation
2815 (4.30/29). In particular:
2817 . If the quota is zero, the maildirsize file is maintained, but no quota is
2820 . If the maildir directory does not exist, it is created before any attempt
2821 to write a maildirsize file.
2823 . The quota value in the file is just a cache; if the quota is changed in
2824 the transport, the new value overrides.
2826 . A regular expression is available for excluding directories from the
2829 20. The autoreply transport checks the characters in options that define the
2830 message's headers; it allows continued headers, but it was checking with
2831 isspace() after an embedded newline instead of explicitly looking for a
2834 21. If all the "regular" hosts to which an address was routed had passed their
2835 expiry times, and had not reached their retry times, the address was
2836 bounced, even if fallback hosts were defined. Now Exim should go on to try
2839 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2840 equivalent code in the SMTP transport. Some hosts send humungous responses
2841 to HELO/EHLO, more than 1024 it seems.
2843 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2844 gives compiler warnings in some environments. I've now done it "properly",
2847 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2848 (because of problems with the built-in one) was declared to return uschar *
2849 instead of char *, causing compiler failure.
2851 25. Fixed a file descriptor leak when processing alias/forward files.
2853 26. Fixed a minor format string issue in dbfn.c.
2855 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2857 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2858 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2861 29. When a sender address is verified, it is cached, to save repeating the test
2862 when there is more than one recipient in a message. However, when the
2863 verification involves a callout, it is possible for different callout
2864 options to be set for different recipients. It is too complicated to keep
2865 track of this in the cache, so now Exim always runs a verification when a
2866 callout is required, relying on the callout cache for the optimization.
2867 The overhead is duplication of the address routing, but this should not be
2870 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2871 to be verified with callout=postmaster, and the main callout worked but the
2872 postmaster check failed, the verification correctly failed. However, if a
2873 subsequent RCPT command asked for sender verification *without* the
2874 postmaster check, incorrect caching caused this verification also to fail,
2877 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2878 it was not caching the DNS options (qualify_single, search_parents) that
2879 were used when the lookup failed. A subsequent lookup with different
2880 options therefore always gave the same answer, though there were cases
2881 where it should not have. (Example: a "domains = !$mx_any" option on a
2882 dnslookup router: the "domains" option is always processed without any
2883 widening, but the router might have qualify_single set.) Now Exim uses the
2884 cached value only when the same options are set.
2886 32. Added John Jetmore's "exipick" utility to the distribution.
2888 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2889 than a timeout (e.g. a certificate is required, and is not provided), an
2890 Exim server now closes the connection immediately. Previously it waited for
2891 the client to close - but if the client is SSL, it seems that they each
2892 wait for each other, leading to a delay before one of them times out.
2894 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2895 maintained 0.8.x compatibility because I don't think many are using it, and
2896 it is clearly obsolete.
2898 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2901 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2902 client certificate was expired. A simple patch fixes this, though I don't
2903 understand the full logic of why the verify callback is called multiple
2906 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2907 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2908 which causes problems with some clients (such as the Certicom SSL Plus
2909 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2910 disables the coutermeasure allowing Eudora to connect."
2912 38. Exim was not checking that a write() to a log file succeeded. This could
2913 lead to Bad Things if a log got too big, in particular if it hit a file
2914 size limit. Exim now panics and dies if it cannot write to a log file, just
2915 as it does if it cannot open a log file.
2917 39. Modified OS/Makefile-Linux so that it now contains
2919 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2921 The two -D definitions ensure that Exim is compiled with large file
2922 support, which makes it possible to handle log files that are bigger than
2925 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2926 instance) a domain was checked against a named list that involved a lookup,
2927 causing $domain_data to be set, then another domain was checked against the
2928 same list, then the first domain was re-checked, the value of $domain_data
2929 after the final check could be wrong. In particular, if the second check
2930 failed, it could be set empty. This bug probably also applied to
2933 41. The strip_trailing_dot option was not being applied to the address given
2934 with the -f command-line option.
2936 42. The code for reading a message's header from the spool was incrementing
2937 $received_count, but never initializing it. This meant that the value was
2938 incorrect (doubled) while delivering a message in the same process in which
2939 it was received. In the most common configuration of Exim, this never
2940 happens - a fresh exec is done - but it can happen when
2941 deliver_drop_privilege is set.
2943 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2944 it now includes up to 150 characters of the unexpected data in the log
2947 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2948 and building data strings. The size of both of these buffers was 10 000
2949 bytes - far larger than anybody would *ever* want, thought I. Needless to
2950 say, somebody hit the limit. I have increased the maximum line length to
2951 20 000 and the maximum data length of concatenated lines to 100 000. I have
2952 also fixed two bugs, because there was no checking on these buffers. Tsk,
2953 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2954 buffer is too small.
2956 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2957 lsearch lookups. Now it does.
2959 46. When parsing a route_list item in a manualroute router, a fixed-length
2960 buffer was used for the list of hosts. I made this 1024 bytes long,
2961 thinking that nobody would ever have a list of hosts that long. Wrong.
2962 Somebody had a whole pile of complicated expansion conditions, and the
2963 string was silently truncated, leading to an expansion error. It turns out
2964 that it is easier to change to an unlimited length (owing to other changes
2965 that have happened since this code was originally written) than to build
2966 structure for giving a limitation error. The length of the item that
2967 expands into the list of hosts is now unlimited.
2969 47. The lsearch lookup could not handle data where the length of text line was
2970 more than 4095 characters. Such lines were truncated, leading to shortened
2971 data being returned. It should now handle lines of any length.
2973 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
2974 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
2977 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
2978 replacing strings such as BIN_DIRECTORY in a source file: the replacement
2979 no longer happens in comment lines. A list of replacements is now placed
2980 at the head of all of the source files, except those whose only change is
2981 to replace PERL_COMMAND in the very first #! line.
2983 50. Replaced the slow insertion sort in queue.c, for sorting the list of
2984 messages on the queue, with a bottom-up merge sort, using code contributed
2985 by Michael Haardt. This should make operations like -bp somewhat faster on
2986 large queues. It won't affect queue runners, except when queue_run_in_order
2989 51. Installed eximstats 1.31 in the distribution.
2991 52. Added support for SRV lookups to the dnslookup router.
2993 53. If an ACL referred to $message_body or $message_body_end, the value was not
2994 reset for any messages that followed in the same SMTP session.
2996 54. The store-handling optimization for building very long strings was not
2997 differentiating between the different store pools. I don't think this
2998 actually made any difference in practice, but I've tidied it.
3000 55. While running the routers to verify a sender address, $sender_address
3001 was still set to the sender address. This is wrong, because when routing to
3002 send a bounce to the sender, it would be empty. Therefore, I have changed
3003 it so that, while verifying a sender address, $sender_address is set to <>.
3004 (There is no change to what happens when verifying a recipient address.)
3006 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3007 target A or AAAA records (if not already returned) without resetting the
3008 qualify_single or search_parents options of the DNS resolver. These are
3009 inappropriate in this case because the targets of MX and SRV records must
3010 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3011 target that, when qualified, matched something in the local domain. These
3012 two options are now turned off when doing these lookups.
3014 57. It seems that at least some releases of Reiserfs (which does not have the
3015 concept of a fixed number of inodes) returns zero and not -1 for the
3016 number of available inodes. This interacted badly with check_spool_inodes,
3017 which assumed that -1 was the "no such thing" setting. What I have done is
3018 to check that the total number of inodes is greater than zero before doing
3019 the test of how many are available.
3021 58. When a "warn" ACL statement has a log_message modifier, the message is
3022 remembered, and not repeated. This is to avoid a lot of repetition when a
3023 message has many recipients that cause the same warning to be written.
3024 Howewer, Exim was preserving the list of already written lines for an
3025 entire SMTP session, which doesn't seem right. The memory is now reset if a
3026 new message is started.
3028 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3029 debugging output unless log_rewrite was also set.
3031 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3032 of (int)(handle) when we know that handle contains (void *)(-1).
3034 61. The Exim daemon panic-logs an error return when it closes the incoming
3035 connection. However "connection reset by peer" seems to be common, and
3036 isn't really an error worthy of noting specially, so that particular error
3039 62. When Exim is trying to find all the local interfaces, it used to panic and
3040 die if the ioctl to get the interface flags failed. However, it seems that
3041 on at least one OS (Solaris 9) it is possible to have an interface that is
3042 included in the list of interfaces, but for which you get a failure error
3043 for this call. This happens when the interface is not "plumbed" into a
3044 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3045 failure of the "get flags" call assumes that the interface is down.
3047 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3048 makes life easier for people who are doing arithmetic on fields extracted
3049 from dates, where you often get leading zeros that should not be
3050 interpreted as octal.
3052 64. Added qualify_domain to the redirect router, to override the global
3055 65. If a pathologically long header line contained very many addresses (the
3056 report of this problem mentioned 10 000) and each of them was rewritten,
3057 Exim could use up a very large amount of memory. (It kept on making new
3058 copies of the header line as it rewrote, and never released the old ones.)
3059 At the expense of a bit more processing, the header rewriting function has
3060 been changed so that it no longer eats memory in this way.
3062 66. The generation of the Received: header has been moved from the time that a
3063 message starts to be received, to the time that it finishes. The timestamp
3064 in the Received: header should now be very close to that of the <= log
3065 line. There are two side-effects of this change:
3067 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3068 logged header lines no longer include the local Received: line, because
3069 it has not yet been created. The same applies to a copy of the message
3070 that is returned to a non-SMTP sender when a message is rejected.
3072 (b) When a filter file is tested using -bf, no additional Received: header
3073 is added to the test message. After some thought, I decided that this
3076 This change does not affect the value of $received_for. It is still set
3077 after address rewriting, but before local_scan() is called.
3079 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3081 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3082 gave an unhelpful panic error message, and a defer error. I have managed to
3083 change this behaviour so that it now rejects any supplied certificate,
3084 which seems right, as the list of acceptable certificates is empty.
3086 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3087 gave an unhelpful defer error. I have not managed to make this reject any
3088 supplied certificates, but the error message it gives is "no certificate
3089 supplied", which is not helpful.
3091 70. exigrep's output now also includes lines that are not associated with any
3092 message, but which match the given pattern. Implemented by a patch from
3093 Martin Sluka, which also tidied up the Perl a bit.
3095 71. Recipient callout verification, like sender verification, was using <> in
3096 the MAIL FROM command. This isn't really the right thing, since the actual
3097 sender may affect whether the remote host accepts the recipient or not. I
3098 have changed it to use the actual sender in the callout; this means that
3099 the cache record is now keyed on a recipient/sender pair, not just the
3100 recipient address. There doesn't seem to be a real danger of callout loops,
3101 since a callout by the remote host to check the sender would use <>.
3102 [SEE ABOVE: changed after hitting problems.]
3104 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3105 temporary errors. However, in the case of such a code being given after
3106 the end of a data transmission (i.e. after ".") Exim was failing to write
3107 a retry record for the message. (Yes, there was some broken host that was
3108 actually sending 8xx at this point.)
3110 73. An unknown lookup type in a host list could cause Exim to panic-die when
3111 the list was checked. (An example that provoked this was putting <; in the
3112 middle of a list instead of at the start.) If this happened during a DATA
3113 ACL check, a -D file could be left lying around. This kind of configuration
3114 error no longer causes Exim to die; instead it causes a defer errror. The
3115 incident is still logged to the main and panic logs.
3117 74. Buglet left over from Exim 3 conversion. The message "too many messages
3118 in one connection" was written to the rejectlog but not the mainlog, except
3119 when address rewriting (yes!) was being logged.
3121 75. Added write_rejectlog option.
3123 76. When a system filter was run not as root (that is, when system_filter_user
3124 was set), the values of the $n variables were not being returned to the
3125 main process; thus, they were not subsequently available in the $sn
3128 77. Added +return_path_on_delivery log selector.
3130 78. A connection timeout was being treated differently from recipients deferred
3131 when testing hosts_max_try with a message that was older than the host's
3132 retry timeout. (The host should not be counted, thus allowing all hosts to
3133 be tried at least once before bouncing.) This may have been the cause of an
3134 occasionally reported bug whereby a message would remain on the queue
3135 longer than the retry timeout, but would be bounced if a delivery was
3136 forced. I say "may" because I never totally pinned down the problem;
3137 setting up timeout/retry tests is difficult. See also the next item.
3139 79. The ultimate address timeout was not being applied to errors that involved
3140 a combination of host plus message (for example, a timeout on a MAIL
3141 command). When an address resolved to a number of possible hosts, and they
3142 were not all tried for each delivery (e.g. because of hosts_max_try), a
3143 message could remain on the queue longer than the retry timeout.
3145 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3148 81. Fixed an obscure SMTP outgoing bug which required at least the following
3149 conditions: (a) there was another message waiting for the same server;
3150 (b) the server returned 5xx to all RCPT commands in the first message so
3151 that the message was not completed; (c) the server dropped the connection
3152 or gave a negative response to the RSET that Exim sends to abort the
3153 transaction. The observed case was a dropped connection after DATA that had
3154 been sent in pipelining mode. That is, the server had advertised PIPELINING
3155 but was not implementing it correctly. The effect of the bug was incorrect
3156 behaviour, such as trying another host, and this could lead to a crash.
3162 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3163 and daemon.c were passed as pointers to ints; they should have been
3164 pointers to socklen_t variables (which are typically unsigned ints).
3166 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3169 3. Fixed a really odd bug that affected only the testing scheme; patching a
3170 certain fixed string in the binary changed the value of another string that
3171 happened to be identical to the end of the original first string.
3173 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3174 name", it returns that address as the IP address. On some operating
3175 systems (e.g. Solaris), it also passes back the IP address string as the
3176 "host name". However, on others (e.g. Linux), it passes back an empty
3177 string. Exim wasn't checking for this, and was changing the host name to an
3178 empty string, assuming it had been canonicized.
3180 5. Although rare, it is permitted to have more than one PTR record for a given
3181 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3182 all the names associated with an address, because they do in Solaris.
3183 However, it seems that they do not in Linux for data that comes from the
3184 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3185 I found this out when I moved to a new Linux workstation and tried to run
3186 the Exim test suite.
3188 To get round this problem I have changed the code so that it now does its
3189 own call to the DNS to look up PTR records when searching for a host name.
3190 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3191 addresses that are only in /etc/hosts are still found.
3193 This behaviour is, however, controlled by an option called host_lookup_
3194 order, which defaults to "bydns:byaddr". If people want to use the other
3195 order, or indeed, just use one or the other means of lookup, they can
3196 specify it in this variable.
3198 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3199 some operating systems, this comes back from gethostbyaddr() as an empty
3200 string, and this is what Exim used to test for. However, it seems that in
3201 other systems, "." is yielded. Exim now tests for this case too.
3203 7. The values of check_spool_space and check_log_space are now held internally
3204 as a number of kilobytes instead of an absolute number of bytes. If a
3205 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3206 kilobyte. This means that much larger values can be stored.
3208 8. Exim monitor: an attempt to get the action menu when not actually pointing
3209 at a message produces an empty menu entitled "No message selected". This
3210 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3211 no entries in it ("Shell widget menu has zero width and/or height"). So I
3212 have added a single, blank menu entry in this case.
3214 9. Added ${quote_local_part.
3216 10. MIME decoding is now applied to the contents of Subject: header lines when
3219 11. Now that a reference to $sender_host_address automatically causes a reverse
3220 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3221 host lookup before query-style lookups in lists that might use this
3222 variable. This has therefore been abolished, and the "net-" prefix is no
3223 longer necessary for query-style lookups.
3225 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3226 have been a typo for LFLAGS, so it has been changed.
3228 13. The install script calls Exim with "-C /dev/null" in order to find the
3229 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3230 to be output. Howeve, since Exim outputs its version number before the
3231 error, it didn't break the script. It just looked ugly. I fixed this by
3232 always allowing "-C /dev/null" if the caller is root.
3234 14. Ignore overlarge ACL variable number when reading spool file - insurance
3235 against a later release with more variables having written the file.
3237 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3238 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3239 this, as well as the form without the "IPv6" on the front.
3241 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3242 OS/Makefile-Darwin file.
3244 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3245 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3247 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3248 variable. However, some OS use ints for inodes. I've added cast to long int
3249 to get rid of the compiler warning.
3251 19. I had forgotten to lock out "/../" in configuration file names when
3252 ALT_CONFIG_PREFIX was set.
3254 20. Routers used for verification do not need to specify transports. However,
3255 if such a router generated a host list, and callout was configured, Exim
3256 crashed, because it could not find a port number from the (non-existent)
3257 transport. It now assumes port 25 in this circumstance.
3259 21. Added the -t option to exigrep.
3261 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3262 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3263 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3264 them is set, LOOKUP_LSEARCH is forced.
3266 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3268 24. Added sender and host information to the "rejected by local_scan()" log
3269 line; previously there was no indication of these.
3271 25. Added .include_if_exists.
3273 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3274 Linux. It turns out that not all file systems support this. Apparently some
3275 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3276 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3277 not supported on the file descriptor, is now ignored when Exim is trying to
3278 sync a directory. This applies only to Linux.
3280 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3282 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3283 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3286 29. Added support for maildirsize files from supplied patch (modified a bit).
3288 30. The use of :fail: followed by an empty string could lead Exim to respond to
3289 sender verification failures with (e.g.):
3291 550 Verification failed for <xxx>
3292 550 Sender verify failed
3294 where the first response line was missing the '-' that indicates it is not
3295 the final line of the response.
3297 31. The loop for finding the name of the user that called Exim had a hardwired
3298 limit of 10; it now uses the value of finduser_retries, which is used for
3299 all other user lookups.
3301 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3304 33. Exim was neglecting to zero errno before one call of strtol() when
3305 expanding a string and expecting an integer value. On some systems this
3306 resulted in spurious "integer overflow" errors. Also, it was casting the
3307 result into an int without checking.
3309 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3310 did not work. The code looks as if it has *never* worked, though it appears
3311 to have been documented since at least releast 1.62. I have made it work.
3313 35. The "timeout_DNS" error in retry rules, also documented since at least
3314 1.62, also never worked. As it isn't clear exactly what this means, and
3315 clearly it isn't a major issue, I have abolished the feature by treating it
3316 as "timeout", and writing a warning to the main and panic logs.
3318 36. The display of retry rules for -brt wasn't always showing the error code
3321 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3322 timeout_connect_A, timeout_connect_MX.
3324 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3325 to the empty sender.
3327 39. The daemon was not analysing the content of -oX till after it had closed
3328 stderr and disconnected from the controlling terminal. This meant that any
3329 syntax errors were only noted on the panic log, and the return code from
3330 the command was 0. By re-arranging the code a little, I've made the
3331 decoding happen first, so such errors now appear on stderr, and the return
3332 code is 1. However, the actual setting up of the sockets still happens in
3333 the disconnected process, so errors there are still only recorded on the
3336 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3337 connections (as happens on some IP stacks) was logged at start up time as
3338 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3339 it from "IPv6 and IPv4", which means that two separate sockets are being
3342 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3343 says whether AF_INET or AF_INET6 was passed as an argument.
3345 42. Exiwhat output was messed up when time zones were included in log
3348 43. Exiwhat now gives more information about the daemon's listening ports,
3349 and whether -tls-on-connect was used.
3351 44. The "port" option of the smtp transport is now expanded.
3353 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3354 silently ignored. Now an error message is written to the main and panic
3357 46. There's a new ACL modifier called "logwrite" which writes to a log file
3358 as soon as it is encountered.
3360 47. Added $local_user_uid and $local_user_gid at routing time.
3362 48. Exim crashed when trying to verify a sender address that was being
3365 49. Exim was recognizing only a space character after ".include". It now also
3366 recognizes a tab character.
3368 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3369 extracting the relevant information from the specification. The man page no
3370 longer contains scrambled data for the -d option, and I've added a section
3371 at the front about calling Exim under different names.
3373 51. Added "extra_headers" argument to the "mail" command in filter files.
3375 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3378 53. Installed eximstats 1.29.
3380 54. Added transport_filter_timeout as a generic transport option.
3382 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3383 Cc: header lines. This was required by RFC 822, but it not required by RFC
3386 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3387 incoming messages that did not have them. Now it does so only if the
3388 message originates locally, that is, if there is no associated remote host
3389 address. When Resent- header lines are present, this applies to the Resent-
3390 lines rather than the non-Resent- lines.
3392 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3393 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3395 58. Messages for configuration errors now include the name of the main
3396 configuration file - useful now that there may be more than one file in a
3397 list (.included file names were always shown).
3399 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3400 for those rare installations that do not start the daemon as root or run it
3401 setuid root. I've cut out the call to initgroups() if the daemon is not
3404 60. The Exim user and group can now be bound into the binary as text strings
3405 that are looked up at the start of Exim's processing.
3407 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3409 62. Added $mailstore_basename variable.
3411 63. Installed patch to sieve.c from Michael Haardt.
3413 64. When Exim failed to open the panic log after failing to open the main log,
3414 the original message it was trying to log was written to stderr and debug
3415 output, but if they were not available (the usual case in production), it
3416 was lost. Now it is written to syslog before the two lines that record the
3417 failures to open the logs.
3419 65. Users' Exim filters run in subprocesses under the user's uid. It is
3420 possible for a "deliver" command or an alias in a "personal" command to
3421 provoke an address rewrite. If logging of address rewriting is configured,
3422 this fails because the process is not running as root or exim. There may be
3423 a better way of dealing with this, but for the moment (because 4.30 needs
3424 to be released), I have disabled address rewrite logging when running a
3425 filter in a non-root, non-exim process.
3431 1. The buildconfig auxiliary program wasn't quoting the value set for
3432 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3433 not defined. This bug was masked in 4.22 by the effect that was fixed in
3436 2. Some messages that were rejected after a message id was allocated were
3437 shown as "incomplete" by exigrep. It no longer does this for messages that
3438 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3440 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3441 have allow_domain_literals set, the ID did not get logged in the <= line.
3442 Domain literals are now always recognized in Message-ID: header lines.
3444 4. The first argument for a ${extract expansion item is the key name or field
3445 number. Leading and trailing spaces in this item were not being ignored,
3446 causing some misleading effects.
3448 5. When deliver_drop_privilege was set, single queue runner processes started
3449 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3450 same command in the process it spins off) were not dropping privilege.
3452 6. When the daemon running as "exim" started a queue runner, it always
3453 re-executed Exim in the spun-off process. This is a waste of effort when
3454 deliver_drop_privilege is set. The new process now just calls the
3455 queue-runner function directly.
3461 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3464 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3465 ignored. Though the use of -oP was forcing the writing of a pid file, it
3466 was always written to the default place.
3468 3. If the message "no IP address found for host xxxx" is generated during
3469 incoming verification, it is now followed by identification of the incoming
3470 connection (so you can more easily find what provoked it).
3472 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3474 5. Added some features to "harden" Exim a bit more against certain attacks:
3476 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3477 be put in Local/Makefile. This is like the never_users runtime option,
3478 but it cannot be overridden. The default setting is "root".
3480 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3481 prefix string with which any file named in a -C command line option
3484 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3485 is retained for -C and -D only if the caller of Exim is root. Without
3486 it, the exim user may also use -C and -D and retain privilege.
3488 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3489 command line option is disabled.
3491 6. Macro names set by the -D option must start with an upper case letter, just
3492 like macro names defined in the configuration file.
3494 7. Added "dereference=" facility to LDAP.
3496 8. Two instances of the typo "uknown" in the source files are fixed.
3498 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3499 the Configure-Makefile script screwed up while processing it.
3501 10. Incorporated PCRE 4.4.
3503 11. The SMTP synchronization check was not operating right at the start of an
3504 SMTP session. For example, it could not catch a HELO sent before the client
3505 waited for the greeting. There is now a check for outstanding input at the
3506 point when the greeting is written. Because of the duplex, asynchronous
3507 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3508 way, but not yet received, when the check is performed.
3510 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3511 on TCP/IP sockets, because this apparently causes some broken clients to
3514 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3515 unchanged) from the Cygwin maintainer.
3517 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3518 was supported instead of testing for mbx. Effectively a typo.
3520 15. The spa authenticator server code was not checking that the input it
3521 received was valid base64.
3523 16. The debug output line for the "set" modifier in ACLs was not showing the
3524 name of the variable that was being set.
3526 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3528 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3529 lookup on its own. Something else was needed to trigger the lookup. For
3530 example, a match in host_lookup or the need for a host name in a host list.
3531 Now, if $sender_host_name is referenced and the host name has not yet been
3532 looked up, a lookup is performed. If the lookup fails, the variable remains
3533 empty, and $host_lookup_failed is set to "1".
3535 19. Added "eqi" as a case-independent comparison operator.
3537 20. The saslauthd authentication condition could segfault if neither service
3538 nor realm was specified.
3540 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3541 error message that was logged was misleading, and incoming SMTP
3542 connections were dropped. The message is now more accurate, and temporary
3543 errors are given to SMTP connections.
3545 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3546 (except RSET) with a 421 error, until QUIT is received. However, it was
3547 failing to send a response to QUIT.
3549 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3550 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3551 ACL is now run after the helo_try_verify_hosts code.
3553 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3554 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3555 case-independent, so other case variants are also recognized). Apparently
3556 some systems use these upper case variants.
3558 25. If more than two messages were waiting for the same host, and a transport
3559 filter was specified for the transport, Exim sent two messages over the
3560 same TCP/IP connection, and then failed with "socket operation on non-
3561 socket" when it tried to send the third.
3563 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3565 27. The extern definition of crypt16() in expand.c was not being excluded when
3566 the OS had its own crypt16() function.
3568 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3569 as a preferred synonym for return_size_limit, both as an option and as an
3572 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3574 30. Changed the default configuration ACL to relax the local part checking rule
3575 for addresses that are not in any local domains. For these addresses,
3576 slashes and pipe symbols are allowed within local parts, but the sequence
3577 /../ is explicitly forbidden.
3579 31. SPA server authentication was not clearing the challenge buffer before
3582 32. log_message in a "warn" ACL statement was writing to the reject log as
3583 well as to the main log, which contradicts the documentation and doesn't
3584 seem right (because no rejection is happening). So I have stopped it.
3586 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3587 However, I am unable to do any testing of this.
3589 34. Fixed an infelicity in the appendfile transport. When checking directories
3590 for a mailbox, to see if any needed to be created, it was accidentally
3591 using path names with one or more superfluous leading slashes; tracing
3592 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3594 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3595 contents are added to the log line that is written for every discarded
3596 recipient. (Previously a log_message setting was ignored.)
3598 36. The ${quote: operator now quotes the string if it is empty.
3600 37. The install script runs exim in order to find its version number. If for
3601 some reason other than non-existence or emptiness, which it checks, it
3602 could not run './exim', it was installing it with an empty version number,
3603 i.e. as "exim-". This error state is now caught, and the installation is
3606 38. An argument was missing from the function that creates an error message
3607 when Exim fails to connect to the socket for saslauthd authentication.
3608 This could cause Exim to crash, or give a corrupted message.
3610 39. Added isip, isip4, and isip6 to ${if conditions.
3612 40. The ACL variables $acl_xx are now saved with the message, and can be
3613 accessed later in routers, transports, and filters.
3615 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3616 strings in the file are not string-expanded.
3618 42. If a MAIL command specified a SIZE value that was too large to fit into an
3619 int variable, the check against message_size_limit failed. Such values are
3620 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3621 day this will have to be increased, but I don't think I want to be around
3622 when emails are that large.
3629 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3630 iconv() is not standard in FreeBSD.
3632 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3633 IPv6 enabled. The observed symptom was a segmentation fault on return from
3634 the function os_common_find_running_interfaces() in src/os.c.
3636 3. In the check_special_case() function in daemon.c I had used "errno" as an
3637 argument name, which causes warnings on some systems. This was basically a
3638 typo, since it was named "eno" in the comments!
3640 4. The code that waits for the clock to tick (at a resolution of some fraction
3641 of a second) so as to ensure message-id uniqueness was always waiting for
3642 at least one whole tick, when it could have waited for less. [This is
3643 almost certainly not relevant at current processor speeds, where it is
3644 unlikely to ever wait at all. But we try to future-proof.]
3646 5. The function that sleeps for a time interval that includes fractions of a
3647 second contained a race. It did not block SIGALRM between setting the
3648 timer, and suspending (a couple of lines later). If the interval was short
3649 and the sigsuspend() was delayed until after it had expired, the suspension
3650 never ended. On busy systems this could lead to processes getting stuck for
3653 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3654 process, before it forks any delivery processes. The open lookup caching
3655 mechanism meant that the open file or database connection was passed into
3656 the delivery process. The problem was that delivery processes always tidy
3657 up cached lookup data. This could cause a problem for the next delivery
3658 process started by the queue runner, because the external queue runner
3659 process does not know about the closure. So the next delivery process
3660 still has data in the lookup cache. In the case of a file lookup, there was
3661 no problem because closing a file descriptor in a subprocess doesn't affect
3662 the parent. However, if the lookup was caching a connection to a database,
3663 the connection was closed, and the second delivery process was likely to
3664 see errors such as "PGSQL: query failed: server closed the connection
3665 unexpectedly". The problem has been fixed by closing all cached lookups
3666 in a queue runner before running a delivery process.
3668 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3669 seem to have the "const" qualifier which it has on other OS. I've
3672 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3673 *of the same type* (client or server) with the same public name that an
3674 error should be diagnosed.
3676 9. When Exim looked up a host name for an IP address, but failed to find the
3677 original IP address when looking up the host name (a safety check), it
3678 output the message "<ip address> does not match any IP for NULL", which was
3679 confusing, to say the least. The bug was that the host name should have
3680 appeared instead of "NULL".
3682 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3683 user that is built into the binary, and the -C or -D options is used, root
3684 privilege is dropped before the configuration file is read. In addition,
3685 logging is switched to stderr instead of the normal log files. If the
3686 configuration then re-defines the Exim user, the unprivileged environment
3687 is probably not what is expected, so Exim logs a panic warning message (but
3690 However, if deliver_drop_privilege is set, the unprivileged state may well
3691 be exactly what is intended, so the warning has been cut out in that case,
3692 and Exim is allowed to try to write to its normal log files.
3698 1. smtp_return_error_details was not giving details for temporary sender
3699 or receiver verification errors.
3701 2. Diagnose a configuration error if two authenticators have the same public
3704 3. Exim used not to create the message log file for a message until the first
3705 delivery attempt. This could be confusing when incoming messages were held
3706 for policy or load reasons. The message log file is now created at the time
3707 the message is received, and an initial "Received" line is written to it.
3709 4. The automatically generated man page for command line options had a minor
3710 bug that caused no ill effects; however, a more serious problem was that
3711 the procedure for building the man page automatically didn't always
3712 operate. Consequently, release 4.20 contains an out-of-date version. This
3713 shouldn't happen again.
3715 5. When building Exim with embedded Perl support, the script that builds the
3716 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3717 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3719 6. The freeze_tell option was not being used for messages that were frozen on
3720 arrival, either by an ACL or by local_scan().
3722 7. Added the smtp_incomplete_transaction log selector.
3724 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3725 was accepting AUTH without a new EHLO.
3727 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3728 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3731 10. Logging of TCP/IP connections (when configured) now happens in the main
3732 daemon process instead of the child process, so that the TCP/IP connection
3733 count is more accurate (but it can never be perfect).
3735 11. The use of "drop" in a nested ACL was not being handled correctly in the
3736 outer ACL. Now, if condition failure induced by the nested "drop" causes
3737 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3738 or "require"), the connection is dropped.
3740 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3741 that yield "discard" can now be used with an "accept" or a "discard" verb,
3742 but an error is generated for any others (because I can't see a useful way
3743 to define what should happen).
3745 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3746 are now processed in the same way as lines in the Exim configuration file.
3747 In particular, continuation lines are supported.
3749 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3751 15. Added -ti meaning -t -i.
3753 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3754 domains, and warn by logging if others are found.
3756 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3757 returned from GIFCONF when Exim is trying to find the list of interfaces on
3758 a host. The code in os.c has been modified to copy each ifreq to an aligned
3759 structure in all cases.
3761 Also, in some cases, the returned ifreq's were being copied to a 'struct
3762 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3763 means the last couple of bytes of an IPv6 address could be chopped if the
3764 ifreq contained only a normal sockaddr (14 bytes storage).
3766 18. Named domain lists were not supported in the hosts_treat_as_local option.
3767 An entry such as +xxxx was not recognized, and was treated as a literal
3770 19. Ensure that header lines added by a DATA ACL are included in the reject log
3771 if the ACL subsequently rejects the message.
3773 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3774 MD5 (which is deprecated).
3776 21. When testing a filter file using -bf, Exim was writing a message when it
3777 took the sender from a "From " line in the message, but it was not doing so
3778 when it took $return_path from a Return-Path: header line. It now does.
3780 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3781 with a valid header line field name (a series of printing characters
3782 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3784 23. Changed "disc" in the source to "disk" to conform to the documentation and
3785 the book and for uniformity.
3787 24. Ignore Sendmail's -Ooption=value command line item.
3789 25. When execve() failed while trying to run a command in a pipe transport,
3790 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3791 could be confused with a return value of 69 from the command itself. This
3792 has been changed to 127, the value the shell returns if it is asked to run
3793 a non-existent command. The wording for the related log line suggests a
3794 non-existent command as the problem.
3796 26. If received_header_text expands to an empty string, do not add a Received:
3797 header line to the message. (Well, it adds a token one on the spool, but
3798 marks it "old" so that it doesn't get used or transmitted.)
3800 27. Installed eximstats 1.28 (addition of -nt option).
3802 28. There was no check for failure on the call to getsockname() in the daemon
3803 code. This can fail if there is a shortage of resources on the system, with
3804 ENOMEM, for example. A temporary error is now given on failure.
3806 29. Contrary to the C standard, it seems that in some environments, the
3807 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3808 program. Exim now does this explicitly; it affects the formatting of
3809 timestamps using strftime().
3811 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3812 complaints. I've now initialized all the variables, to avoid this.
3814 32. Header lines added by a system filter were not being "seen" during
3815 transport-time rewrites.
3817 33. The info_callback() function passed to OpenSSL is set up with type void
3818 (*)(SSL *, int, int), as described somewhere. However, when calling the
3819 function (actually a macro) that sets it up, the type void(*)() is
3820 expected. I've put in a cast to prevent warnings from picky compilers.
3822 34. If a DNS black list lookup found a CNAME record, but there were no A
3823 records associated with the domain it pointed at, Exim crashed.
3825 35. If a DNS black list lookup returned more than one A record, Exim ignored
3826 all but the first. It now scans all returned addresses if a particular IP
3827 value is being sought. In this situation, the contents of the
3828 $dnslist_value variable are a list of all the addresses, separated by a
3831 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3832 to accept a host name and all its aliases if the forward lookup for any of
3833 them yielded the IP address of the incoming connection. Now it accepts only
3834 those names whose forward lookup yields the correct IP address. Any other
3835 names are discarded. This closes a loophole whereby a rogue DNS
3836 administrator could create reverse DNS records to break through a
3837 wildcarded host restriction in an ACL.
3839 37. If a user filter or a system filter that ran in a subprocess used any of
3840 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3841 the wrong values were passed to the pipe command ($thisaddress had the
3842 value of $0, $0 had the value of $1, etc). This bug was introduced by
3843 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3845 38. Improved the line breaking for long SMTP error messages from ACLs.
3846 Previously, if there was no break point between 40 and 75 characters, Exim
3847 left the rest of the message alone. Two changes have been made: (a) I've
3848 reduced the minimum length to 35 characters; (b) if it can't find a break
3849 point between 35 and 75 characters, it looks ahead and uses the first one
3850 that it finds. This may give the occasional overlong line, but at least the
3851 remaining text gets split now.
3853 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3854 file descriptors might be low, and that setting 1000 would always raise it.
3855 It turns out that in some environments, the limit is already over 1000 and
3856 that lowering it causes trouble. So now Exim takes care not to decrease it.
3858 40. When delivering a message, the value of $return_path is set to $sender_
3859 address at the start of routing (routers may change the value). By an
3860 oversight, this default was not being set up when an address was tested by
3861 -bt or -bv, which affected the outcome if any router or filter referred to
3864 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3865 log only when "message" or "log_message" are set. However, if one of the
3866 conditions was an address verification, or a call to a nested ACL, the
3867 messages generated by the underlying test were being passed through. This
3868 no longer happens. The underlying message is available in $acl_verify_
3869 message for both "message" and "log_message" expansions, so it can be
3870 passed through if needed.
3872 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3873 new expansion $bh_ to give the encoded byte string without charset
3874 translation. Translation happens only if iconv() is available; HAVE_ICONV
3875 indicates this at build time. HEADERS_CHARSET gives the charset to
3876 translate to; headers_charset can change it in the configuration, and
3877 "headers charset" can change it in an individual filter file.
3879 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3880 that creates RFC 2047 encoded "words" labels them as that charset instead
3881 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3882 expansion operator; (ii) when Exim creates a From: line for a local
3883 message; (iii) when a header line is rewritten to include a "phrase" part.
3885 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3886 buggy, causing it to skip the first lines of messages whose message ID
3887 ended in 'D'. This would not have bitten before Exim release 4.14, because
3888 message IDs were unlikely to end in 'D' before then. The effect was to have
3889 incorrect size information for certain domains.
3891 45. #include "config.h" was missing at the start of the crypt16.c module. This
3892 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3895 46. If there was a timeout during a "random" callout check, Exim treated it as
3896 a failure of the random address, and carried on sending RSET and the real
3897 address. If the delay was just some slowness somewhere, the response to the
3898 original RCPT would be taken as a response to RSET and so on, causing
3899 mayhem of various kinds.
3901 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3902 when I implemented it. It didn't allow for the fact that some option values
3903 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3904 the right test for positive values.
3906 48. Domain names in DNS records are case-independent. Exim always looks them up
3907 in lower case. Some resolvers return domain names in exactly the case they
3908 appear in the zone file, that is, they may contain uppercase letters. Not
3909 all resolvers do this - some return always lower case. Exim was treating a
3910 change of case by a resolver as a change of domain, similar to a widening
3911 of a domain abbreviation. This triggered its re-routing code and so it was
3912 trying to route what was effectively the same domain again. This normally
3913 caused routing to fail (because the router wouldn't handle the domain
3914 twice). Now Exim checks for this case specially, and just changes the
3915 casing of the domain that it ultimately uses when it transmits the message
3918 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3921 50. If a filter generated a file delivery with a non-absolute name (possible if
3922 no home directory exists for the router), the forbid_file option was not
3925 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3926 the existing equality matching.
3928 52. Exim was using ints instead of ino_t variables in some places where it was
3929 dealing with inode numbers.
3931 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3932 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3933 called TMPDIR, and if it finds it is different, it changes its value.
3935 54. The smtp_printf() function is now made available to local_scan() so
3936 additional output lines can be written before returning. There is also an
3937 smtp_fflush() function to enable the detection of a dropped connection.
3938 The variables smtp_input and smtp_batched_input are exported to
3941 55. Changed the default runtime configuration: the message "Unknown user"
3942 has been removed from the ACL, and instead placed on the localuser router,
3943 using the cannot_route_message feature. This means that any verification
3944 failures that generate their own messages won't get overridden. Similarly,
3945 the "Unrouteable address" message that was in the ACL for unverifiable
3946 relay addresses has also been removed.
3948 56. Added hosts_avoid_esmtp to the smtp transport.
3950 57. The exicyclog script was not checking for the esoteric option
3951 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3952 will work only if exicyclog is run under the appropriate euid.
3954 58. Following a discussion on the list, the rules by which Exim recognises line
3955 endings on incoming messages have been changed. The -dropcr and drop_cr
3956 options are now no-ops, retained only for backwards compatibility. The
3957 following line terminators are recognized: LF CRLF CR. However, special
3958 processing applies to CR:
3960 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3961 nor a local message in the state where . is a terminator.
3963 (ii) If a bare CR is encountered in a header line, an extra space is added
3964 after the line terminator so as not to end the header. The reasoning
3965 behind this is that bare CRs in header lines are most likely either
3966 to be mistakes, or people trying to play silly games.
3968 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
3969 was being incorrectly given as 18 bytes larger than it should have been.
3970 This is a VOB (very old bug).
3972 60. This may never have affected anything current, but just in case it has:
3973 When the local host is found other than at the start of a list of hosts,
3974 the local host, those with the same MX, and any that follow, are discarded.
3975 When the list in question was part of a longer list of hosts, the following
3976 hosts (not currently being processed) were also being discarded. This no
3977 longer happens. I'm not sure if this situation could ever has previously
3980 61. Added the "/MX" feature to lists of hosts in the manualroute and query
3983 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
3984 header. This is something that is recommended in a new Internet Draft, and
3985 is something that is documented as being done by Sendmail. There are two
3986 possible values. For messages generated by the autoreply transport, Exim
3989 Auto-Submitted: auto-replied
3991 whereas for all other generated messages (e.g. bounces) it adds
3993 Auto-Submitted: auto-generated
3995 63. The "personal" condition in filters now includes a test for the
3996 Auto-Submitted: header. If it contains the string "auto-" the message it
3997 not considered personal.
3999 64. Added rcpt_include_affixes as a generic transport option.
4001 65. Added queue_only_override (default true).
4003 66. Added the syslog_duplication option.
4005 67. If what should have been the first header line of a message consisted of
4006 a space followed by a colon, Exim was mis-interpreting it as a header line.
4007 It isn't of course - it is syntactically invalid and should therefore be
4008 treated as the start of the message body. The misbehaviour could have
4009 caused a number of strange effects, including loss of data in subsequent
4010 header lines, and spool format errors.
4012 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4013 client host had authenticated. This control can now be exercised by an ACL
4014 for more flexibility.
4016 69. By default, callouts do not happen when testing with -bh. There is now a
4017 variant, -bhc, which does actually run the callout code, including
4018 consulting and updating the callout cache.
4020 70. Added support for saslauthd authentication, courtesy of Alexander
4023 71. If statvfs() failed on the spool or log directories while checking their
4024 size for availability, Exim confusingly gave the error "space shortage".
4025 Furthermore, in debugging mode it crashed with a floating point exception.
4026 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4027 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4028 serious problem, Exim now writes to the main and panic logs when this
4029 happens, with details of the failure. It then refuses to accept the
4030 incoming message, giving the message "spool directory problem" or "log
4031 directory problem" with a 421 code for SMTP messages.
4033 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4034 0, 1, and 2 exist, because some OS complain for execs without them (see
4035 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4036 descriptors. However, the code omitted to check that the open succeeded,
4037 causing mysterious errors if for some reason the permissions on /dev/null
4038 got screwed. Now Exim writes a message to the main and panic logs, and
4039 bombs out if it can't open /dev/null.
4041 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4042 interact so that it is all more flexible. It is supposed to remain
4043 backwards compatible. Also added extra_local_interfaces.
4045 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4046 to bomb out with an assertion failure - to the client this appears as a
4047 connection drop. This problem occurs in the part of the code that was taken
4048 from the Samba project. Fortunately, the assertion is in a very simple
4049 function, so I have fixed this by reproducing the function inline in the
4050 one place where it is called, and arranging for authentication to fail
4051 instead of killing the process with assert().
4053 75. The SPA client code was not working when the server requested OEM rather
4054 than Unicode encoding.
4056 76. Added code to make require_files with a specific uid setting more usable in
4057 the case where statting the file as root fails - usually a non-root-mounted
4058 NFS file system. When this happens and the failure is EACCES, Exim now
4059 forks a subprocess and does the per-uid checking as the relevant uid.
4061 77. Added process_log_path.
4063 78. If log_file_path was not explicitly set, a setting of check_log_space or
4064 check_log_inodes was ignored.
4066 79. If a space check for the spool or log partitions fails, the incident is now
4067 logged. Of course, in the latter case the data may get lost...
4069 80. Added the %p formatting code to string_format() so that it can be used to
4070 print addresses in debug_print(). Adjusted all the address printing in the
4071 debugging in store.c to use %p rather than %d.
4073 81. There was a concern that a line of code in smtp_in.c could overflow a
4074 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4075 initially expressed, the concern was not well-founded, because trailing
4076 spaces are removed early. However, if the trailing spaces were followed by
4077 a NULL, they did not get removed, so the overflow was possible. Two fixes
4080 (a) I re-wrote the offending code in a cleaner fashion.
4081 (b) If an incoming SMTP command contains a NULL character, it is rejected
4084 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4085 runs initgroups(), so that if the Exim user is in any additional groups,
4086 they will be used during message reception.
4092 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob