1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.385 2006/07/28 11:39:11 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 ${quote_pgsql to quote ' as '' instead of \' because of a
60 security issue with \' (bugzilla #107). I could not use the
61 PQescapeStringConn() function, because it needs a PGconn value as one of
64 PH/08 When testing addresses using -bt, indicate those final addresses that
65 are duplicates that would not cause an additional delivery. At least one
66 person was confused, thinking that -bt output corresponded to deliveries.
67 (Suppressing duplicates isn't a good idea as you lose the information
68 about possibly different redirections that led to the duplicates.)
70 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
71 systems where poll() doesn't work, in particular OS X.
73 PH/10 Added more information to debugging output for retry time not reached.
75 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
76 operations in malware.c.
78 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
81 PH/13 If write_rejectlog was set false when logging was sent to syslog with
82 syslog_duplication set false, log lines that would normally be written
83 both the the main log and to the reject log were not written to syslog at
86 PH/14 In the default configuration, change the use of "message" in ACL warn
87 statements to "add_header".
89 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
90 not followed by a command (e.g. "seen endif").
92 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
93 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
96 PH/17 Added extra conditions to the default value of delay_warning_condition
100 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
101 { match{$h_precedence:}{(?i)bulk|list|junk} } \
102 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
105 The Auto-Submitted: and various List- headers are standardised, whereas I
106 don't think Precedence: ever was.
108 PH/18 Refactored debugging code in route_finduser() to show more information,
109 in particular, the error code if getpwnam() issues one.
111 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
112 This is apparently needed in addition to the PH/07 change above to avoid
113 any possible encoding problems.
115 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
116 but not after initializing Perl.
118 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
119 output them only if debugging. By default they are written stderr,
120 apparently, which is not desirable.
122 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
125 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
128 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
130 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
131 authenticated or an ident call has been made. Suppress the default
132 values for $authenticated_id and $authenticated_sender (but permit -oMai
133 and -oMas) when testing with -bh.
135 PH/24 Re-jigged the order of the tests in the default configuration so that the
136 tests for valid domains and recipients precede the DNS black list and CSA
137 tests, on the grounds that those ones are more expensive.
139 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
140 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
141 This bug exists in every version of Exim that I still have, right back to
144 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
145 However, an attempt to turn on case-sensitivity in a regex key by
146 including (?-i) didn't work because the subject string was already
147 lowercased, and the effects were non-intuitive. It turns out that a
148 one-line patch can be used to allow (?-i) to work as expected.
154 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
155 other effects) broke the use of negated acl sub-conditions.
157 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
160 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
161 "Deny" causes Exim to reject the incoming connection with a 554 error.
162 Unfortunately, if there is a major crisis, such as a disk failure,
163 tcp-wrappers gives "deny", whereas what one would like would be some
164 kind of temporary error. A kludge has been added to help with this.
165 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
166 554 error is used if errno is still zero or contains ENOENT (which occurs
167 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
170 PH/03 Add -lutil to the default FreeBSD LIBS setting.
172 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
173 errors. Otherwise a message that provokes a temporary error (when other
174 messages do not) can cause a whole host to time out.
176 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
177 addresses were routed directly to files or pipes from a redirect router.
178 File deliveries just didn't batch; pipe deliveries might have suffered
181 PH/06 A failure to get a lock for a hints database would erroneously always say
182 "Failed to get write lock", even when it was really a read lock.
184 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
185 of 0600. This has been changed to use the value of the lockfile_mode
186 option (which defaults to 0600).
188 PH/08 Applied small patch from the Sieve maintainer.
190 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
191 folder from quota calculations, a direct delivery into this folder messed
192 up the contents of the maildirsize file. This was because the regex was
193 used only to exclude .Trash (or whatever) when the size of the mailbox
194 was calculated. There was no check that a delivery was happening into an
195 excluded directory. This bug has been fixed by ignoring all quota
196 processing for deliveries into excluded directories.
198 PH/10 Added the maildirfolder_create_regex option to appendfile.
204 PH/01 The code for finding all the local interface addresses on a FreeBSD
205 system running IPv6 was broken. This may well have applied to all BSD
206 systems, as well as to others that have similar system calls. The broken
207 code found IPv4 interfaces correctly, but gave incorrect values for the
208 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
209 that it would not match correctly against @[] and not recognize the IPv6
212 PH/02 The ipliteral router was not recognizing addresses of the form user@
213 [ipv6:....] because it didn't know about the "ipv6:" prefix.
215 PH/03 Added disable_ipv6.
217 PH/04 Changed $reply_address to use the raw form of the headers instead of the
218 decoded form, because it is most often used to construct To: headers
219 lines in autoreplies, and the decoded form may well be syntactically
220 invalid. However, $reply_address has leading white space removed, and all
221 newlines turned into spaces so that the autoreply transport does not
224 PH/05 If group was specified without a user on a router, and no group or user
225 was specified on a transport, the group from the router was ignored.
227 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
228 for visible compile-time settings that can be used to change these
229 numbers, for those that want even more. Backwards compatibility with old
230 spool files has been maintained. However, going back to a previous Exim
231 release will lost any variables that are in spool files.
233 PH/07 Two small changes when running in the test harness: increase delay when
234 passing a TCP/IP connection to a new process, in case the original
235 process has to generate a bounce, and remove special handling of
236 127.0.0.2 (sic), which is no longer necessary.
238 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
239 be the same on different OS.
241 PH/09 Moved a debug statement in filter processing to avoid a race problem when
244 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
245 whether --show-vars was specified or not
247 JJ/02 exipick: Added support for new ACL variable spool format introduced
250 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
251 syntactically invalid From: or Reply-to: line, and a filter used this to
252 generate an autoreply, and therefore failed to obtain an address for the
253 autoreply, Exim could try to deliver to a non-existent relative file
254 name, causing unrelated and misleading errors. What now happens is that
255 it logs this as a hard delivery error, but does not attempt to create a
258 PH/11 The exinext utility has a -C option for testing purposes, but although
259 the given file was scanned by exinext itself; it wasn't being passed on
260 when Exim was called.
262 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
263 an end-of-file indication when reading a command response.
265 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
266 compiled. In many other places in Exim, IPv6 addresses are always
267 recognized, so I have changed this. It also means that IPv4 domain
268 literals of the form [IPV4:n.n.n.n] are now always recognized.
270 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
271 used if the router is not running as root, for example, when verifying at
272 ACL time, or when using -bh. The debugging output from this situation was
273 non-existent - all you got was a failure to exec. I have made two
276 (a) Failures to set uid/gid, the current directory, or a process leader
277 in a subprocess such as that created by queryprogram now generate
278 suitable debugging ouput when -d is set.
280 (b) The queryprogram router detects when it is not running as root,
281 outputs suitable debugging information if -d is set, and then runs
282 the subprocess without attempting to change uid/gid.
284 PH/15 Minor change to Makefile for building test_host (undocumented testing
287 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
288 additional section of a DNS packet that returns MX or SRV records.
289 Instead, it always explicitly searches for A/AAAA records. This avoids
290 major problems that occur when a DNS server includes only records of one
291 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
292 fixed another bug: if SRV records were looked up and the corresponding
293 address records were *not* found in the additional section, the port
294 values from the SRV records were lost.
296 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
297 using the correct key (the original address) when searching the retry
298 rules in order to find which one to use for generating the retry hint.
300 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
301 adding the default one. Similarly, if it contains a Reply-To: header, the
302 errors_reply_to option, if set, is not used.
304 PH/19 When calculating a retry time, Exim used to measure the "time since
305 failure" by looking at the "first failed" field in the retry record. Now
306 it does not use this if it is later than than the arrival time of the
307 message. Instead it uses the arrival time. This makes for better
308 behaviour in cases where some deliveries succeed, thus re-setting the
309 "first failed" field. An example is a quota failure for a huge message
310 when small messages continue to be delivered. Without this change, the
311 "time since failure" will always be short, possible causing more frequent
312 delivery attempts for the huge message than are intended.
313 [Note: This change was subsequently modified - see PH/04 for 4.62.]
315 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
316 $1, $2, $3) because the numerical variables can be reset during some
317 expansion items (e.g. "match"), thereby losing the authentication data.
319 PH/21 Make -bV show the size of off_t variables so that the test suite can
320 decide whether to run tests for quotas > 2G.
322 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
323 mailbox_size, and mailbox_filecount in the appendfile transport. If a
324 filecount value is greater than 2G or if a quota value is greater than 2G
325 on a system where the size of off_t is not greater than 4, a panic error
328 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
329 never match. The debug and -bh output now contains an explicit error
330 message indicating a malformed IPv4 address or mask.
332 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
333 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
336 PH/25 Do not write to syslog when running in the test harness. The only
337 occasion when this arises is a failure to open the main or panic logs
338 (for which there is an explicit test).
340 PH/26 Added the /no_tell option to "control=freeze".
342 PH/27 If a host name lookup failed very early in a connection, for example, if
343 the IP address matched host_lookup and the reverse lookup yielded a name
344 that did not have a forward lookup, an error message of the form "no IP
345 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
346 could be logged. Now it outputs the IP address instead of "NULL".
348 PH/28 An enabling patch from MH: add new function child_open_exim2() which
349 allows the sender and the authenticated sender to be set when
350 submitting a message from within Exim. Since child_open_exim() is
351 documented for local_scan(), the new function should be too.
353 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
354 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
355 results in an empty string is now treated as unset.
357 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
359 PH/31 Added sender_verify_fail logging option.
361 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
362 needed by RFC 822 but not by RFC 2822 was commented out. I have now
363 tidied the source and removed it altogether.
365 PH/33 When a queue run was abandoned because the load average was too high, a
366 log line was always written; now it is written only if the queue_run log
367 selector is set. In addition, the log line for abandonment now contains
368 information about the queue run such as the pid. This is always present
369 in "start" and "stop" lines but was omitted from the "abandon" line.
371 PH/34 Omit spaces between a header name and the colon in the error message that
372 is given when verify = headers_syntax fails (if there are lots of them,
373 the message gets confusing).
375 PH/35 Change the default for dns_check_names_pattern to allow slashes within
376 names, as there are now some PTR records that contain slashes. This check
377 is only to protect against broken name servers that fall over on strange
378 characters, so the fact that it applies to all lookups doesn't matter.
380 PH/36 Now that the new test suite is complete, we can remove some of the
381 special code in Exim that was needed for the old test suite. For example,
382 sorting DNS records because real resolvers return them in an arbitrary
383 order. The new test suite's fake resolver always returns records in the
386 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
387 bounces) except when queue_only is set, to avoid logging races between
388 the different processes.
390 PH/38 Panic-die if .include specifies a non-absolute path.
392 PH/39 A tweak to the "H" retry rule from its user.
394 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
395 a label. They prevented compilation on older perls.
397 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
398 a warning to be raised on newish perls.
400 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
401 on queue. Changes to match documented behaviour of showing count of
402 messages matching specified criteria.
404 PH/40 Changed the default ident timeout from 30s to 5s.
406 PH/41 Added support for the use of login_cap features, on those BSD systems
407 that have them, for controlling the resources used by pipe deliveries.
409 PH/42 The content-scanning code uses fopen() to create files in which to put
410 message data. Previously it was not paying any attention to the mode of
411 the files. Exim runs with umask(0) because the rest of the code creates
412 files with open(), and sets the required mode explicitly. Thus, these
413 files were ending up world-writeable. This was not a big issue, because,
414 being within the spool directory, they were not world-accessible. I have
415 created a function called modefopen, which takes an additional mode
416 argument. It sets umask(777), creates the file, chmods it to the required
417 mode, then resets the umask. All the relevant calls to fopen() in the
418 content scanning code have been changed to use this function.
420 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
421 to 24 hours. This avoids potential overflow problems when processing G
422 and H retry rules. I suspect nobody ever tinkers with this value.
424 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
426 PH/45 When the plaintext authenticator is running as a client, the server's
427 challenges are checked to ensure they are valid base64 strings. By
428 default, the authentication attempt is cancelled if an invalid string is
429 received. Setting client_ignore_invalid_base64 true ignores these errors.
430 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
431 they are received. Thus, the responses can be made to depend on the
432 challenges. If an invalid string is ignored, an empty string is placed in
435 PH/46 Messages that are created by the autoreply transport now contains a
436 References: header, in accordance with RFCs 2822 and 3834.
438 PH/47 Added authenticated_sender_force to the smtp transport.
440 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
442 PH/49 Installed latest patch from the Sieve maintainer.
444 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
445 was also set, the appendfile transport was unnecessarily scanning a
446 directory of message files (e.g. for maildir delivery) to find the count
447 of files (along with the size), even though it did not need this
448 information. It now does the scan only if it needs to find either the
449 size of the count of files.
451 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
453 PH/52 Two bugs concerned with error handling when the smtp transport is
456 (i) Exim was not creating retry information for temporary errors given
457 for individual recipients after the DATA command when the smtp transport
458 was used in LMTP mode. This meant that they could be retried too
459 frequently, and not timed out correctly.
461 (ii) Exim was setting the flag that allows error details to be returned
462 for LMTP errors on RCPT commands, but not for LMTP errors for individual
463 recipients that were returned after the DATA command.
465 PH/53 This is related to PH/52, but is more general: for any failing address,
466 when detailed error information was permitted to be returned to the
467 sender, but the error was temporary, then after the final timeout, only
468 "retry timeout exceeded" was returned. Now it returns the full error as
469 well as "retry timeout exceeded".
471 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
472 do this, and (what is worse) MTAs that accept it.
474 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
475 will now be deprecated.
477 PH/56 New os.c-cygwin from the Cygwin maintainer.
479 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
480 formats (previously only available in exim formats via -bpr, -bpru,
481 and -bpra. Now also available in native and exiqgrep formats)
483 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
484 with very large, slow to parse queues
486 JJ/08 exipick: added ! as generic prefix to negate any criteria format
488 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
490 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
491 responses to authentication challenges, though it was showing the
492 challenges; (ii) I've removed the CR characters from the debug output for
495 PH/58 Allow for the insertion of a newline as well as a space when a string
496 is turned into more than one encoded-word during RFC 2047 encoding. The
497 Sieve code now uses this.
499 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
500 data_4xx, lost_connection, tls_required.
502 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
503 message was being sent as an SMTP response.
505 PH/61 Add -l and -k options to exicyclog.
507 PH/62 When verifying, if an address was redirected to one new address, so that
508 verification continued, and the new address failed or deferred after
509 having set something in $address_data, the value of $address_data was not
510 passed back to the ACL. This was different to the case when no
511 redirection occurred. The value is now passed back in both cases.
513 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
514 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
515 use login_cap.h, so on its own it isn't the distinguishing feature. The
516 new name refers directly to the setclassresources() function.
518 PH/65 Added configuration files for NetBSD3.
520 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
522 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
523 is preferred over IPv4.
525 PH/68 The bounce_return_message and bounce_return_body options were not being
526 honoured for bounces generated during the reception of non-SMTP messages.
527 In particular, this applied to messages rejected by the ACL. This bug has
528 been fixed. However, if bounce_return_message is true and bounce_return_
529 body is false, the headers that are returned for a non-SMTP message
530 include only those that have been read before the error was detected.
531 (In the case of an ACL rejection, they have all been read.)
533 PH/69 The HTML version of the specification is now built in a directory called
534 spec_html instead of spec.html, because the latter looks like a path with
535 a MIME-type, and this confuses some software.
537 PH/70 Catch two compiler warnings in sieve.c.
539 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
540 function verify_get_ident() calls ip_connect() to connect a socket, but
541 if the "connect()" function timed out, ip_connect() used to close the
542 socket. However, verify_get_ident() also closes the socket later, and in
543 between Exim writes to the log, which may get opened at this point. When
544 the socket was closed in ip_connect(), the log could get the same file
545 descriptor number as the socket. This naturally causes chaos. The fix is
546 not to close the socket in ip_connect(); the socket should be closed by
547 the function that creates it. There was only one place in the code where
548 this was missing, in the iplookup router, which I don't think anybody now
549 uses, but I've fixed it anyway.
551 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
552 well as to direct DNS lookups. Otherwise the handling of names in host
553 lists is inconsistent and therefore confusing.
559 PH/01 Two changes to the default runtime configuration:
561 (1) Move the checks for relay_from_hosts and authenticated clients from
562 after to before the (commented out) DNS black list checks.
564 (2) Add control=submission to the relay_from_hosts and authenticated
565 clients checks, on the grounds that messages accepted by these
566 statements are most likely to be submissions.
568 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
570 (1) Generate an error if the third argument for the ${prvs expansion is
573 (2) Treat a missing third argument of ${prvscheck as if it were an empty
576 (3) Reset the variables that are obtained from the first argument of
577 ${prvscheck and used in the second argument before leaving the code,
578 because their memory is reclaimed, so using them afterwards may do
581 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
582 one (it's much easier than Tom thought :-).
584 (5) Because of (4), we can now allow for the use of $prvscheck_result
585 inside the third argument.
587 PH/03 For some reason, the default setting of PATH when running a command from
588 a pipe transport was just "/usr/bin". I have changed it to
591 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
592 anything to be listed in the output from -bV.
594 PH/05 When a filter generated an autoreply, the entire To: header line was
595 quoted in the delivery log line, like this:
597 => >A.N.Other <ano@some.domain> <original@ddress> ...
599 This has been changed so that it extracts the operative address. There
600 may be more than one such address. If so, they are comma-separated, like
603 => >ano@some.domain,ona@other.domain <original@ddress> ...
605 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
606 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
607 not being looked up in the rDNS to get a host name, Exim was showing the
608 IP address twice in Received: lines, even though the IP addresses were
609 identical. For example:
611 Received: from [1.2.3.4] (helo=[1.2.3.4])
613 However, if the real host name was known, it was omitting the HELO data
614 if it matched the actual IP address. This has been tidied up so that it
615 doesn't show the same IP address twice.
617 PH/07 When both +timestamp and +memory debugging was on, the value given by
618 $tod_xxx expansions could be wrong, because the tod_stamp() function was
619 called by the debug printing, thereby overwriting the timestamp buffer.
620 Debugging no longer uses the tod_stamp() function when +timestamp is set.
622 PH/08 When the original message was included in an autoreply transport, it
623 always said "this is a copy of the message, including all the headers",
624 even if body_only or headers_only was set. It now gives an appropriate
627 PH/09 Applied a patch from the Sieve maintainer which:
629 o fixes some comments
630 o adds the (disabled) notify extension core
631 o adds some debug output for the result of if/elsif tests
632 o points to the current vacation draft in the documentation
633 and documents the missing references header update
637 o fixes a bug in processing the envelope test (when testing
638 multiple envelope elements, the last element determinted the
641 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
642 Electronic Mail") by including:
644 Auto-submitted: auto-generated
646 in the messages that it generates (bounce messages and others, such as
647 warnings). In the case of bounce messages for non-SMTP mesages, there was
648 also a typo: it was using "Auto_submitted" (underscore instead of
649 hyphen). Since every message generated by Exim is necessarily in response
650 to another message, thes have all been changed to:
652 Auto-Submitted: auto-replied
654 in accordance with these statements in the RFC:
656 The auto-replied keyword:
658 - SHOULD be used on messages sent in direct response to another
659 message by an automatic process,
661 - MUST NOT be used on manually-generated messages,
663 - MAY be used on Delivery Status Notifications (DSNs) and Message
664 Disposition Notifications (MDNs),
666 - MUST NOT be used on messages generated by automatic or periodic
667 processes, except for messages which are automatic responses to
670 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
671 to the default Received: header definition.
673 PH/12 Added log selector acl_warn_skipped (default on).
675 PH/13 After a successful wildlsearch lookup, discard the values of numeric
676 variables because (a) they are in the wrong storage pool and (b) even if
677 they were copied, it wouldn't work properly because of the caching.
679 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
680 checking when decoding. Apparently there are clients that generate
681 overlong encoded strings. Why am I not surprised?
683 PH/15 If the first argument of "${if match_address" was not empty, but did not
684 contain an "@" character, Exim crashed. Now it writes a panic log message
685 and treats the condition as false.
687 PH/16 In autoreply, treat an empty string for "once" the same as unset.
689 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
690 extension "envelope-auth". The code is finished and in agreement with
691 other implementations, but there is no documentation so far and in fact,
692 nobody wrote the draft yet. This extension is currently #undef'ed, thus
693 not changing the active code.
695 Print executed "if" and "elsif" statements when debugging is used. This
696 helps a great deal to understand what a filter does.
698 Document more things not specified clearly in RFC3028. I had all this
699 sorted out, when out of a sudden new issues came to my mind. Oops."
701 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
704 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
705 canonical form (as documented). However, after a host name lookup from
706 the IP address, check_host() was doing a simple string comparison with
707 addresses acquired from the DNS when checking that the found name did
708 have the original IP as one of its addresses. Since any found IPv6
709 addresses are likely to be in abbreviated form, the comparison could
710 fail. Luckily, there already exists a function for doing the comparison
711 by converting both addresses to binary, so now that is used instead of
714 PH/20 There was another similar case to PH/19, when a complete host name was
715 given in a host list; looking up its IP address could give an abbreviated
716 form, whereas the current host's name might or might not be abbreviated.
717 The same fix has been applied.
723 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
724 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
727 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
728 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
730 PH/03 Typo: missing ".o" in src/pcre/Makefile.
732 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
733 header line, restrict the check to what is listed in RFCs 2369 and 2929.
734 Also, for "Auto-Submitted", treat anything other than "no" as
735 non-personal, in accordance with RFC 3834. (Previously it treated
736 anything starting "auto-" as non-personal.)
738 TF/01 The control=submission/name=... option had a problem with syntax
739 errors if the name included a slash character. The /name= option
740 now slurps the rest of the string, so it can include any characters
741 but it must come last in the list of options (after /sender_retain
744 PH/05 Some modifications to the interface to the fake nameserver for the new
752 TK/01 Added the "success_on_redirect" address verification option. See
753 NewStuff for rationale and an example.
755 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
757 PH/02 Patch to exigrep to allow it to work on syslog lines.
759 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
760 fread() to skip over the body file's header line, because in Cygwin the
761 header line is locked and is inaccessible.
763 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
764 co-exist for some time) to make it clear that it is the Exim ID that is
765 referenced, not the Message-ID: header line.
767 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
768 string_format() function, because snprintf() does not exist on all
771 PH/06 The use of forbid_filter_existstest now also locks out the use of the
772 ${stat: expansion item.
774 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
775 protocol synchronization error", to keep the pedants happy.
777 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
778 well as for IRIX systems, when gcc is being used. See the host.c source
781 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
783 PH/10 Named domain lists were not working if used in a queue_smtp_domains
786 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
787 transport and to the smtp transport in LMTP mode.
789 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
791 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
792 run a filter in a subprocess. This could lead to confusion in subsequent
793 lookups in the parent process. There should also be a search_tidyup() at
794 the end of the subprocess.
796 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
797 only if the host matched helo_try_verify_hosts, which caused the
798 verification to occur when the EHLO/HELO command was issued. The ACL just
799 tested the remembered result. Now, if a previous verification attempt has
800 not happened, "verify = helo" does it there and then.
802 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
804 TK/03 Fix log output including CR from clamd.
806 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
807 exist provoked a memory error which could cause a segfault.
809 PH/15 Installed PCRE 6.2
811 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
813 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
814 of the problem. Specifically, suggested +O2 rather than +O1 for the
817 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
819 PH/20 If a delivery was routed to a non-standard port by means of an SRV
820 record, the port was not correctly logged when the outgoing_port log
821 selector was set (it logged the transort's default port).
823 PH/21 Added support for host-specific ports to manualroute, queryprogram,
824 fallback_hosts, and "hosts" in the smtp transport.
826 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
827 host errors such as "Connection refused".
829 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
830 authentication with radiusclient 0.4.9:
832 - Error returned from rc_read_config was caught wrongly
833 - Username/password not passed on to radius server due to wrong length.
835 The presumption is that some radiusclient API changes for 4.51/PH/17
836 were not taken care of correctly. The code is still untested by me (my
837 Linux distribution still has 0.3.2 of radiusclient), but it was
838 contributed by a Radius user.
840 PH/24 When doing a callout, the value of $domain wasn't set correctly when
841 expanding the "port" option of the smtp transport.
843 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
844 while reading a MIME header. Thanks to Tom Hughes for a patch.
846 PH/24 Include config.h inside local_scan.h so that configuration settings are
849 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
850 that in an ACL for RCPT (for example), you can examine exactly what was
853 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
854 commands, but it was not correctly comparing the address with the actual
855 client host address. Thus, it would show the EHLO address in Received:
856 header lines when this was not necessary.
858 PH/27 Added the % operator to ${eval:}.
860 PH/28 Exim tries to create and chdir to its spool directory when it starts;
861 it should be ignoring failures (because with -C, for example, it has lost
862 privilege). It wasn't ignoring creation failures other than "already
865 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
868 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
869 because an input line was too long, either on its own, or by virtue of
870 too many continuations, the temporary file was not being removed, and the
871 return code was incorrect.
873 PH/31 Missing "BOOL" in function definition in filtertest.c.
875 PH/32 Applied Sieve patches from the maintainer.
877 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
879 PH/33 Added "verify = not_blind".
881 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
882 Local/Makefile (with some defaults set). These are used in built scripts
883 such as exicyclog, but they have never been used in the exim_install
884 script (though there are many overriding facilities there). I have
885 arranged that the exim_install script now takes note of these two
888 PH/35 Installed configuration files for Dragonfly.
890 PH/36 When a locally submitted message by a trusted user did not contain a
891 From: header, and the sender address was obtained from -f or from an SMTP
892 MAIL command, and the trusted user did not use -F to supply a sender
893 name, $originator_name was incorrectly used when constructing a From:
894 header. Furthermore, $originator_name was used for submission mode
895 messages from external hosts without From: headers in a similar way,
896 which is clearly wrong.
898 PH/37 Added control=suppress_local_fixups.
900 PH/38 When log_selector = +received_sender was set, and the addition of the
901 sender made the log line's construction buffer exactly full, or one byte
902 less than full, an overflow happened when the terminating "\n" was
905 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
906 when the result of a list match is failure because a DNS lookup failed.
908 PH/40 RM_COMMAND is now used in the building process.
910 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
911 the "build-* directories that it finds.
913 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
914 address in a domain literal was a prefix of an interface address.
916 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
917 when verifying a sender address, unless rewrite_headers is false.
919 PH/44 Wrote a long comment about why errors_to addresses are verified as
920 recipients, not senders.
922 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
923 the ratelimit ACL was added.
925 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
927 PH/46 Added extra information about PostgreSQL errors to the error string.
929 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
930 avoiding the need to install special zones in a real server. This is
931 backwards compatible; if it can't find the fake resolver, it drops back.
932 Thus, both old and new test suites can be run.
934 TF/02 Added util/ratelimit.pl
936 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
937 clock is set back in time.
939 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
940 Candler <B.Candler@pobox.com>.
942 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
943 OK for addresses that are the result of redirections.
945 PH/48 A number of further additions for the benefit of the new test suite,
946 including a fake gethostbyname() that interfaces to the fake DNS resolver
949 TF/06 The fix for widen_domains has also been applied to qualify_single and
950 search_parents which are the other dnslookup options that can cause
953 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
956 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
958 TF/07 Exim produced the error message "an SRV record indicated no SMTP
959 service" if it encountered an MX record with an empty target hostname.
960 The message is now "an MX or SRV record indicated no SMTP service".
962 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
963 if the DNS of the sending site is misconfigured. This is quite a
964 common situation. This change restores the behaviour of treating a
965 helo verification defer as a failure.
967 PH/51 If self=fail was set on a router, the bounce message did not include the
968 actual error message.
974 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
976 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
977 command itself ended in error, the underlying message about the transport
978 filter timeout was being overwritten with the pipe command error. Now the
979 underlying error message should be appended to the second error message.
981 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
983 PH/02 Reduce the amount of output that "make" produces by default. Full output
984 can still be requested.
986 PH/03 The warning log line about a condition test deferring for a "warn" verb
987 was being output only once per connection, rather than after each
988 occurrence (because it was using the same function as for successful
989 "warn" verbs). This seems wrong, so I have changed it.
991 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
992 it should not have, which might have caused a crash in the right
993 circumstances, but probably never did.
995 PH/04 Installed a modified version of Tony Finch's patch to make submission
996 mode fix the return path as well as the Sender: header line, and to
997 add a /name= option so that you can make the user's friendly name appear
1000 TF/03 Added the control = fakedefer ACL modifier.
1002 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1003 Mark Lowes for thorough testing.
1005 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1007 TK/03 Merged latest SRS patch from Miles Wilton.
1009 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1010 with the definition in sysexits.h (which is #included earlier).
1011 Fortunately, Exim does not actually use EX_OK. The code used to try to
1012 preserve the sysexits.h value, by assumimg that macro definitions were
1013 scanned for macro replacements. I have been disabused of this notion,
1014 so now the code just undefines EX_OK before #including unistd.h.
1016 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1017 in the smtp transport. When a block could not be written in a single
1018 write() function, the timeout was being re-applied to each part-write.
1019 This seems wrong - if the receiver was accepting one byte at a time it
1020 would take for ever. The timeout is now adjusted when this happens. It
1021 doesn't have to be particularly precise.
1023 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1024 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1026 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1027 without a domain if the check to <postmaster@domain> fails.
1029 SC/01 Eximstats: added -xls and the ability to specify output files
1030 (patch written by Frank Heydlauf).
1032 SC/02 Eximstats: use FileHandles for outputing results.
1034 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1036 SC/04 Eximstats: fixed display of large numbers with -nvr option
1038 SC/05 Eximstats: fixed merging of reports with empty tables.
1040 SC/06 Eximstats: added the -include_original_destination flag
1042 SC/07 Eximstats: removed tabs and trailing whitespace.
1044 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1046 TK/06 MBOX spool code: Add real "From " MBOX separator line
1047 so the .eml file is really in mbox format (even though
1048 most programs do not really care). Patch from Alex Miller.
1050 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1051 The latter is generated from $received_to and is only set if the
1052 message has one envelope recipient. SA can use these headers,
1053 obviously out-of-the-box. Patch from Alex Miller.
1055 PH/08 The ${def test on a variable was returning false if the variable's
1056 value was "0", contrary to what the specification has always said!
1057 The result should be true unless the variable is empty.
1059 PH/09 The syntax error of a character other than { following "${if
1060 def:variable_name" (after optional whitespace) was not being diagnosed.
1061 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1062 accidental colon was present, for example, could give incorrect results.
1064 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1065 result is used (not including appendfile, where other changes are about
1068 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1069 This involved changing a lot of size variables from int to off_t. It
1070 should work with maildirs and everything.
1072 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1073 spamd dying while we are connected to it.
1075 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1078 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1079 function for BDB 4.3.
1081 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1083 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1084 the sources for PCRE have been re-arranged and more files are now
1087 PH/15 The code I had for printing potentially long long variables in PH/11
1088 above was not the best (it lost precision). The length of off_t variables
1089 is now inspected at build time, and an appropriate printing format (%ld
1090 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1091 to be "long long int" or "long int". This is needed for the internal
1092 formatting function string_vformat().
1094 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1095 the configuration file to be ":syslog", then the script "guesses" where
1096 the logs files are, rather than using the compiled in default. In our
1097 case the guess is not the same as the compiled default, so the script
1098 suddenly stopped working when I started to use syslog. The patch checks
1099 to see if log_file_path is "". If so, it attempts to read it from exim
1100 with no configuration file to get the compiled in version, before it
1101 falls back to the previous guessing code."
1103 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1104 implementing BATV in an Exim configuration. See NewStuff for the gory
1107 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1108 Makefile that are specific to HP-UX.
1110 PH/18 If the "use_postmaster" option was set for a recipient callout together
1111 with the "random" option, the postmaster address was used as the MAIL
1112 FROM address for the random test, but not for the subsequent recipient
1113 test. It is now used for both.
1115 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1116 patch removes a few documentation additions to RFC 3028, because the
1117 latest draft now contains them. It adds the new en;ascii-case comparator
1118 and a new error check for 8bit text in MIME parts. Comparator and
1119 require names are now matched exactly. I enabled the subaddress
1120 extension, but it is not well tested yet (read: it works for me)."
1122 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1123 rework some of the code of TK/09 above to avoid the hardwired use of
1124 "%lld" and "long long". Replaced the call to snprintf() with a call to
1127 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1128 records point to non-existent hosts", "retry timeout exceeded", and
1129 "retry time not reached for any host after a long failure period".
1131 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1132 experimental DomainKeys support:
1134 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1135 (2) On an error such as an illegally used "control", the wrong name for
1136 the control was given.
1138 These problems did NOT occur unless DomainKeys support was compiled.
1140 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1142 PH/24 Added ${if match_ip condition.
1144 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1145 databases so that it will be absolutely obvious if a crash occurs in the
1146 DB library. This is a regular occurrence (often caused by mis-matched
1149 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1150 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1151 picked up on a user's system that detects such things. There doesn't seem
1152 to be a gcc warning option for this - only an attribute that has to be
1153 put on the function's prototype. It seems that in Fedora Core 4 they have
1154 set this on a number of new functions. No doubt there will be more in due
1157 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1158 specify a transport. However, if an address that was verified by such a
1159 router was the subject of a callout, Exim crashed because it tried to
1160 read the rcpt_include_affixes from the non-existent transport. Now it
1161 just assumes that the setting of that option is false. This bug was
1162 introduced by 4.51/PH/31.
1164 PH/28 Changed -d+all to exclude +memory, because that information is very
1165 rarely of interest, but it makes the output a lot bigger. People tend to
1166 do -d+all out of habit.
1168 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1169 code in os-type was giving problems when libc.so lives in lib64, like on
1172 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1173 aren't the modern standard, and it seems that some systems' include files
1174 don't always have them. Exim was already checking for some of the newer
1175 ones like T_AAAA, and defining it itself. I've added checks for all the
1176 record types that Exim uses.
1178 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1179 not automatically generating a new one, as it is supposed to. This
1180 prevented TLS from working. If the file did exist, but contained invalid
1181 data, a new version was generated, as expected. It was only the case of a
1182 non-existent file that was broken.
1184 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1185 with a change in libdomainkeys > 0.64.
1187 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1188 from DNS. If the selector record carries the flag, it now has
1189 precedence over the domain-wide flag.
1191 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1193 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1194 the use of a port name that isn't defined in /etc/services) occurred, the
1195 message was deferred as in a normal delivery, and thus remained on the
1196 spool, instead of being failed because of the mua_wrapper setting. This
1197 is now fixed, and I tidied up some of the mua_wrapper messages at the
1200 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1201 the messages in a hash of arrays rather than using individual hashes.
1202 This is a bit cleaner and results in dramatic memory savings, albeit
1203 at a slight CPU cost.
1205 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1206 as requested by Marc Sherman.
1208 SC/10 Eximstats: added histograms for user specified patterns as requested
1211 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1213 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1214 fopen() in the content-scanning modules that did not already have it.
1220 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1221 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1223 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1225 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1227 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1228 to test Sieve filters that use "vacation".
1230 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1231 that changes the way the GnuTLS parameters are stored in the cache file.
1232 The new format can be generated externally. For backward compatibility,
1233 if the data in the cache doesn't make sense, Exim assumes it has read an
1234 old-format file, and it generates new data and writes a new file. This
1235 means that you can't go back to an older release without removing the
1238 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1239 work if there are any delivery delays because "one_time" forces the
1240 parent to be marked "delivered", so its unseen clone is never tried
1241 again. For this reason, Exim now forbids the simultaneous setting of
1244 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1245 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1246 entry if you want to know the details. The fix, however, neglected to
1247 consider the case when local delivery batching is involved. The test for
1248 "previously delivered" was not happening when checking to see if an
1249 address could be batched with a previous (undelivered) one; under
1250 certain circumstances this could lead to multiple deliveries to the same
1253 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1254 in its include files, and this causes problems building Exim.
1256 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1257 header_syntax) but Exim was just ignoring anything given after a slash.
1258 In particular, this caused confusion with an attempt to use "verify =
1259 reverse_host_lookup/defer_ok". An error is now given when options are
1260 supplied for verify items that do not have them. (Maybe reverse_host_
1261 lookup should have a defer_ok option, but that's a different point.)
1263 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1264 defined by RFC 821) to 2048, because there were problems with some AUTH
1265 commands, and RFC 1869 says the size should be increased for extended
1266 SMTP commands that take arguments.
1268 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1271 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1272 "unknown" error; now it says that the functionality isn't in the binary.
1274 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1275 an address' error message when a string expansion fails (syntax or
1276 whatever). Otherwise the password may appear in the log. Following change
1277 PH/42 below, there is no longer a chance of it appearing in a bounce
1280 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1282 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1283 of its addresses was ever tried. (Bugzilla bug #2.)
1285 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1286 the result incorrectly in the debug output. (It correctly added a newline
1287 to what was transported.)
1289 TF/01 Added $received_time.
1291 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1292 commented out examples of how to interface to a virus scanner and to
1293 SpamAssassin. Also added commented examples of av_scanner and
1294 spamd_address settings.
1296 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1297 and controls are allowed in which ACLs. There were a couple of minor
1298 errors. Some of the entries in the conditions table (which is a table of
1299 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1300 negation of where the condition IS allowed.
1302 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1304 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1305 header file does not have a version number, so I've had to invent a new
1306 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1307 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1308 radiusclient), but it was contributed by a Radius user.
1310 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1311 files or directories, for OpenSSL.
1313 PH/19 When an Exim process that is running as root has to create an Exim log
1314 file, it does so in a subprocess that runs as exim:exim so as to get the
1315 ownership right at creation (otherwise, other Exim processes might see
1316 the file with the wrong ownership). There was no test for failure of this
1317 fork() call, which would lead to the process getting stuck as it waited
1318 for a non-existent subprocess. Forks do occasionally fail when resources
1319 run out. I reviewed all the other calls to fork(); they all seem to check
1322 PH/20 When checking for unexpected SMTP input at connect time (before writing
1323 the banner), Exim was not dealing correctly with a non-positive return
1324 from the read() function. If the client had disconnected by this time,
1325 the result was a log entry for a synchronization error with an empty
1326 string after "input=" when read() returned zero. If read() returned -1
1327 (an event I could not check), uninitialized data bytes were printed.
1328 There were reports of junk text (parts of files, etc) appearing after
1331 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1333 PH/22 Added support for macro redefinition, and (re)definition in between
1334 driver and ACL definitions.
1336 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1337 forgetting to use the resulting value; it was using the unexpanded value.
1339 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1340 hadn't been configured. The fix is from Juergen Kreileder, who
1341 understands it better than I do:
1343 "Here's what I see happening with three configured cyrus_sasl
1344 authenticators configured (plain, login, cram-md5):
1346 On startup auth_cyrus_sasl_init() gets called for each of these.
1347 This means three calls to sasl_listmech() without a specified mech_list.
1348 => SASL tests which mechs of all available mechs actually work
1349 => three warnings about OTP not working
1350 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1352 With the patch, sasl_listmech() also gets called three times. But now
1353 SASL's mech_list option is set to the server_mech specified in the the
1354 authenticator. Or in other words, the answer from sasl_listmech()
1355 gets limited to just the mech you're testing for (which is different
1357 => the return list contains just 'plain' or 'login', 'cram-md5' or
1358 nothing depending on the value of ob->server_mech.
1360 I've just tested the patch: Authentication still works fine,
1361 unavailable mechs specified in the exim configuration are still
1362 caught, and the auth.log warnings about OTP are gone."
1364 PH/25 When debugging is enabled, the contents of the command line are added
1365 to the debugging output, even when log_selector=+arguments is not
1368 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1369 answer is "GNU", and only if the return is "GNU/something" is the answer
1372 PH/27 $acl_verify_message is now set immediately after the failure of a
1373 verification in an ACL, and so is available in subsequent modifiers. In
1374 particular, the message can be preserved by coding like this:
1376 warn !verify = sender
1377 set acl_m0 = $acl_verify_message
1379 Previously, $acl_verify_message was set only while expanding "message"
1380 and "log_message" when a very denied access.
1382 PH/28 Modified OS/os.c-Linux with
1384 -#ifndef OS_LOAD_AVERAGE
1385 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1387 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1388 nomenclature these days.)
1390 PH/29 Installed patch from the Sieve maintainer that adds the options
1391 sieve_useraddress and sieve_subaddress to the redirect router.
1393 PH/30 In these circumstances:
1394 . Two addresses routed to the same list of hosts;
1395 . First host does not offer TLS;
1396 . First host accepts first address;
1397 . First host gives temporary error to second address;
1398 . Second host offers TLS and a TLS session is established;
1399 . Second host accepts second address.
1400 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1401 and peerdn, if requested) that were in fact used only for the second
1404 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1405 attention to any local part prefix or suffix that was matched by the
1406 router that accepted the address. It now behaves in the same way as it
1407 does for delivery: the affixes are removed from the local part unless
1408 rcpt_include_affixes is set on the transport.
1410 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1411 timeout during the DATA phase of an incoming message.
1413 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1414 applied a patch sanctioned by the Sieve maintainer.
1416 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1417 the uid or gid is negative. A case of a negative gid caused this to be
1418 noticed. The fix allows for either to be negative.
1420 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1421 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1424 PH/36 The change PH/12 above was broken. Fixed it.
1426 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1427 the grounds that routing the same address twice would always produce the
1428 same answer. This might have been true once, but it is certainly no
1429 longer true now. Routing a child address may depend on the previous
1430 routing that produced that child. Some complicated redirection strategies
1431 went wrong when messages had multiple recipients, and made Exim's
1432 behaviour dependent on the order in which the addresses were given.
1434 I have moved the duplicate checking until after the routing is complete.
1435 Exim scans the addresses that are assigned to local and remote
1436 transports, and removes any duplicates. This means that more work will be
1437 done, as duplicates will always all be routed, but duplicates are
1438 presumably rare, so I don't expect this is of any significance.
1440 For deliveries to pipes, files, and autoreplies, the duplicate checking
1441 still happens during the routing process, since they are not going to be
1444 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1445 It corrects a timeout issue with spamd. This is Ian's comment: "The
1446 background is that sometimes spamd either never reads data from a
1447 connection it has accepted, or it never writes response data. The exiscan
1448 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1449 blindly assumes that writes won't block so it may never time out."
1451 PH/39 Allow G after quota size as well as K and M.
1453 PH/40 The value set for $authenticated_id in an authenticator may not contain
1454 binary zeroes or newlines because the value is written to log lines and
1455 to spool files. There was no check on this. Now the value is run through
1456 the string_printing() function so that such characters are converted to
1457 printable escape sequences.
1459 PH/41 $message_linecount is a new variable that contains the total number of
1460 lines in the message. Compare $body_linecount, which is the count for the
1463 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1464 bounce and delay warning messages, except in certain special cases, which
1467 (a) An SMTP error message from a remote host;
1468 (b) A message specified in a :fail: redirection;
1469 (c) A message specified in a "fail" command in a system filter;
1470 (d) A message specified in a FAIL return from the queryprogram router;
1471 (e) A message specified by the cannot_route_message router option.
1473 In these cases only, Exim does include the error details in bounce and
1474 warning messages. There are also a few cases where bland messages such
1475 as "unrouteable address" or "local delivery error" are given.
1477 PH/43 $value is now also set for the "else" part of a ${run expansion.
1479 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1480 being worked on, but at least Exim now implements the latest version to
1483 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1484 process to complete was treated as a delivery failure, a timeout while
1485 writing the message to the pipe was logged, but erroneously treated as a
1486 successful delivery. Such timeouts include transport filter timeouts. For
1487 consistency with the overall process timeout, these timeouts are now
1488 treated as errors, giving rise to delivery failures by default. However,
1489 there is now a new Boolean option for the pipe transport called
1490 timeout_defer, which, if set TRUE, converts the failures into defers for
1491 both kinds of timeout. A transport filter timeout is now identified in
1494 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1495 systems where "make" and "gmake" are different, calling "gmake" at top
1496 level broke things. I've arranged for the value of $(MAKE) to be passed
1497 from the Makefile to this script so that it can call the same version of
1501 A note about Exim versions 4.44 and 4.50
1502 ----------------------------------------
1504 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1505 changes of various kinds. As a consequence, a big documentation update was
1506 needed. This delayed the release for rather longer than seemed good, especially
1507 in the light of a couple of (minor) security issues. Therefore, the changes
1508 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1509 release. So 4.44 and 4.50 are in effect two different branches that both start
1512 I have left the 4.50 change log unchanged; it contains all the changes since
1513 4.43. The change log for 4.44 is below; many of its items are identical to
1514 those for 4.50. This seems to be the most sensible way to preserve the
1515 historical information.
1521 1. Minor wording change to the doc/README.SIEVE file.
1523 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1524 computation of the current number of files was incorrect.
1526 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1527 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1530 4. Give more explanation in the error message when the command for a transport
1531 filter fails to execute.
1533 5. There are several places where Exim runs a non-Exim command in a
1534 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1535 was being done only for the command run by the queryprogram router. It is
1536 now done for all such subprocesses. The other cases are: ${run, transport
1537 filters, and the commands run by the lmtp and pipe transports.
1539 6. Added CONFIGURE_GROUP build-time option.
1541 7. Some older OS have a limit of 256 on the maximum number of file
1542 descriptors. Exim was using setrlimit() to set 1000 as a large value
1543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1544 systems. I've change it so that if it can't get 1000, it tries for 256.
1546 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1547 was an oversight, and furthermore, ever since the addition of extra
1548 controls (e.g. 4.43/32), the checks on when to allow different forms of
1549 "control" were broken. There should now be diagnostics for all cases when a
1550 control that does not make sense is encountered.
1552 9. Added the /retain_sender option to "control=submission".
1554 10. $recipients is now available in the predata ACL (oversight).
1556 11. Tidy the search cache before the fork to do a delivery from a message
1557 received from the command line. Otherwise the child will trigger a lookup
1558 failure and thereby defer the delivery if it tries to use (for example) a
1559 cached ldap connection that the parent has called unbind on.
1561 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1562 of $address_data from the recipient verification was clobbered by the
1563 sender verification.
1565 13. The value of address_data from a sender verification is now available in
1566 $sender_address_data in subsequent conditions in the ACL statement.
1568 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1570 15. Added a new option "connect=<time>" to callout options, to set a different
1573 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1574 was its contents. (It was OK if the option was not defined at all.)
1576 17. A "Completed" log line is now written for messages that are removed from
1577 the spool by the -Mrm option.
1579 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1580 information about exactly what failed.
1582 19. Added -dd to debug only the daemon process.
1584 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1585 handles timeouts, both on the server side and network timeouts. Renamed the
1586 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1588 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1589 It is now set to "smtps".
1591 22. $host_address is now set to the target address during the checking of
1592 ignore_target_hosts.
1594 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1595 being passed; this would have caused $sender_host_name to have been used if
1596 matching the list had actually called for a host name (not very likely,
1597 since this list is usually IP addresses). A host name is now passed as
1600 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1601 code that specifies a non-restarting handler (typically sigaction() in
1602 modern systems) in an attempt to fix a rare and obscure crash bug.
1604 25. Narrowed the window for a race in the daemon that could cause it to ignore
1605 SIGCHLD signals. This is not a major problem, because they are used only to
1606 wake it up if nothing else does.
1608 26. A malformed maildirsize file could cause Exim to calculate negative values
1609 for the mailbox size or file count. Odd effects could occur as a result.
1610 The maildirsize information is now recalculated if the size or filecount
1613 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1614 support for a long time. Removed HAVE_SYS_VFS_H.
1616 28. Installed the latest version of exipick from John Jetmore.
1618 29. In an address list, if the pattern was not a regular expression, an empty
1619 subject address (from a bounce message) matched only if the pattern was an
1620 empty string. Non-empty patterns were not even tested. This was the wrong
1621 because it is perfectly reasonable to use an empty address as part of a
1622 database query. An empty address is now tested by patterns that are
1623 lookups. However, all the other forms of pattern expect the subject to
1624 contain a local part and a domain, and therefore, for them, an empty
1625 address still always fails if the pattern is not itself empty.
1627 30. Exim went into a mad DNS loop when attempting to do a callout where the
1628 host was specified on an smtp transport, and looking it up yielded more
1629 than one IP address.
1631 31. Re-factored the code for checking spool and log partition space into a
1632 function that finds that data and another that does the check. The former
1633 is then used to implement four new variables: $spool_space, $log_space,
1634 $spool_inodes, and $log_inodes.
1636 32. The RFC2047 encoding function was originally intended for short strings
1637 such as real names; it was not keeping to the 75-character limit for
1638 encoded words that the RFC imposes. It now respects the limit, and
1639 generates multiple encoded words if necessary. To be on the safe side, I
1640 have increased the buffer size for the ${rfc2047: expansion operator from
1643 33. It is now permitted to omit both strings after an "if" condition; if the
1644 condition is true, the result is "true". As before, when the second string
1645 is omitted, a false condition yields an empty string. This makes it less
1646 cumbersome to write custom ACL and router conditions.
1648 34. Failure to deliver a bounce message always caused it to be frozen, even if
1649 there was an errors_to setting on the router. The errors_to setting is now
1652 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1653 canonical form (fully expanded) before being placed in
1654 $sender_host_address.
1656 36. The table in the code that translates DNS record types into text (T_A to
1657 "A" for instance) was missing entries for NS and CNAME. It is just possible
1658 that this could have caused confusion if both these types were looked up
1659 for the same domain, because the text type is used as part of Exim's
1660 per-process caching. But the chance of anyone hitting this buglet seems
1663 37. The dnsdb lookup has been extended in a number of ways.
1665 (1) There is a new type, "zns", which walks up the domain tree until it
1666 finds some nameserver records. It should be used with care.
1668 (2) There is a new type, "mxh", which is like "mx" except that it returns
1669 just the host names, not the priorities.
1671 (3) It is now possible to give a list of domains (or IP addresses) to be
1672 looked up. The behaviour when one of the lookups defers can be
1673 controlled by a keyword.
1675 (4) It is now possible to specify the separator character for use when
1676 multiple records are returned.
1678 38. The dnslists ACL condition has been extended: it is now possible to supply
1679 a list of IP addresses and/or domains to be looked up in a particular DNS
1682 39. Added log_selector=+queue_time_overall.
1684 40. When running the queue in the test harness, wait just a tad after forking a
1685 delivery process, to get repeatability of debugging output.
1687 41. Include certificate and key file names in error message when GnuTLS fails
1688 to set them up, because the GnuTLS error message doesn't include the name
1689 of the failing file when there is a problem reading it.
1691 42. Allow both -bf and -bF in the same test run.
1693 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1695 44. The "Exiscan patch" is now merged into the mainline Exim source.
1697 45. Sometimes the final signoff response after QUIT could fail to get
1698 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1699 < 0 before doing a fflush(). This bug looks as though it goes back to the
1700 introduction of TLS in release 3.20, but "sometimes" must have been rare
1701 because the tests only now provoked it.
1703 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1704 (this can affect the format of dates).
1706 47. exim_tidydb, when checking for the continued existence of a message for
1707 which it has found a message-specific retry record, was not finding
1708 messages that were in split spool directories. Consequently, it was
1709 deleting retry records that should have stayed in existence.
1711 48. Steve fixed some bugs in eximstats.
1713 49. The SPA authentication driver was not abandoning authentication and moving
1714 on to the next authenticator when an expansion was forced to fail,
1715 contradicting the general specification for all authenticators. Instead it
1716 was generating a temporary error. It now behaves as specified.
1718 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1719 (the order specifies the preference for clients). The order is now AES256,
1720 AES128, 3DES, ARCFOUR128.
1722 51. Small patch to Sieve code - explicitly set From: when generating an
1725 52. Exim crashed if a remote delivery caused a very long error message to be
1726 recorded - for instance if somebody sent an entire SpamAssassin report back
1727 as a large number of 550 error lines. This bug was coincidentally fixed by
1728 increasing the size of one of Exim's internal buffers (big_buffer) that
1729 happened as part of the Exiscan merge. However, to be on the safe side, I
1730 have made the code more robust (and fixed the comments that describe what
1733 53. Now that there can be additional text after "Completed" in log lines (if
1734 the queue_time_overall log selector is set), a one-byte patch to exigrep
1735 was needed to allow it to recognize "Completed" as not the last thing in
1738 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1739 patch that reportedly fixes this has been added. I am not expert enough to
1740 create a test for it. This is what the patch creator wrote:
1742 "I found a little strange behaviour of ldap code when working with
1743 Windows 2003 AD Domain, where users was placed in more than one
1744 Organization Units. When I tried to give exim partial DN, the exit code
1745 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1746 But simultaneously result of request was absolutely normal ldap result,
1747 so I produce this patch..."
1749 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1750 REFERENCE, so I have modified the code to exclude the patch when that macro
1753 55. Some experimental protocols are using DNS PTR records for new purposes. The
1754 keys for these records are domain names, not reversed IP addresses. The
1755 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1756 leaves it alone. Component reversal etc. now happens only for IP addresses.
1759 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1761 57. Double the size of the debug message buffer (to 2048) so that more of very
1762 long debug lines gets shown.
1764 58. The exicyclog utility now does better if the number of log files to keep
1765 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1767 59. Two changes related to the smtp_active_hostname option:
1769 (1) $smtp_active_hostname is now available as a variable.
1770 (2) The default for smtp_banner uses $smtp_active_hostname instead
1771 of $primary_hostname.
1773 60. The host_aton() function is supposed to be passed a string that is known
1774 to be a valid IP address. However, in the case of IPv6 addresses, it was
1775 not checking this. This is a hostage to fortune. Exim now panics and dies
1776 if the condition is not met. A case was found where this could be provoked
1777 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1778 components; fortuitously, this particular loophole had already been fixed
1779 by change 4.50/55 above.
1781 If there are any other similar loopholes, the new check in host_aton()
1782 itself should stop them being exploited. The report I received stated that
1783 data on the command line could provoke the exploit when Exim was running as
1784 exim, but did not say which command line option was involved. All I could
1785 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1786 running as the user.
1789 61. There was a buffer overflow vulnerability in the SPA authentication code
1790 (which came originally from the Samba project). I have added a test to the
1791 spa_base64_to_bits() function which I hope fixes it.
1794 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1795 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1797 63. The daemon start-up calls getloadavg() while still root for those OS that
1798 need the first call to be done as root, but it missed one case: when
1799 deliver_queue_load_max is set with deliver_drop_privilege. This is
1800 necessary for the benefit of the queue runner, because there is no re-exec
1801 when deliver_drop_privilege is set.
1803 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1804 This has been fixed.
1806 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1807 was in use, was not putting the data itself into the right store pool;
1808 consequently, it could be overwritten for a subsequent message in the same
1809 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1812 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1814 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1815 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1816 the function were treating the return as a boolean value, which happened to
1817 work because 0=false and not-0=true, but is not correct code.
1819 68. The host_aton() function was not handling scoped IPv6 addresses (those
1820 with, for example, "%eth0" on the end) correctly.
1822 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1823 negated items (that is, ~something) in unsigned ints. Some compilers
1824 apparently mutter when there is no cast.
1826 70. If an address verification called from an ACL failed, and did not produce a
1827 user-specific message (i.e. there was only a "system" message), nothing was
1828 put in $acl_verify_message. In this situation, it now puts the system
1831 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1832 session; change 4.31/43 added the unwanted input to the log line - except
1833 that it did not do this in the start of session case. It now does.
1835 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1836 This is wrong and can cause the other end to generate a synchronization
1837 error if it is another Exim or anything else that does the synchronization
1838 check. A QUIT command is no longer sent after a timeout.
1840 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1841 during host lookups.
1843 74. The defer_ok option of callout verification was not working if it was used
1844 when verifying addresses in header lines, that is, for this case:
1846 verify = header_sender/callout=defer_ok
1848 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1849 those file descriptors could be used for SMTP connections. If anything
1850 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1851 could be sent to the SMTP client, causing chaos. The daemon now opens
1852 stdin, stdout, and stderr to /dev/null when it puts itself into the
1855 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1856 log by default. The user can override this with suitable Perl magic.
1858 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1859 the log message when discard triggers, was not working for the DATA ACL or
1860 for the non-SMTP ACL.
1862 78. Error message wording change in sieve.c.
1864 79. If smtp_accept_max_per_host was set, the number of connections could be
1865 restricted to fewer than expected, because the daemon was trying to set up
1866 a new connection before checking whether the processes handling previous
1867 connections had finished. The check for completed processes is now done
1868 earlier. On busy systems, this bug wouldn't be noticed because something
1869 else would have woken the daemon, and it would have reaped the completed
1872 80. If a message was submitted locally by a user whose login name contained one
1873 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1874 It caused a spool format error. I have fixed the spool reading code. A
1875 related problem was that the "from" clause in the Received: line became
1876 illegal because of the space(s). It is now covered by ${quote_local_part.
1878 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1881 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1883 83. Patch from Sieve maintainer to fix unterminated string problem in
1884 "vacation" handling.
1886 84. Some minor changes to the Linux configuration files to help with other
1887 OS variants using glibc.
1889 85. One more patch for Sieve to update vacation handling to latest spec.
1892 ----------------------------------------------------
1893 See the note above about the 4.44 and 4.50 releases.
1894 ----------------------------------------------------
1900 1. Change 4.43/35 introduced a bug that caused file counts to be
1901 incorrectly computed when quota_filecount was set in an appendfile
1904 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1905 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1908 3. Give more explanation in the error message when the command for a transport
1909 filter fails to execute.
1911 4. There are several places where Exim runs a non-Exim command in a
1912 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1913 was being done only for the command run by the queryprogram router. It is
1914 now done for all such subprocesses. The other cases are: ${run, transport
1915 filters, and the commands run by the lmtp and pipe transports.
1917 5. Some older OS have a limit of 256 on the maximum number of file
1918 descriptors. Exim was using setrlimit() to set 1000 as a large value
1919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1920 systems. I've change it so that if it can't get 1000, it tries for 256.
1922 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1923 was an oversight, and furthermore, ever since the addition of extra
1924 controls (e.g. 4.43/32), the checks on when to allow different forms of
1925 "control" were broken. There should now be diagnostics for all cases when a
1926 control that does not make sense is encountered.
1928 7. $recipients is now available in the predata ACL (oversight).
1930 8. Tidy the search cache before the fork to do a delivery from a message
1931 received from the command line. Otherwise the child will trigger a lookup
1932 failure and thereby defer the delivery if it tries to use (for example) a
1933 cached ldap connection that the parent has called unbind on.
1935 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1936 of $address_data from the recipient verification was clobbered by the
1937 sender verification.
1939 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1940 was its contents. (It was OK if the option was not defined at all.)
1942 11. A "Completed" log line is now written for messages that are removed from
1943 the spool by the -Mrm option.
1945 12. $host_address is now set to the target address during the checking of
1946 ignore_target_hosts.
1948 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1949 being passed; this would have caused $sender_host_name to have been used if
1950 matching the list had actually called for a host name (not very likely,
1951 since this list is usually IP addresses). A host name is now passed as
1954 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1955 code that specifies a non-restarting handler (typically sigaction() in
1956 modern systems) in an attempt to fix a rare and obscure crash bug.
1958 15. Narrowed the window for a race in the daemon that could cause it to ignore
1959 SIGCHLD signals. This is not a major problem, because they are used only to
1960 wake it up if nothing else does.
1962 16. A malformed maildirsize file could cause Exim to calculate negative values
1963 for the mailbox size or file count. Odd effects could occur as a result.
1964 The maildirsize information is now recalculated if the size or filecount
1967 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1968 support for a long time. Removed HAVE_SYS_VFS_H.
1970 18. Updated exipick to current release from John Jetmore.
1972 19. Allow an empty sender to be matched against a lookup in an address list.
1973 Previously the only cases considered were a regular expression, or an
1976 20. Exim went into a mad DNS lookup loop when doing a callout where the
1977 host was specified on the transport, if the DNS lookup yielded more than
1980 21. The RFC2047 encoding function was originally intended for short strings
1981 such as real names; it was not keeping to the 75-character limit for
1982 encoded words that the RFC imposes. It now respects the limit, and
1983 generates multiple encoded words if necessary. To be on the safe side, I
1984 have increased the buffer size for the ${rfc2047: expansion operator from
1987 22. Failure to deliver a bounce message always caused it to be frozen, even if
1988 there was an errors_to setting on the router. The errors_to setting is now
1991 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1992 canonical form (fully expanded) before being placed in
1993 $sender_host_address.
1995 24. Updated eximstats to version 1.33
1997 25. Include certificate and key file names in error message when GnuTLS fails
1998 to set them up, because the GnuTLS error message doesn't include the name
1999 of the failing file when there is a problem reading it.
2001 26. Expand error message when OpenSSL has problems setting up cert/key files.
2004 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2005 (this can affect the format of dates).
2007 28. exim_tidydb, when checking for the continued existence of a message for
2008 which it has found a message-specific retry record, was not finding
2009 messages that were in split spool directories. Consequently, it was
2010 deleting retry records that should have stayed in existence.
2012 29. eximstats updated to version 1.35
2013 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2014 1.35 - bugfix such that pie charts by volume are generated correctly
2016 30. The SPA authentication driver was not abandoning authentication and moving
2017 on to the next authenticator when an expansion was forced to fail,
2018 contradicting the general specification for all authenticators. Instead it
2019 was generating a temporary error. It now behaves as specified.
2021 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2022 (the order specifies the preference for clients). The order is now AES256,
2023 AES128, 3DES, ARCFOUR128.
2025 31. Small patch to Sieve code - explicitly set From: when generating an
2028 32. Exim crashed if a remote delivery caused a very long error message to be
2029 recorded - for instance if somebody sent an entire SpamAssassin report back
2030 as a large number of 550 error lines. This bug was coincidentally fixed by
2031 increasing the size of one of Exim's internal buffers (big_buffer) that
2032 happened as part of the Exiscan merge. However, to be on the safe side, I
2033 have made the code more robust (and fixed the comments that describe what
2036 33. Some experimental protocols are using DNS PTR records for new purposes. The
2037 keys for these records are domain names, not reversed IP addresses. The
2038 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2039 leaves it alone. Component reversal etc. now happens only for IP addresses.
2042 34. The host_aton() function is supposed to be passed a string that is known
2043 to be a valid IP address. However, in the case of IPv6 addresses, it was
2044 not checking this. This is a hostage to fortune. Exim now panics and dies
2045 if the condition is not met. A case was found where this could be provoked
2046 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2047 components; fortuitously, this particular loophole had already been fixed
2048 by change 4.50/55 or 4.44/33 above.
2050 If there are any other similar loopholes, the new check in host_aton()
2051 itself should stop them being exploited. The report I received stated that
2052 data on the command line could provoke the exploit when Exim was running as
2053 exim, but did not say which command line option was involved. All I could
2054 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2055 running as the user.
2058 35. There was a buffer overflow vulnerability in the SPA authentication code
2059 (which came originally from the Samba project). I have added a test to the
2060 spa_base64_to_bits() function which I hope fixes it.
2063 36. The daemon start-up calls getloadavg() while still root for those OS that
2064 need the first call to be done as root, but it missed one case: when
2065 deliver_queue_load_max is set with deliver_drop_privilege. This is
2066 necessary for the benefit of the queue runner, because there is no re-exec
2067 when deliver_drop_privilege is set.
2069 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2070 was in use, was not putting the data itself into the right store pool;
2071 consequently, it could be overwritten for a subsequent message in the same
2072 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2075 38. Sometimes the final signoff response after QUIT could fail to get
2076 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2077 < 0 before doing a fflush(). This bug looks as though it goes back to the
2078 introduction of TLS in release 3.20, but "sometimes" must have been rare
2079 because the tests only now provoked it.
2085 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2086 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2087 Now it returns an integer. A number of calls were still expecting a T/F
2088 return. Fortuitously, in all cases, the tests worked in OK situations,
2089 which is the norm. However, things would have gone wrong on any write
2090 failures on the smtp file descriptor. This function is used when sending
2091 messages over SMTP and also when doing verify callouts.
2093 2. When Exim is called to do synchronous delivery of a locally submitted
2094 message (the -odf or -odi options), it no longer closes stderr before doing
2097 3. Implemented the mua_wrapper option.
2099 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2101 5. Implemented the functions header_remove(), header_testname(),
2102 header_add_at_position(), and receive_remove_recipient(), and exported them
2105 6. If an ACL "warn" statement specified the addition of headers, Exim already
2106 inserted X-ACL-Warn: at the start if there was no header name. However, it
2107 was not making this test for the second and subsequent header lines if
2108 there were newlines in the string. This meant that an invalid header could
2109 be inserted if Exim was badly configured.
2111 7. Allow an ACL "warn" statement to add header lines at the start or after all
2112 the Received: headers, as well as at the end.
2114 8. Added the rcpt_4xx retry error code.
2116 9. Added postmaster_mailfrom=xxx to callout verification option.
2118 10. Added mailfrom=xxxx to the callout verification option, for verify=
2121 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2122 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2124 12. Inserted some casts to stop certain compilers warning when using pointer
2125 differences as field lengths or precisions in printf-type calls (mostly
2126 affecting debugging statements).
2128 13. Added optional readline() support for -be (dynamically loaded).
2130 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2131 same clock tick as a message's arrival, so that its received time was the
2132 same as the "first fail" time on the retry record, and that message
2133 remained on the queue past the ultimate address timeout, every queue runner
2134 would try a delivery (because it was past the ultimate address timeout) but
2135 after another failure, the ultimate address timeout, which should have then
2136 bounced the address, did not kick in. This was a "< instead of <=" error;
2137 in most cases the first failure would have been in the next clock tick
2138 after the received time, and all would be well.
2140 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2141 being recognized when the domain list was tested by the match_domain
2142 condition in an expansion string.
2144 16. Added the ${str2b64: operator.
2146 17. Exim was always calling setrlimit() to set a large limit for the number of
2147 processes, without checking whether the existing limit was already
2148 adequate. (It did check for the limit on file descriptors.) Furthermore,
2149 errors from getrlimit() and setrlimit() were being ignored. Now they are
2150 logged to the main and panic logs, but Exim does carry on, to try to do its
2151 job under whatever limits there are.
2153 18. Imported PCRE 5.0.
2155 19. Trivial typo in log message " temporarily refused connection" (the leading
2158 20. If the log selector return_path_on_delivery was set and an address was
2159 redirected to /dev/null, the delivery process crashed because it assumed
2160 that a return path would always be set for a "successful" delivery. In this
2161 case, the whole delivery is bypassed as an optimization, and therefore no
2164 21. Internal re-arrangement: the function for sending a challenge and reading
2165 a response while authentication was assuming a zero-terminated challenge
2166 string. It's now changed to take a pointer and a length, to allow for
2167 binary data in such strings.
2169 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2171 23. Exim was not respecting finduser_retries when seeking the login of the
2172 uid under which it was called; it was always trying 10 times. (The default
2173 setting of finduser_retries is zero.) Also, it was sleeping after the final
2174 failure, which is pointless.
2176 24. Implemented tls_on_connect_ports.
2178 25. Implemented acl_smtp_predata.
2180 26. If the domain in control=submission is set empty, Exim assumes that the
2181 authenticated id is a complete email address when it generates From: or
2182 Sender: header lines.
2184 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2185 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2186 chown and chgrp in /bin and hostname in /usr/bin.
2188 28. Exim was keeping the "process log" file open after each use, just as it
2189 does for the main log. This opens the possibility of it remaining open for
2190 long periods when the USR1 signal hits a daemon. Occasional processlog
2191 errors were reported, that could have been caused by this. Anyway, it seems
2192 much more sensible not to leave this file open at all, so that is what now
2195 29. The long-running daemon process does not normally write to the log once it
2196 has entered its main loop, and it closes the log before doing so. This is
2197 so that log files can straightforwardly be renamed and moved. However,
2198 there are a couple of unusual error situations where the daemon does write
2199 log entries, and I had neglected to close the log afterwards.
2201 30. The text of an SMTP error response that was received during a remote
2202 delivery was being truncated at 512 bytes. This is too short for some of
2203 the long messages that one sometimes sees. I've increased the limit to
2206 31. It is now possible to make retry rules that apply only when a message has a
2207 specific sender, in particular, an empty sender.
2209 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2210 it possible to be selective about when SMTP synchronization is enforced.
2212 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2214 32. Implemented hosts_connection_nolog.
2216 33. Added an ACL for QUIT.
2218 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2221 35. Added mailbox_size and mailbox_filecount to appendfile.
2223 36. Added control = no_multiline_responses to ACLs.
2225 37. There was a bug in the logic of the code that waits for the clock to tick
2226 in the case where the clock went backwards by a substantial amount such
2227 that the microsecond fraction of "now" was more than the microsecond
2228 fraction of "then" (but the whole seconds number was less).
2230 38. Added support for the libradius Radius client library this is found on
2231 FreeBSD (previously only the radiusclient library was supported).
2237 1. When certain lookups returned multiple values in the form name=value, the
2238 quoting of the values was not always being done properly. Specifically:
2239 (a) If the value started with a double quote, but contained no whitespace,
2241 (b) If the value contained whitespace other than a space character (i.e.
2242 tabs or newlines or carriage returns) it was not quoted.
2243 This fix has been applied to the mysql and pgsql lookups by writing a
2244 separate quoting function and calling it from the lookup code. The fix
2245 should probably also be applied to nisplus, ibase and oracle lookups, but
2246 since I cannot test any of those, I have not disturbed their existing code.
2248 2. A hit in the callout cache for a specific address caused a log line with no
2249 reason for rejecting RCPT. Now it says "Previous (cached) callout
2250 verification failure".
2252 3. There was an off-by-one bug in the queryprogram router. An over-long
2253 return line was truncated at 256 instead of 255 characters, thereby
2254 overflowing its buffer with the terminating zero. As well as fixing this, I
2255 have increased the buffer size to 1024 (and made a note to document this).
2257 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2258 when Exim is waiting for an SMTP response from a remote server, Exim
2259 restarts its select() call on the socket, thereby resetting its timeout.
2260 This is not a problem when such interrupts are rare. Somebody set up a cron
2261 job to run exiwhat every 2 minutes, which is less than the normal select()
2262 timeout (5 or 10 minutes). This meant that the select() timeout never
2263 kicked in because it was always reset. I have fixed this by comparing the
2264 time when an interrupt arrives with the time at the start of the first call
2265 to select(). If more time than the timeout has elapsed, the interrupt is
2266 treated as a timeout.
2268 5. Some internal re-factoring in preparation for the addition of Sieve
2269 extensions (by MH). In particular, the "personal" test is moved to a
2270 separate function, and given an option for scanning Cc: and Bcc: (which is
2271 not set for Exim filters).
2273 6. When Exim created an email address using the login of the caller as the
2274 local part (e.g. when creating a From: or Sender: header line), it was not
2275 quoting the local part when it contained special characters such as @.
2277 7. Installed new OpenBSD configuration files.
2279 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2280 try to make them clearer.
2282 9. Callout options, other than the timeout value, were being ignored when
2283 verifying sender addresses in header lines. For example, when using
2285 verify = header_sender/callout=no_cache
2287 the cache was (incorrectly) being used.
2289 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2290 only the Cygwin environment.
2292 11. When return_path_on_delivery was set as a log selector, if different remote
2293 addresses in the same message used different return paths and parallel
2294 remote delivery occurred, the wrong values would sometimes be logged.
2295 (Whenever a remote delivery process finished, the return path value from
2296 the most recently started remote delivery process was logged.)
2298 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2299 lines when AUTH and/or TLS are in use. This is the "received protocol"
2300 field. Exim used to use "asmtp" for authenticated SMTP, without any
2301 indication (in the protocol name) for TLS use. Now it follows the RFC and
2302 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2303 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2304 names appear in log lines as well as in Received: header lines.
2306 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2307 extensions, and comparison tests, and to fix some bugs.
2309 14. Changes to the "personal" filter test:
2311 (1) The test was buggy in that it was just doing the equivalent of
2312 "contains" tests on header lines. For example, if a user's address was
2313 anne@some.where, the "personal" test would incorrectly be true for
2315 To: susanne@some.where
2317 This test is now done by extracting each address from the header in turn,
2318 and checking the entire address. Other tests that are part of "personal"
2319 are now done using regular expressions (for example, to check local parts
2320 of addresses in From: header lines).
2322 (2) The list of non-personal local parts in From: addresses has been
2323 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2324 taken from the Sieve specification recommendations.
2326 (3) If the message contains any header line starting with "List-" it is
2327 treated as non-personal.
2329 (4) The test for "circular" in the Subject: header line has been removed
2330 because it now seems ill-conceived.
2332 15. Minor typos in src/EDITME comments corrected.
2334 16. Installed latest exipick from John Jetmore.
2336 17. If headers_add on a router specified a text string that was too long for
2337 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2338 of string_sprintf() is now avoided.
2340 18. $message_body_size was not set (it was always zero) when running the DATA
2341 ACL and the local_scan() function.
2343 19. For the "mail" command in an Exim filter, no default was being set for
2344 the once_repeat time, causing a random time value to be used if "once" was
2345 specified. (If the value happened to be <= 0, no repeat happened.) The
2346 default is now 0s, meaning "never repeat". The "vacation" command was OK
2347 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2348 (I found it when inspecting the code).
2350 20. There is now an overall timeout for performing a callout verification. It
2351 defaults to 4 times the callout timeout, which applies to individual SMTP
2352 commands during the callout. The overall timeout applies when there is more
2353 than one host that can be tried. The timeout is checked before trying the
2354 next host. This prevents very long delays if there are a large number of
2355 hosts and all are timing out (e.g. when the network connections are timing
2356 out). The value of the overall timeout can be changed by specifying an
2357 additional sub-option for "callout", called "maxwait". For example:
2359 verify = sender/callout=5s,maxwait=20s
2361 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2362 to the end before writing, but this should make it even safer).
2364 22. Exim was forgetting that it had advertised PIPELINING for the second and
2365 subsequent messages on an SMTP connection. It was also not resetting its
2366 memory on STARTTLS and an internal HELO.
2368 23. When Exim logs an SMTP synchronization error within a session, it now
2369 records whether PIPELINING has been advertised or not.
2371 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2372 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2373 rather than long int.
2375 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2377 26. Added the never_mail option to autoreply.
2383 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2384 crash if the getsockname() call failed; this can happen if a connection is
2385 closed very soon after it is established. The problem was simply in the
2386 order in which certain operations were done, causing Exim to try to write
2387 to the SMTP stream before it had set up the file descriptor. The bug has
2388 been fixed by making things happen in the correct order.
2394 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2395 before the connection was closed, thus losing the rejection response.
2397 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2398 some early Solaris releases, but causes trouble in current releases where
2399 socklen_t is defined.
2401 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2404 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2407 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2408 that releasing the top of it at the end releases what was used for sub-
2409 expansions (unless the block got too big). However, discard this block if
2410 the first thing is a variable or header, so that we can use its block when
2411 it is dynamic (useful for very large $message_headers, for example).
2413 6. Lookups now cache *every* query, not just the most recent. A new, separate
2414 store pool is used for this. It can be recovered when all lookup caches are
2415 flushed. Lookups now release memory at the end of their result strings.
2416 This has involved some general refactoring of the lookup sources.
2418 7. Some code has been added to the store_xxx() functions to reduce the amount
2419 of flapping under certain conditions.
2421 8. log_incoming_interface used to affect only the <= reception log lines. Now
2422 it causes the local interface and port to be added to several more SMTP log
2423 lines, for example "SMTP connection from", and rejection lines.
2425 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2427 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2429 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2430 could be overwritten at the end of the current message (or the start of a
2431 new message if it was set in a HELO ACL). The value is now preserved for
2432 the duration of the SMTP connection.
2434 12. If a transport had a headers_rewrite setting, and a matching header line
2435 contained an unqualified address, that address was qualified, even if it
2436 did not match any rewriting rules. The underlying bug was that the values
2437 of the flags that permit the existence of unqualified sender and recipient
2438 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2439 non-local messages, and by -bnq for local messages) were not being
2440 preserved with the message after it was received.
2442 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2443 "next input=" as part of the text comprising the host identity instead of
2444 the correct text. The code was using the same buffer for two different
2445 strings. However, depending on which order the printing function evaluated
2446 its arguments, the bug did not always show up. Under Linux, for example, my
2447 test suite worked just fine.
2449 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2450 doesn't work with some older versions of Perl. It has been changed to "my",
2451 which in any case is probably the better facility to use.
2453 15. A really picky compiler found some instances of statements for creating
2454 error messages that either had too many or two few arguments for the format
2457 16. The size of the buffer for calls to the DNS resolver has been increased
2458 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2459 for addresses that have a lot of PTR records. This alleviates a problem; it
2460 does not fully solve it.
2462 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2463 buffer now truncates the list and logs the incident, which is the same
2464 action as happens when Exim is looking up a host name and its aliases.
2465 Previously in this situation something unpredictable would happen;
2466 sometimes it was "internal error: store_reset failed".
2468 18. If a server dropped the connection unexpectedly when an Exim client was
2469 using GnuTLS and trying to read a response, the client delivery process
2470 crashed while trying to generate an error log message.
2472 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2473 string, for example:
2475 warn message = H1: something\nH2: something
2477 the text was added as a single header line from Exim's point of view
2478 though it ended up OK in the delivered message. However, searching for the
2479 second and subsequent header lines using $h_h2: did not work. This has been
2480 fixed. Similarly, if a system filter added multiple headers in this way,
2481 the routers could not see them.
2483 20. Expanded the error message when iplsearch is called with an invalid key to
2484 suggest using net-iplsearch in a host list.
2486 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2487 ACLs are no longer actually imposed (and a message to that effect is
2490 22. If a "gecos" field in a passwd entry contained escaped characters, in
2491 particular, if it contained a \" sequence, Exim got it wrong when building
2492 a From: or a Sender: header from that name. A second bug also caused
2493 incorrect handling when an unquoted " was present following a character
2494 that needed quoting.
2496 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2497 was not being matched caselessly.
2499 24. Arranged for all hyphens in the exim.8 source to be escaped with
2502 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2503 properly. Recipient callout cache records were still being keyed to include
2504 the sender, even when use_sender was set false. This led to far more
2505 callouts that were necessary. The sender is no longer included in the key
2506 when use_sender is false.
2508 26. Added "control = submission" modifier to ACLs.
2510 27. Added the ${base62d: operator to decode base 62 numbers.
2512 28. dnsdb lookups can now access SRV records.
2514 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2515 the configuration file.
2517 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2518 (-v) mode. This makes the output for a verbose queue run more intelligible.
2520 31. Added a use_postmaster feature to recipient callouts.
2522 32. Added the $body_zerocount variable, containing the number of binary zero
2523 bytes in the message body.
2525 33. The time of last modification of the "new" subdirectory is now used as the
2526 "mailbox time last read" when there is a quota error for a maildir
2529 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2531 35. Added +ignore_unknown as a special item in host lists.
2533 36. Code for decoding IPv6 addresses in host lists is now included, even if
2534 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2535 address was recognized as an IP address, but was then not correctly decoded
2536 into binary, causing unexpected and incorrect effects when compared with
2543 1. Very minor rewording of debugging text in manualroute to say "list of
2544 hosts" instead of "hostlist".
2546 2. If verify=header_syntax was set, and a header line with an unqualified
2547 address (no domain) and a large number of spaces between the end of the
2548 name and the colon was received, the reception process suffered a buffer
2549 overflow, and (when I tested it) crashed. This was caused by some obsolete
2550 code that should have been removed. The fix is to remove it!
2552 3. When running in the test harness, delay a bit after writing a bounce
2553 message to get a bit more predictability in the log output.
2555 4. Added a call to search_tidyup() just before forking a reception process. In
2556 theory, someone could use a lookup in the expansion of smtp_accept_max_
2557 per_host which, without the tidyup, could leave open a database connection.
2559 5. Added the variables $recipient_data and $sender_data which get set from a
2560 lookup success in an ACL "recipients" or "senders" condition, or a router
2561 "senders" option, similar to $domain_data and $local_part_data.
2563 6. Moved the writing of debug_print from before to after the "senders" test
2566 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2567 problems for message scanning, either using a data ACL, or using
2568 local_scan() because the Received: header was not generated till after they
2569 were called (in order to set the time as the time of reception completion).
2570 I have revised the way this works. The header is now generated after the
2571 body is received, but before the ACL or local_scan() are called. After they
2572 are run, the timestamp in the header is updated.
2578 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2579 before starting a queue runner without re-exec. This happened only when
2580 deliver_drop_privilege was set or when the Exim user was set to root. The
2581 effect of the bug was that timeouts during subsequent deliveries caused
2582 crashes instead of being properly handled. The handler is now left at its
2583 default (and expected) setting.
2585 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2586 message, again when deliver_drop_privilege is set or Exim is run as root.
2587 The bug described in (1) was not present in this case, but the tidying up
2588 of the other signals was missing. I have made the two cases consistent.
2590 3. The ignore_target_hosts setting on a manualroute router was being ignored
2591 for hosts that were looked up using the /MX notation.
2593 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2596 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2597 operated on the sender address. After changing the $sender_address to <>
2598 for the sender address verify, Exim was re-instated it as the original
2599 (before rewriting) address, but remembering that it had rewritten it, so it
2600 wasn't rewriting it again. This bug also had the effect of breaking the
2601 sender address verification caching when the sender address was rewritten.
2603 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2604 This has been changed so that if the ip literal address matches
2605 ignore_target_hosts, the router declines.
2607 7. Added expansion conditions match_domain, match_address, and match_local_
2608 part (NOT match_host).
2610 8. The placeholder for the Received: header didn't have a length field set.
2612 9. Added code to Exim itself and to exim_lock to test for a specific race
2613 condition that could lead to file corruption when using MBX delivery. The
2614 issue is with the lockfile that is created in /tmp. If this file is removed
2615 after a process has opened it but before that process has acquired a lock,
2616 there is the potential for a second process to recreate the file and also
2617 acquire a lock. This could lead to two Exim processes writing to the file
2618 at the same time. The added code performs the same test as UW imapd; it
2619 checks after acquiring the lock that its file descriptor still refers to
2620 the same named file.
2622 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2623 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2626 11. Added the smtp_active_hostname option. If used, this will typically be made
2627 to depend on the incoming interface address. Because $interface_address is
2628 not set up until the daemon has forked a reception process, error responses
2629 that can happen earlier (such as "too many connections") no longer contain
2632 12. If an expansion in a condition on a "warn" statement fails because a lookup
2633 defers, the "warn" statement is abandoned, and the next ACL statement is
2634 processed. Previously this caused the whole ACL to be aborted.
2636 13. Added the iplsearch lookup type.
2638 14. Added ident_timeout as a log selector.
2640 15. Added tls_certificate_verified as a log selector.
2642 16. Added a global option tls_require_ciphers (compare the smtp transport
2643 option of the same name). This controls incoming TLS connections.
2645 17. I finally figured out how to make tls_require_ciphers do a similar thing
2646 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2647 before starting the TLS session.
2649 18. Tabs are now shown as \t in -bP output.
2651 19. If the log selector return_path_on_delivery was set, Exim crashed when
2652 bouncing a message because it had too many Received: header lines.
2654 20. If two routers both had headers_remove settings, and the first one included
2655 a superfluous trailing colon, the final name in the first list and the
2656 first name in the second list were incorrectly joined into one item (with a
2657 colon in the middle).
2663 1. Added -C and -D options to the exinext utility, mainly to make it easier
2664 to include in the automated testing, but these could be helpful when
2665 multiple configurations are in use.
2667 2. The exinext utility was not formatting the output nicely when there was
2668 an alternate port involved in the retry record key, nor when there was a
2669 message id as well (for retries that were specific to a specific message
2670 and a specific host). It was also confused by IPv6 addresses, because of
2671 the additional colons they contain. I have fixed the IPv4 problem, and
2672 patched it up to do a reasonable job for IPv6.
2674 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2675 delivery, the log line now contains "pipelined" if PIPELINING was used.
2677 4. An SMTP transport process used to panic and die if the bind() call to set
2678 an explicit outgoing interface failed. This has been changed; it is now
2679 treated in the same way as a connect() failure.
2681 5. A reference to $sender_host_name in the part of a conditional expansion
2682 that was being skipped was still causing a DNS lookup. This no longer
2685 6. The def: expansion condition was not recognizing references to header lines
2686 that used bh_ and bheader_.
2688 7. Added the _cache feature to named lists.
2690 8. The code for checking quota_filecount in the appendfile transport was
2691 allowing one more file than it should have been.
2693 9. For compatibility with Sendmail, the command line option
2701 and sets the incoming protocol and host name (for trusted callers). The
2702 host name and its colon can be omitted when only the protocol is to be set.
2703 Note the Exim already has two private options, -pd and -ps, that refer to
2704 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2705 "s", but I don't think that's a major issue.
2707 10. A number of refactoring changes to the code, none of which should affect
2710 (a) The number of logging options was getting close to filling up the
2711 32-bit word that was used as a bit map. I have split them into two classes:
2712 those that are passed in the argument to log_write(), and those that are
2713 only ever tested independently outside of that function. These are now in
2714 separate 32-bit words, so there is plenty of room for expansion again.
2715 There is no change in the user interface or the logging behaviour.
2717 (b) When building, for example, log lines, the code previously used a
2718 macro that called string_cat() twice, in order to add two strings. This is
2719 not really sufficiently general. Furthermore, there was one instance where
2720 it was actually wrong because one of the argument was used twice, and in
2721 one call a function was used. (As it happened, calling the function twice
2722 did not affect the overall behaviour.) The macro has been replaced by a
2723 function that can join an arbitrary number of extra strings onto a growing
2726 (c) The code for expansion conditions now uses a table and a binary chop
2727 instead of a serial search (which was left over from when there were very
2728 few conditions). Also, it now recognizes conditions like "pam" even when
2729 the relevant support is not compiled in: a suitably worded error message is
2730 given if an attempt is made to use such a condition.
2732 11. Added ${time_interval:xxxxx}.
2734 12. A bug was causing one of the ddress fields not to be passed back correctly
2735 from remote delivery subprocesses. The field in question was not being
2736 subsequently used, so this caused to problems in practice.
2738 13. Added new log selectors queue_time and deliver_time.
2740 14. Might have fixed a bug in maildirsizefile handling that threw up
2741 "unexpected character" debug warnings, and recalculated the data
2742 unnecessarily. In any case, I expanded the warning message to give more
2745 15. Added the message "Restricted characters in address" to the statements in
2746 the default ACL that block characters like @ and % in local parts.
2748 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2749 Three changes have been made:
2751 (a) There was a serious bug; a negative response to MAIL caused the whole
2752 recipient domain to be cached as invalid, thereby blocking all messages
2753 to all local parts at the same domain, from all senders. This bug has
2754 been fixed. The domain is no longer cached after a negative response to
2755 MAIL if the sender used is not empty.
2757 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2760 (c) A new callout option, "use_sender" has been added for people who want
2761 the modified behaviour.
2767 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2770 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2771 indeed breaks things for older releases.
2773 3. Added additional logging to the case where there is a problem reading data
2774 from a filter that is running in a subprocess using a pipe, in order to
2775 try to track down a specific problem.
2777 4. Testing facility fudge: when running in the test harness and attempting
2778 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2779 getting "No route to host". Convert this to a timeout.
2781 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2784 6. Some OS don't have socklen_t but use size_t instead. This affects the
2785 fifth argument of getsockopt() amongst other things. This is now
2786 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2787 can be set for individual OS. I have set it for SunOS5, OSF1, and
2788 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2789 some earlier ones do not.
2791 7. Change 4.30/15 was not doing the test caselessly.
2793 8. The standard form for an IPv6 address literal was being rejected by address
2794 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2795 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2796 this, as well as the form without the "IPv6" on the front (but only when
2797 address literals are enabled, of course).
2799 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2801 10. Exim crashed if a message with an empty sender address specified by -f
2802 encountered a router with an errors_to setting. This could be provoked only
2803 by a command such as
2807 where an empty string was supplied; "<>" did not hit this bug.
2809 11. Installed PCRE release 4.5.
2811 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2812 remained set. It is now erased.
2814 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2815 times from message ids (which are base 36 rather than the normal 62).
2817 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2818 were being counted as actual protocol errors, and logged if the log
2819 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2820 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2821 and DATA following a set of rejected RCPTs do not count as protocol errors.
2822 In other words, Exim assumes they were pipelined, though this may not
2823 actually be the case. Of course, in all cases the client gets an
2824 appropriate error code.
2826 15. If a lookup fails in an ACL condition, a message about the failure may
2827 be available; it is used if testing the ACL cannot continue, because most
2828 such messages specify what the cause of the deferral is. However, some
2829 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2830 that caused an old message to be retained and used if a later statement
2831 caused a defer, replacing the real cause of the deferral.
2833 16. If an IP address had so many PTR records that the DNS lookup buffer
2834 was not large enough to hold them, Exim could crash while trying to process
2835 the truncated data. It now detects and logs this case.
2837 17. Further to 4.21/58, another change has been made: if (and only if) the
2838 first line of a message (the first header line) ends with CRLF, a bare LF
2839 in a subsequent header line has a space inserted after it, so as not to
2840 terminate the header.
2842 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2843 unnecessarily, used atoi() instead of strtol(), and didn't check the
2844 termination when getting file sizes from file names by regex.
2846 19. Completely re-implemented the support for maildirsize files, in the light
2847 of a number of problems with the previous contributed implementation
2848 (4.30/29). In particular:
2850 . If the quota is zero, the maildirsize file is maintained, but no quota is
2853 . If the maildir directory does not exist, it is created before any attempt
2854 to write a maildirsize file.
2856 . The quota value in the file is just a cache; if the quota is changed in
2857 the transport, the new value overrides.
2859 . A regular expression is available for excluding directories from the
2862 20. The autoreply transport checks the characters in options that define the
2863 message's headers; it allows continued headers, but it was checking with
2864 isspace() after an embedded newline instead of explicitly looking for a
2867 21. If all the "regular" hosts to which an address was routed had passed their
2868 expiry times, and had not reached their retry times, the address was
2869 bounced, even if fallback hosts were defined. Now Exim should go on to try
2872 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2873 equivalent code in the SMTP transport. Some hosts send humungous responses
2874 to HELO/EHLO, more than 1024 it seems.
2876 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2877 gives compiler warnings in some environments. I've now done it "properly",
2880 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2881 (because of problems with the built-in one) was declared to return uschar *
2882 instead of char *, causing compiler failure.
2884 25. Fixed a file descriptor leak when processing alias/forward files.
2886 26. Fixed a minor format string issue in dbfn.c.
2888 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2890 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2891 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2894 29. When a sender address is verified, it is cached, to save repeating the test
2895 when there is more than one recipient in a message. However, when the
2896 verification involves a callout, it is possible for different callout
2897 options to be set for different recipients. It is too complicated to keep
2898 track of this in the cache, so now Exim always runs a verification when a
2899 callout is required, relying on the callout cache for the optimization.
2900 The overhead is duplication of the address routing, but this should not be
2903 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2904 to be verified with callout=postmaster, and the main callout worked but the
2905 postmaster check failed, the verification correctly failed. However, if a
2906 subsequent RCPT command asked for sender verification *without* the
2907 postmaster check, incorrect caching caused this verification also to fail,
2910 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2911 it was not caching the DNS options (qualify_single, search_parents) that
2912 were used when the lookup failed. A subsequent lookup with different
2913 options therefore always gave the same answer, though there were cases
2914 where it should not have. (Example: a "domains = !$mx_any" option on a
2915 dnslookup router: the "domains" option is always processed without any
2916 widening, but the router might have qualify_single set.) Now Exim uses the
2917 cached value only when the same options are set.
2919 32. Added John Jetmore's "exipick" utility to the distribution.
2921 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2922 than a timeout (e.g. a certificate is required, and is not provided), an
2923 Exim server now closes the connection immediately. Previously it waited for
2924 the client to close - but if the client is SSL, it seems that they each
2925 wait for each other, leading to a delay before one of them times out.
2927 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2928 maintained 0.8.x compatibility because I don't think many are using it, and
2929 it is clearly obsolete.
2931 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2934 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2935 client certificate was expired. A simple patch fixes this, though I don't
2936 understand the full logic of why the verify callback is called multiple
2939 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2940 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2941 which causes problems with some clients (such as the Certicom SSL Plus
2942 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2943 disables the coutermeasure allowing Eudora to connect."
2945 38. Exim was not checking that a write() to a log file succeeded. This could
2946 lead to Bad Things if a log got too big, in particular if it hit a file
2947 size limit. Exim now panics and dies if it cannot write to a log file, just
2948 as it does if it cannot open a log file.
2950 39. Modified OS/Makefile-Linux so that it now contains
2952 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
2954 The two -D definitions ensure that Exim is compiled with large file
2955 support, which makes it possible to handle log files that are bigger than
2958 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
2959 instance) a domain was checked against a named list that involved a lookup,
2960 causing $domain_data to be set, then another domain was checked against the
2961 same list, then the first domain was re-checked, the value of $domain_data
2962 after the final check could be wrong. In particular, if the second check
2963 failed, it could be set empty. This bug probably also applied to
2966 41. The strip_trailing_dot option was not being applied to the address given
2967 with the -f command-line option.
2969 42. The code for reading a message's header from the spool was incrementing
2970 $received_count, but never initializing it. This meant that the value was
2971 incorrect (doubled) while delivering a message in the same process in which
2972 it was received. In the most common configuration of Exim, this never
2973 happens - a fresh exec is done - but it can happen when
2974 deliver_drop_privilege is set.
2976 43. When Exim logs an SMTP synchronization error - client data sent too soon -
2977 it now includes up to 150 characters of the unexpected data in the log
2980 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
2981 and building data strings. The size of both of these buffers was 10 000
2982 bytes - far larger than anybody would *ever* want, thought I. Needless to
2983 say, somebody hit the limit. I have increased the maximum line length to
2984 20 000 and the maximum data length of concatenated lines to 100 000. I have
2985 also fixed two bugs, because there was no checking on these buffers. Tsk,
2986 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
2987 buffer is too small.
2989 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
2990 lsearch lookups. Now it does.
2992 46. When parsing a route_list item in a manualroute router, a fixed-length
2993 buffer was used for the list of hosts. I made this 1024 bytes long,
2994 thinking that nobody would ever have a list of hosts that long. Wrong.
2995 Somebody had a whole pile of complicated expansion conditions, and the
2996 string was silently truncated, leading to an expansion error. It turns out
2997 that it is easier to change to an unlimited length (owing to other changes
2998 that have happened since this code was originally written) than to build
2999 structure for giving a limitation error. The length of the item that
3000 expands into the list of hosts is now unlimited.
3002 47. The lsearch lookup could not handle data where the length of text line was
3003 more than 4095 characters. Such lines were truncated, leading to shortened
3004 data being returned. It should now handle lines of any length.
3006 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3007 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3010 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3011 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3012 no longer happens in comment lines. A list of replacements is now placed
3013 at the head of all of the source files, except those whose only change is
3014 to replace PERL_COMMAND in the very first #! line.
3016 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3017 messages on the queue, with a bottom-up merge sort, using code contributed
3018 by Michael Haardt. This should make operations like -bp somewhat faster on
3019 large queues. It won't affect queue runners, except when queue_run_in_order
3022 51. Installed eximstats 1.31 in the distribution.
3024 52. Added support for SRV lookups to the dnslookup router.
3026 53. If an ACL referred to $message_body or $message_body_end, the value was not
3027 reset for any messages that followed in the same SMTP session.
3029 54. The store-handling optimization for building very long strings was not
3030 differentiating between the different store pools. I don't think this
3031 actually made any difference in practice, but I've tidied it.
3033 55. While running the routers to verify a sender address, $sender_address
3034 was still set to the sender address. This is wrong, because when routing to
3035 send a bounce to the sender, it would be empty. Therefore, I have changed
3036 it so that, while verifying a sender address, $sender_address is set to <>.
3037 (There is no change to what happens when verifying a recipient address.)
3039 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3040 target A or AAAA records (if not already returned) without resetting the
3041 qualify_single or search_parents options of the DNS resolver. These are
3042 inappropriate in this case because the targets of MX and SRV records must
3043 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3044 target that, when qualified, matched something in the local domain. These
3045 two options are now turned off when doing these lookups.
3047 57. It seems that at least some releases of Reiserfs (which does not have the
3048 concept of a fixed number of inodes) returns zero and not -1 for the
3049 number of available inodes. This interacted badly with check_spool_inodes,
3050 which assumed that -1 was the "no such thing" setting. What I have done is
3051 to check that the total number of inodes is greater than zero before doing
3052 the test of how many are available.
3054 58. When a "warn" ACL statement has a log_message modifier, the message is
3055 remembered, and not repeated. This is to avoid a lot of repetition when a
3056 message has many recipients that cause the same warning to be written.
3057 Howewer, Exim was preserving the list of already written lines for an
3058 entire SMTP session, which doesn't seem right. The memory is now reset if a
3059 new message is started.
3061 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3062 debugging output unless log_rewrite was also set.
3064 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3065 of (int)(handle) when we know that handle contains (void *)(-1).
3067 61. The Exim daemon panic-logs an error return when it closes the incoming
3068 connection. However "connection reset by peer" seems to be common, and
3069 isn't really an error worthy of noting specially, so that particular error
3072 62. When Exim is trying to find all the local interfaces, it used to panic and
3073 die if the ioctl to get the interface flags failed. However, it seems that
3074 on at least one OS (Solaris 9) it is possible to have an interface that is
3075 included in the list of interfaces, but for which you get a failure error
3076 for this call. This happens when the interface is not "plumbed" into a
3077 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3078 failure of the "get flags" call assumes that the interface is down.
3080 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3081 makes life easier for people who are doing arithmetic on fields extracted
3082 from dates, where you often get leading zeros that should not be
3083 interpreted as octal.
3085 64. Added qualify_domain to the redirect router, to override the global
3088 65. If a pathologically long header line contained very many addresses (the
3089 report of this problem mentioned 10 000) and each of them was rewritten,
3090 Exim could use up a very large amount of memory. (It kept on making new
3091 copies of the header line as it rewrote, and never released the old ones.)
3092 At the expense of a bit more processing, the header rewriting function has
3093 been changed so that it no longer eats memory in this way.
3095 66. The generation of the Received: header has been moved from the time that a
3096 message starts to be received, to the time that it finishes. The timestamp
3097 in the Received: header should now be very close to that of the <= log
3098 line. There are two side-effects of this change:
3100 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3101 logged header lines no longer include the local Received: line, because
3102 it has not yet been created. The same applies to a copy of the message
3103 that is returned to a non-SMTP sender when a message is rejected.
3105 (b) When a filter file is tested using -bf, no additional Received: header
3106 is added to the test message. After some thought, I decided that this
3109 This change does not affect the value of $received_for. It is still set
3110 after address rewriting, but before local_scan() is called.
3112 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3114 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3115 gave an unhelpful panic error message, and a defer error. I have managed to
3116 change this behaviour so that it now rejects any supplied certificate,
3117 which seems right, as the list of acceptable certificates is empty.
3119 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3120 gave an unhelpful defer error. I have not managed to make this reject any
3121 supplied certificates, but the error message it gives is "no certificate
3122 supplied", which is not helpful.
3124 70. exigrep's output now also includes lines that are not associated with any
3125 message, but which match the given pattern. Implemented by a patch from
3126 Martin Sluka, which also tidied up the Perl a bit.
3128 71. Recipient callout verification, like sender verification, was using <> in
3129 the MAIL FROM command. This isn't really the right thing, since the actual
3130 sender may affect whether the remote host accepts the recipient or not. I
3131 have changed it to use the actual sender in the callout; this means that
3132 the cache record is now keyed on a recipient/sender pair, not just the
3133 recipient address. There doesn't seem to be a real danger of callout loops,
3134 since a callout by the remote host to check the sender would use <>.
3135 [SEE ABOVE: changed after hitting problems.]
3137 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3138 temporary errors. However, in the case of such a code being given after
3139 the end of a data transmission (i.e. after ".") Exim was failing to write
3140 a retry record for the message. (Yes, there was some broken host that was
3141 actually sending 8xx at this point.)
3143 73. An unknown lookup type in a host list could cause Exim to panic-die when
3144 the list was checked. (An example that provoked this was putting <; in the
3145 middle of a list instead of at the start.) If this happened during a DATA
3146 ACL check, a -D file could be left lying around. This kind of configuration
3147 error no longer causes Exim to die; instead it causes a defer errror. The
3148 incident is still logged to the main and panic logs.
3150 74. Buglet left over from Exim 3 conversion. The message "too many messages
3151 in one connection" was written to the rejectlog but not the mainlog, except
3152 when address rewriting (yes!) was being logged.
3154 75. Added write_rejectlog option.
3156 76. When a system filter was run not as root (that is, when system_filter_user
3157 was set), the values of the $n variables were not being returned to the
3158 main process; thus, they were not subsequently available in the $sn
3161 77. Added +return_path_on_delivery log selector.
3163 78. A connection timeout was being treated differently from recipients deferred
3164 when testing hosts_max_try with a message that was older than the host's
3165 retry timeout. (The host should not be counted, thus allowing all hosts to
3166 be tried at least once before bouncing.) This may have been the cause of an
3167 occasionally reported bug whereby a message would remain on the queue
3168 longer than the retry timeout, but would be bounced if a delivery was
3169 forced. I say "may" because I never totally pinned down the problem;
3170 setting up timeout/retry tests is difficult. See also the next item.
3172 79. The ultimate address timeout was not being applied to errors that involved
3173 a combination of host plus message (for example, a timeout on a MAIL
3174 command). When an address resolved to a number of possible hosts, and they
3175 were not all tried for each delivery (e.g. because of hosts_max_try), a
3176 message could remain on the queue longer than the retry timeout.
3178 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3181 81. Fixed an obscure SMTP outgoing bug which required at least the following
3182 conditions: (a) there was another message waiting for the same server;
3183 (b) the server returned 5xx to all RCPT commands in the first message so
3184 that the message was not completed; (c) the server dropped the connection
3185 or gave a negative response to the RSET that Exim sends to abort the
3186 transaction. The observed case was a dropped connection after DATA that had
3187 been sent in pipelining mode. That is, the server had advertised PIPELINING
3188 but was not implementing it correctly. The effect of the bug was incorrect
3189 behaviour, such as trying another host, and this could lead to a crash.
3195 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3196 and daemon.c were passed as pointers to ints; they should have been
3197 pointers to socklen_t variables (which are typically unsigned ints).
3199 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3202 3. Fixed a really odd bug that affected only the testing scheme; patching a
3203 certain fixed string in the binary changed the value of another string that
3204 happened to be identical to the end of the original first string.
3206 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3207 name", it returns that address as the IP address. On some operating
3208 systems (e.g. Solaris), it also passes back the IP address string as the
3209 "host name". However, on others (e.g. Linux), it passes back an empty
3210 string. Exim wasn't checking for this, and was changing the host name to an
3211 empty string, assuming it had been canonicized.
3213 5. Although rare, it is permitted to have more than one PTR record for a given
3214 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3215 all the names associated with an address, because they do in Solaris.
3216 However, it seems that they do not in Linux for data that comes from the
3217 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3218 I found this out when I moved to a new Linux workstation and tried to run
3219 the Exim test suite.
3221 To get round this problem I have changed the code so that it now does its
3222 own call to the DNS to look up PTR records when searching for a host name.
3223 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3224 addresses that are only in /etc/hosts are still found.
3226 This behaviour is, however, controlled by an option called host_lookup_
3227 order, which defaults to "bydns:byaddr". If people want to use the other
3228 order, or indeed, just use one or the other means of lookup, they can
3229 specify it in this variable.
3231 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3232 some operating systems, this comes back from gethostbyaddr() as an empty
3233 string, and this is what Exim used to test for. However, it seems that in
3234 other systems, "." is yielded. Exim now tests for this case too.
3236 7. The values of check_spool_space and check_log_space are now held internally
3237 as a number of kilobytes instead of an absolute number of bytes. If a
3238 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3239 kilobyte. This means that much larger values can be stored.
3241 8. Exim monitor: an attempt to get the action menu when not actually pointing
3242 at a message produces an empty menu entitled "No message selected". This
3243 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3244 no entries in it ("Shell widget menu has zero width and/or height"). So I
3245 have added a single, blank menu entry in this case.
3247 9. Added ${quote_local_part.
3249 10. MIME decoding is now applied to the contents of Subject: header lines when
3252 11. Now that a reference to $sender_host_address automatically causes a reverse
3253 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3254 host lookup before query-style lookups in lists that might use this
3255 variable. This has therefore been abolished, and the "net-" prefix is no
3256 longer necessary for query-style lookups.
3258 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3259 have been a typo for LFLAGS, so it has been changed.
3261 13. The install script calls Exim with "-C /dev/null" in order to find the
3262 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3263 to be output. Howeve, since Exim outputs its version number before the
3264 error, it didn't break the script. It just looked ugly. I fixed this by
3265 always allowing "-C /dev/null" if the caller is root.
3267 14. Ignore overlarge ACL variable number when reading spool file - insurance
3268 against a later release with more variables having written the file.
3270 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3271 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3272 this, as well as the form without the "IPv6" on the front.
3274 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3275 OS/Makefile-Darwin file.
3277 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3278 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3280 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3281 variable. However, some OS use ints for inodes. I've added cast to long int
3282 to get rid of the compiler warning.
3284 19. I had forgotten to lock out "/../" in configuration file names when
3285 ALT_CONFIG_PREFIX was set.
3287 20. Routers used for verification do not need to specify transports. However,
3288 if such a router generated a host list, and callout was configured, Exim
3289 crashed, because it could not find a port number from the (non-existent)
3290 transport. It now assumes port 25 in this circumstance.
3292 21. Added the -t option to exigrep.
3294 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3295 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3296 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3297 them is set, LOOKUP_LSEARCH is forced.
3299 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3301 24. Added sender and host information to the "rejected by local_scan()" log
3302 line; previously there was no indication of these.
3304 25. Added .include_if_exists.
3306 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3307 Linux. It turns out that not all file systems support this. Apparently some
3308 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3309 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3310 not supported on the file descriptor, is now ignored when Exim is trying to
3311 sync a directory. This applies only to Linux.
3313 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3315 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3316 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3319 29. Added support for maildirsize files from supplied patch (modified a bit).
3321 30. The use of :fail: followed by an empty string could lead Exim to respond to
3322 sender verification failures with (e.g.):
3324 550 Verification failed for <xxx>
3325 550 Sender verify failed
3327 where the first response line was missing the '-' that indicates it is not
3328 the final line of the response.
3330 31. The loop for finding the name of the user that called Exim had a hardwired
3331 limit of 10; it now uses the value of finduser_retries, which is used for
3332 all other user lookups.
3334 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3337 33. Exim was neglecting to zero errno before one call of strtol() when
3338 expanding a string and expecting an integer value. On some systems this
3339 resulted in spurious "integer overflow" errors. Also, it was casting the
3340 result into an int without checking.
3342 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3343 did not work. The code looks as if it has *never* worked, though it appears
3344 to have been documented since at least releast 1.62. I have made it work.
3346 35. The "timeout_DNS" error in retry rules, also documented since at least
3347 1.62, also never worked. As it isn't clear exactly what this means, and
3348 clearly it isn't a major issue, I have abolished the feature by treating it
3349 as "timeout", and writing a warning to the main and panic logs.
3351 36. The display of retry rules for -brt wasn't always showing the error code
3354 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3355 timeout_connect_A, timeout_connect_MX.
3357 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3358 to the empty sender.
3360 39. The daemon was not analysing the content of -oX till after it had closed
3361 stderr and disconnected from the controlling terminal. This meant that any
3362 syntax errors were only noted on the panic log, and the return code from
3363 the command was 0. By re-arranging the code a little, I've made the
3364 decoding happen first, so such errors now appear on stderr, and the return
3365 code is 1. However, the actual setting up of the sockets still happens in
3366 the disconnected process, so errors there are still only recorded on the
3369 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3370 connections (as happens on some IP stacks) was logged at start up time as
3371 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3372 it from "IPv6 and IPv4", which means that two separate sockets are being
3375 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3376 says whether AF_INET or AF_INET6 was passed as an argument.
3378 42. Exiwhat output was messed up when time zones were included in log
3381 43. Exiwhat now gives more information about the daemon's listening ports,
3382 and whether -tls-on-connect was used.
3384 44. The "port" option of the smtp transport is now expanded.
3386 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3387 silently ignored. Now an error message is written to the main and panic
3390 46. There's a new ACL modifier called "logwrite" which writes to a log file
3391 as soon as it is encountered.
3393 47. Added $local_user_uid and $local_user_gid at routing time.
3395 48. Exim crashed when trying to verify a sender address that was being
3398 49. Exim was recognizing only a space character after ".include". It now also
3399 recognizes a tab character.
3401 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3402 extracting the relevant information from the specification. The man page no
3403 longer contains scrambled data for the -d option, and I've added a section
3404 at the front about calling Exim under different names.
3406 51. Added "extra_headers" argument to the "mail" command in filter files.
3408 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3411 53. Installed eximstats 1.29.
3413 54. Added transport_filter_timeout as a generic transport option.
3415 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3416 Cc: header lines. This was required by RFC 822, but it not required by RFC
3419 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3420 incoming messages that did not have them. Now it does so only if the
3421 message originates locally, that is, if there is no associated remote host
3422 address. When Resent- header lines are present, this applies to the Resent-
3423 lines rather than the non-Resent- lines.
3425 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3426 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3428 58. Messages for configuration errors now include the name of the main
3429 configuration file - useful now that there may be more than one file in a
3430 list (.included file names were always shown).
3432 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3433 for those rare installations that do not start the daemon as root or run it
3434 setuid root. I've cut out the call to initgroups() if the daemon is not
3437 60. The Exim user and group can now be bound into the binary as text strings
3438 that are looked up at the start of Exim's processing.
3440 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3442 62. Added $mailstore_basename variable.
3444 63. Installed patch to sieve.c from Michael Haardt.
3446 64. When Exim failed to open the panic log after failing to open the main log,
3447 the original message it was trying to log was written to stderr and debug
3448 output, but if they were not available (the usual case in production), it
3449 was lost. Now it is written to syslog before the two lines that record the
3450 failures to open the logs.
3452 65. Users' Exim filters run in subprocesses under the user's uid. It is
3453 possible for a "deliver" command or an alias in a "personal" command to
3454 provoke an address rewrite. If logging of address rewriting is configured,
3455 this fails because the process is not running as root or exim. There may be
3456 a better way of dealing with this, but for the moment (because 4.30 needs
3457 to be released), I have disabled address rewrite logging when running a
3458 filter in a non-root, non-exim process.
3464 1. The buildconfig auxiliary program wasn't quoting the value set for
3465 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3466 not defined. This bug was masked in 4.22 by the effect that was fixed in
3469 2. Some messages that were rejected after a message id was allocated were
3470 shown as "incomplete" by exigrep. It no longer does this for messages that
3471 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3473 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3474 have allow_domain_literals set, the ID did not get logged in the <= line.
3475 Domain literals are now always recognized in Message-ID: header lines.
3477 4. The first argument for a ${extract expansion item is the key name or field
3478 number. Leading and trailing spaces in this item were not being ignored,
3479 causing some misleading effects.
3481 5. When deliver_drop_privilege was set, single queue runner processes started
3482 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3483 same command in the process it spins off) were not dropping privilege.
3485 6. When the daemon running as "exim" started a queue runner, it always
3486 re-executed Exim in the spun-off process. This is a waste of effort when
3487 deliver_drop_privilege is set. The new process now just calls the
3488 queue-runner function directly.
3494 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3497 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3498 ignored. Though the use of -oP was forcing the writing of a pid file, it
3499 was always written to the default place.
3501 3. If the message "no IP address found for host xxxx" is generated during
3502 incoming verification, it is now followed by identification of the incoming
3503 connection (so you can more easily find what provoked it).
3505 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3507 5. Added some features to "harden" Exim a bit more against certain attacks:
3509 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3510 be put in Local/Makefile. This is like the never_users runtime option,
3511 but it cannot be overridden. The default setting is "root".
3513 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3514 prefix string with which any file named in a -C command line option
3517 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3518 is retained for -C and -D only if the caller of Exim is root. Without
3519 it, the exim user may also use -C and -D and retain privilege.
3521 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3522 command line option is disabled.
3524 6. Macro names set by the -D option must start with an upper case letter, just
3525 like macro names defined in the configuration file.
3527 7. Added "dereference=" facility to LDAP.
3529 8. Two instances of the typo "uknown" in the source files are fixed.
3531 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3532 the Configure-Makefile script screwed up while processing it.
3534 10. Incorporated PCRE 4.4.
3536 11. The SMTP synchronization check was not operating right at the start of an
3537 SMTP session. For example, it could not catch a HELO sent before the client
3538 waited for the greeting. There is now a check for outstanding input at the
3539 point when the greeting is written. Because of the duplex, asynchronous
3540 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3541 way, but not yet received, when the check is performed.
3543 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3544 on TCP/IP sockets, because this apparently causes some broken clients to
3547 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3548 unchanged) from the Cygwin maintainer.
3550 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3551 was supported instead of testing for mbx. Effectively a typo.
3553 15. The spa authenticator server code was not checking that the input it
3554 received was valid base64.
3556 16. The debug output line for the "set" modifier in ACLs was not showing the
3557 name of the variable that was being set.
3559 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3561 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3562 lookup on its own. Something else was needed to trigger the lookup. For
3563 example, a match in host_lookup or the need for a host name in a host list.
3564 Now, if $sender_host_name is referenced and the host name has not yet been
3565 looked up, a lookup is performed. If the lookup fails, the variable remains
3566 empty, and $host_lookup_failed is set to "1".
3568 19. Added "eqi" as a case-independent comparison operator.
3570 20. The saslauthd authentication condition could segfault if neither service
3571 nor realm was specified.
3573 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3574 error message that was logged was misleading, and incoming SMTP
3575 connections were dropped. The message is now more accurate, and temporary
3576 errors are given to SMTP connections.
3578 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3579 (except RSET) with a 421 error, until QUIT is received. However, it was
3580 failing to send a response to QUIT.
3582 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3583 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3584 ACL is now run after the helo_try_verify_hosts code.
3586 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3587 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3588 case-independent, so other case variants are also recognized). Apparently
3589 some systems use these upper case variants.
3591 25. If more than two messages were waiting for the same host, and a transport
3592 filter was specified for the transport, Exim sent two messages over the
3593 same TCP/IP connection, and then failed with "socket operation on non-
3594 socket" when it tried to send the third.
3596 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3598 27. The extern definition of crypt16() in expand.c was not being excluded when
3599 the OS had its own crypt16() function.
3601 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3602 as a preferred synonym for return_size_limit, both as an option and as an
3605 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3607 30. Changed the default configuration ACL to relax the local part checking rule
3608 for addresses that are not in any local domains. For these addresses,
3609 slashes and pipe symbols are allowed within local parts, but the sequence
3610 /../ is explicitly forbidden.
3612 31. SPA server authentication was not clearing the challenge buffer before
3615 32. log_message in a "warn" ACL statement was writing to the reject log as
3616 well as to the main log, which contradicts the documentation and doesn't
3617 seem right (because no rejection is happening). So I have stopped it.
3619 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3620 However, I am unable to do any testing of this.
3622 34. Fixed an infelicity in the appendfile transport. When checking directories
3623 for a mailbox, to see if any needed to be created, it was accidentally
3624 using path names with one or more superfluous leading slashes; tracing
3625 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3627 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3628 contents are added to the log line that is written for every discarded
3629 recipient. (Previously a log_message setting was ignored.)
3631 36. The ${quote: operator now quotes the string if it is empty.
3633 37. The install script runs exim in order to find its version number. If for
3634 some reason other than non-existence or emptiness, which it checks, it
3635 could not run './exim', it was installing it with an empty version number,
3636 i.e. as "exim-". This error state is now caught, and the installation is
3639 38. An argument was missing from the function that creates an error message
3640 when Exim fails to connect to the socket for saslauthd authentication.
3641 This could cause Exim to crash, or give a corrupted message.
3643 39. Added isip, isip4, and isip6 to ${if conditions.
3645 40. The ACL variables $acl_xx are now saved with the message, and can be
3646 accessed later in routers, transports, and filters.
3648 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3649 strings in the file are not string-expanded.
3651 42. If a MAIL command specified a SIZE value that was too large to fit into an
3652 int variable, the check against message_size_limit failed. Such values are
3653 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3654 day this will have to be increased, but I don't think I want to be around
3655 when emails are that large.
3662 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3663 iconv() is not standard in FreeBSD.
3665 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3666 IPv6 enabled. The observed symptom was a segmentation fault on return from
3667 the function os_common_find_running_interfaces() in src/os.c.
3669 3. In the check_special_case() function in daemon.c I had used "errno" as an
3670 argument name, which causes warnings on some systems. This was basically a
3671 typo, since it was named "eno" in the comments!
3673 4. The code that waits for the clock to tick (at a resolution of some fraction
3674 of a second) so as to ensure message-id uniqueness was always waiting for
3675 at least one whole tick, when it could have waited for less. [This is
3676 almost certainly not relevant at current processor speeds, where it is
3677 unlikely to ever wait at all. But we try to future-proof.]
3679 5. The function that sleeps for a time interval that includes fractions of a
3680 second contained a race. It did not block SIGALRM between setting the
3681 timer, and suspending (a couple of lines later). If the interval was short
3682 and the sigsuspend() was delayed until after it had expired, the suspension
3683 never ended. On busy systems this could lead to processes getting stuck for
3686 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3687 process, before it forks any delivery processes. The open lookup caching
3688 mechanism meant that the open file or database connection was passed into
3689 the delivery process. The problem was that delivery processes always tidy
3690 up cached lookup data. This could cause a problem for the next delivery
3691 process started by the queue runner, because the external queue runner
3692 process does not know about the closure. So the next delivery process
3693 still has data in the lookup cache. In the case of a file lookup, there was
3694 no problem because closing a file descriptor in a subprocess doesn't affect
3695 the parent. However, if the lookup was caching a connection to a database,
3696 the connection was closed, and the second delivery process was likely to
3697 see errors such as "PGSQL: query failed: server closed the connection
3698 unexpectedly". The problem has been fixed by closing all cached lookups
3699 in a queue runner before running a delivery process.
3701 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3702 seem to have the "const" qualifier which it has on other OS. I've
3705 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3706 *of the same type* (client or server) with the same public name that an
3707 error should be diagnosed.
3709 9. When Exim looked up a host name for an IP address, but failed to find the
3710 original IP address when looking up the host name (a safety check), it
3711 output the message "<ip address> does not match any IP for NULL", which was
3712 confusing, to say the least. The bug was that the host name should have
3713 appeared instead of "NULL".
3715 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3716 user that is built into the binary, and the -C or -D options is used, root
3717 privilege is dropped before the configuration file is read. In addition,
3718 logging is switched to stderr instead of the normal log files. If the
3719 configuration then re-defines the Exim user, the unprivileged environment
3720 is probably not what is expected, so Exim logs a panic warning message (but
3723 However, if deliver_drop_privilege is set, the unprivileged state may well
3724 be exactly what is intended, so the warning has been cut out in that case,
3725 and Exim is allowed to try to write to its normal log files.
3731 1. smtp_return_error_details was not giving details for temporary sender
3732 or receiver verification errors.
3734 2. Diagnose a configuration error if two authenticators have the same public
3737 3. Exim used not to create the message log file for a message until the first
3738 delivery attempt. This could be confusing when incoming messages were held
3739 for policy or load reasons. The message log file is now created at the time
3740 the message is received, and an initial "Received" line is written to it.
3742 4. The automatically generated man page for command line options had a minor
3743 bug that caused no ill effects; however, a more serious problem was that
3744 the procedure for building the man page automatically didn't always
3745 operate. Consequently, release 4.20 contains an out-of-date version. This
3746 shouldn't happen again.
3748 5. When building Exim with embedded Perl support, the script that builds the
3749 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3750 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3752 6. The freeze_tell option was not being used for messages that were frozen on
3753 arrival, either by an ACL or by local_scan().
3755 7. Added the smtp_incomplete_transaction log selector.
3757 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3758 was accepting AUTH without a new EHLO.
3760 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3761 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3764 10. Logging of TCP/IP connections (when configured) now happens in the main
3765 daemon process instead of the child process, so that the TCP/IP connection
3766 count is more accurate (but it can never be perfect).
3768 11. The use of "drop" in a nested ACL was not being handled correctly in the
3769 outer ACL. Now, if condition failure induced by the nested "drop" causes
3770 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3771 or "require"), the connection is dropped.
3773 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3774 that yield "discard" can now be used with an "accept" or a "discard" verb,
3775 but an error is generated for any others (because I can't see a useful way
3776 to define what should happen).
3778 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3779 are now processed in the same way as lines in the Exim configuration file.
3780 In particular, continuation lines are supported.
3782 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3784 15. Added -ti meaning -t -i.
3786 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3787 domains, and warn by logging if others are found.
3789 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3790 returned from GIFCONF when Exim is trying to find the list of interfaces on
3791 a host. The code in os.c has been modified to copy each ifreq to an aligned
3792 structure in all cases.
3794 Also, in some cases, the returned ifreq's were being copied to a 'struct
3795 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3796 means the last couple of bytes of an IPv6 address could be chopped if the
3797 ifreq contained only a normal sockaddr (14 bytes storage).
3799 18. Named domain lists were not supported in the hosts_treat_as_local option.
3800 An entry such as +xxxx was not recognized, and was treated as a literal
3803 19. Ensure that header lines added by a DATA ACL are included in the reject log
3804 if the ACL subsequently rejects the message.
3806 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3807 MD5 (which is deprecated).
3809 21. When testing a filter file using -bf, Exim was writing a message when it
3810 took the sender from a "From " line in the message, but it was not doing so
3811 when it took $return_path from a Return-Path: header line. It now does.
3813 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3814 with a valid header line field name (a series of printing characters
3815 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3817 23. Changed "disc" in the source to "disk" to conform to the documentation and
3818 the book and for uniformity.
3820 24. Ignore Sendmail's -Ooption=value command line item.
3822 25. When execve() failed while trying to run a command in a pipe transport,
3823 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3824 could be confused with a return value of 69 from the command itself. This
3825 has been changed to 127, the value the shell returns if it is asked to run
3826 a non-existent command. The wording for the related log line suggests a
3827 non-existent command as the problem.
3829 26. If received_header_text expands to an empty string, do not add a Received:
3830 header line to the message. (Well, it adds a token one on the spool, but
3831 marks it "old" so that it doesn't get used or transmitted.)
3833 27. Installed eximstats 1.28 (addition of -nt option).
3835 28. There was no check for failure on the call to getsockname() in the daemon
3836 code. This can fail if there is a shortage of resources on the system, with
3837 ENOMEM, for example. A temporary error is now given on failure.
3839 29. Contrary to the C standard, it seems that in some environments, the
3840 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3841 program. Exim now does this explicitly; it affects the formatting of
3842 timestamps using strftime().
3844 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3845 complaints. I've now initialized all the variables, to avoid this.
3847 32. Header lines added by a system filter were not being "seen" during
3848 transport-time rewrites.
3850 33. The info_callback() function passed to OpenSSL is set up with type void
3851 (*)(SSL *, int, int), as described somewhere. However, when calling the
3852 function (actually a macro) that sets it up, the type void(*)() is
3853 expected. I've put in a cast to prevent warnings from picky compilers.
3855 34. If a DNS black list lookup found a CNAME record, but there were no A
3856 records associated with the domain it pointed at, Exim crashed.
3858 35. If a DNS black list lookup returned more than one A record, Exim ignored
3859 all but the first. It now scans all returned addresses if a particular IP
3860 value is being sought. In this situation, the contents of the
3861 $dnslist_value variable are a list of all the addresses, separated by a
3864 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3865 to accept a host name and all its aliases if the forward lookup for any of
3866 them yielded the IP address of the incoming connection. Now it accepts only
3867 those names whose forward lookup yields the correct IP address. Any other
3868 names are discarded. This closes a loophole whereby a rogue DNS
3869 administrator could create reverse DNS records to break through a
3870 wildcarded host restriction in an ACL.
3872 37. If a user filter or a system filter that ran in a subprocess used any of
3873 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3874 the wrong values were passed to the pipe command ($thisaddress had the
3875 value of $0, $0 had the value of $1, etc). This bug was introduced by
3876 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3878 38. Improved the line breaking for long SMTP error messages from ACLs.
3879 Previously, if there was no break point between 40 and 75 characters, Exim
3880 left the rest of the message alone. Two changes have been made: (a) I've
3881 reduced the minimum length to 35 characters; (b) if it can't find a break
3882 point between 35 and 75 characters, it looks ahead and uses the first one
3883 that it finds. This may give the occasional overlong line, but at least the
3884 remaining text gets split now.
3886 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3887 file descriptors might be low, and that setting 1000 would always raise it.
3888 It turns out that in some environments, the limit is already over 1000 and
3889 that lowering it causes trouble. So now Exim takes care not to decrease it.
3891 40. When delivering a message, the value of $return_path is set to $sender_
3892 address at the start of routing (routers may change the value). By an
3893 oversight, this default was not being set up when an address was tested by
3894 -bt or -bv, which affected the outcome if any router or filter referred to
3897 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3898 log only when "message" or "log_message" are set. However, if one of the
3899 conditions was an address verification, or a call to a nested ACL, the
3900 messages generated by the underlying test were being passed through. This
3901 no longer happens. The underlying message is available in $acl_verify_
3902 message for both "message" and "log_message" expansions, so it can be
3903 passed through if needed.
3905 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3906 new expansion $bh_ to give the encoded byte string without charset
3907 translation. Translation happens only if iconv() is available; HAVE_ICONV
3908 indicates this at build time. HEADERS_CHARSET gives the charset to
3909 translate to; headers_charset can change it in the configuration, and
3910 "headers charset" can change it in an individual filter file.
3912 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3913 that creates RFC 2047 encoded "words" labels them as that charset instead
3914 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3915 expansion operator; (ii) when Exim creates a From: line for a local
3916 message; (iii) when a header line is rewritten to include a "phrase" part.
3918 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3919 buggy, causing it to skip the first lines of messages whose message ID
3920 ended in 'D'. This would not have bitten before Exim release 4.14, because
3921 message IDs were unlikely to end in 'D' before then. The effect was to have
3922 incorrect size information for certain domains.
3924 45. #include "config.h" was missing at the start of the crypt16.c module. This
3925 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3928 46. If there was a timeout during a "random" callout check, Exim treated it as
3929 a failure of the random address, and carried on sending RSET and the real
3930 address. If the delay was just some slowness somewhere, the response to the
3931 original RCPT would be taken as a response to RSET and so on, causing
3932 mayhem of various kinds.
3934 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3935 when I implemented it. It didn't allow for the fact that some option values
3936 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3937 the right test for positive values.
3939 48. Domain names in DNS records are case-independent. Exim always looks them up
3940 in lower case. Some resolvers return domain names in exactly the case they
3941 appear in the zone file, that is, they may contain uppercase letters. Not
3942 all resolvers do this - some return always lower case. Exim was treating a
3943 change of case by a resolver as a change of domain, similar to a widening
3944 of a domain abbreviation. This triggered its re-routing code and so it was
3945 trying to route what was effectively the same domain again. This normally
3946 caused routing to fail (because the router wouldn't handle the domain
3947 twice). Now Exim checks for this case specially, and just changes the
3948 casing of the domain that it ultimately uses when it transmits the message
3951 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
3954 50. If a filter generated a file delivery with a non-absolute name (possible if
3955 no home directory exists for the router), the forbid_file option was not
3958 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
3959 the existing equality matching.
3961 52. Exim was using ints instead of ino_t variables in some places where it was
3962 dealing with inode numbers.
3964 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
3965 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
3966 called TMPDIR, and if it finds it is different, it changes its value.
3968 54. The smtp_printf() function is now made available to local_scan() so
3969 additional output lines can be written before returning. There is also an
3970 smtp_fflush() function to enable the detection of a dropped connection.
3971 The variables smtp_input and smtp_batched_input are exported to
3974 55. Changed the default runtime configuration: the message "Unknown user"
3975 has been removed from the ACL, and instead placed on the localuser router,
3976 using the cannot_route_message feature. This means that any verification
3977 failures that generate their own messages won't get overridden. Similarly,
3978 the "Unrouteable address" message that was in the ACL for unverifiable
3979 relay addresses has also been removed.
3981 56. Added hosts_avoid_esmtp to the smtp transport.
3983 57. The exicyclog script was not checking for the esoteric option
3984 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
3985 will work only if exicyclog is run under the appropriate euid.
3987 58. Following a discussion on the list, the rules by which Exim recognises line
3988 endings on incoming messages have been changed. The -dropcr and drop_cr
3989 options are now no-ops, retained only for backwards compatibility. The
3990 following line terminators are recognized: LF CRLF CR. However, special
3991 processing applies to CR:
3993 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
3994 nor a local message in the state where . is a terminator.
3996 (ii) If a bare CR is encountered in a header line, an extra space is added
3997 after the line terminator so as not to end the header. The reasoning
3998 behind this is that bare CRs in header lines are most likely either
3999 to be mistakes, or people trying to play silly games.
4001 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4002 was being incorrectly given as 18 bytes larger than it should have been.
4003 This is a VOB (very old bug).
4005 60. This may never have affected anything current, but just in case it has:
4006 When the local host is found other than at the start of a list of hosts,
4007 the local host, those with the same MX, and any that follow, are discarded.
4008 When the list in question was part of a longer list of hosts, the following
4009 hosts (not currently being processed) were also being discarded. This no
4010 longer happens. I'm not sure if this situation could ever has previously
4013 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4016 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4017 header. This is something that is recommended in a new Internet Draft, and
4018 is something that is documented as being done by Sendmail. There are two
4019 possible values. For messages generated by the autoreply transport, Exim
4022 Auto-Submitted: auto-replied
4024 whereas for all other generated messages (e.g. bounces) it adds
4026 Auto-Submitted: auto-generated
4028 63. The "personal" condition in filters now includes a test for the
4029 Auto-Submitted: header. If it contains the string "auto-" the message it
4030 not considered personal.
4032 64. Added rcpt_include_affixes as a generic transport option.
4034 65. Added queue_only_override (default true).
4036 66. Added the syslog_duplication option.
4038 67. If what should have been the first header line of a message consisted of
4039 a space followed by a colon, Exim was mis-interpreting it as a header line.
4040 It isn't of course - it is syntactically invalid and should therefore be
4041 treated as the start of the message body. The misbehaviour could have
4042 caused a number of strange effects, including loss of data in subsequent
4043 header lines, and spool format errors.
4045 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4046 client host had authenticated. This control can now be exercised by an ACL
4047 for more flexibility.
4049 69. By default, callouts do not happen when testing with -bh. There is now a
4050 variant, -bhc, which does actually run the callout code, including
4051 consulting and updating the callout cache.
4053 70. Added support for saslauthd authentication, courtesy of Alexander
4056 71. If statvfs() failed on the spool or log directories while checking their
4057 size for availability, Exim confusingly gave the error "space shortage".
4058 Furthermore, in debugging mode it crashed with a floating point exception.
4059 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4060 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4061 serious problem, Exim now writes to the main and panic logs when this
4062 happens, with details of the failure. It then refuses to accept the
4063 incoming message, giving the message "spool directory problem" or "log
4064 directory problem" with a 421 code for SMTP messages.
4066 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4067 0, 1, and 2 exist, because some OS complain for execs without them (see
4068 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4069 descriptors. However, the code omitted to check that the open succeeded,
4070 causing mysterious errors if for some reason the permissions on /dev/null
4071 got screwed. Now Exim writes a message to the main and panic logs, and
4072 bombs out if it can't open /dev/null.
4074 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4075 interact so that it is all more flexible. It is supposed to remain
4076 backwards compatible. Also added extra_local_interfaces.
4078 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4079 to bomb out with an assertion failure - to the client this appears as a
4080 connection drop. This problem occurs in the part of the code that was taken
4081 from the Samba project. Fortunately, the assertion is in a very simple
4082 function, so I have fixed this by reproducing the function inline in the
4083 one place where it is called, and arranging for authentication to fail
4084 instead of killing the process with assert().
4086 75. The SPA client code was not working when the server requested OEM rather
4087 than Unicode encoding.
4089 76. Added code to make require_files with a specific uid setting more usable in
4090 the case where statting the file as root fails - usually a non-root-mounted
4091 NFS file system. When this happens and the failure is EACCES, Exim now
4092 forks a subprocess and does the per-uid checking as the relevant uid.
4094 77. Added process_log_path.
4096 78. If log_file_path was not explicitly set, a setting of check_log_space or
4097 check_log_inodes was ignored.
4099 79. If a space check for the spool or log partitions fails, the incident is now
4100 logged. Of course, in the latter case the data may get lost...
4102 80. Added the %p formatting code to string_format() so that it can be used to
4103 print addresses in debug_print(). Adjusted all the address printing in the
4104 debugging in store.c to use %p rather than %d.
4106 81. There was a concern that a line of code in smtp_in.c could overflow a
4107 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4108 initially expressed, the concern was not well-founded, because trailing
4109 spaces are removed early. However, if the trailing spaces were followed by
4110 a NULL, they did not get removed, so the overflow was possible. Two fixes
4113 (a) I re-wrote the offending code in a cleaner fashion.
4114 (b) If an incoming SMTP command contains a NULL character, it is rejected
4117 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4118 runs initgroups(), so that if the Exim user is in any additional groups,
4119 they will be used during message reception.
4125 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob