1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.397 2006/09/25 10:14:20 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Efrurth's refactoring patch to tidy up spool_mbox.c, and
38 later the small subsequent patch to fix an introduced bug.
40 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
42 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
44 PH/08 An error is now given if message_size_limit is specified negative.
46 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
47 to be given (somewhat) arbitrary names.
49 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
52 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
53 miscellaneous code fixes
55 PH/10 Added the log_reject_target ACL modifier to specify where to log
63 SC/01 Use a glob alias rather than an array ref in eximstats generated
64 parser. This improves both readability and performance.
66 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
67 Don't display local sender or destination tables in eximstats unless
68 there is data to show.
69 Added average volumes into the eximstats top table text output.
71 SC/03 Collect data on the number of addresses (recipients) as well
72 as the number of messages in eximstats.
74 TF/01 Correct an error in the documentation for the redirect router. Exim
75 does (usually) call initgroups() when daemonizing.
77 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
78 with consistent privilege compared to when running as a daemon.
80 TF/03 Note in the spec that $authenticated_id is not set for local
81 submissions from trusted users.
83 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
84 Thanks to Dean Brooks <dean@iglou.com> for the patch.
86 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
87 by adding some example configuration directives to the default
88 configuration file. A little bit of work is required to uncomment the
89 directives and define how usernames and passwords are checked, but
90 there is now a framework to start from.
92 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
93 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
94 without this. I don't know how relevant this is to other LDAP libraries.
96 PH/02 Add the verb name to the "unknown ACL verb" error.
98 PH/03 Magnus Holmgren's patch for filter_prepend_home.
100 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
102 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
103 directory not expanded when it should be if an expanded home directory
104 was set for the address (which is overridden by the transport).
106 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
109 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
110 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
111 because it is too late at that time, and has no effect.
113 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
114 security issue with \' (bugzilla #107). I could not use the
115 PQescapeStringConn() function, because it needs a PGconn value as one of
118 PH/08 When testing addresses using -bt, indicate those final addresses that
119 are duplicates that would not cause an additional delivery. At least one
120 person was confused, thinking that -bt output corresponded to deliveries.
121 (Suppressing duplicates isn't a good idea as you lose the information
122 about possibly different redirections that led to the duplicates.)
124 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
125 systems where poll() doesn't work, in particular OS X.
127 PH/10 Added more information to debugging output for retry time not reached.
129 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
130 operations in malware.c.
132 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
135 PH/13 If write_rejectlog was set false when logging was sent to syslog with
136 syslog_duplication set false, log lines that would normally be written
137 both the the main log and to the reject log were not written to syslog at
140 PH/14 In the default configuration, change the use of "message" in ACL warn
141 statements to "add_header".
143 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
144 not followed by a command (e.g. "seen endif").
146 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
147 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
150 PH/17 Added extra conditions to the default value of delay_warning_condition
154 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
155 { match{$h_precedence:}{(?i)bulk|list|junk} } \
156 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
159 The Auto-Submitted: and various List- headers are standardised, whereas I
160 don't think Precedence: ever was.
162 PH/18 Refactored debugging code in route_finduser() to show more information,
163 in particular, the error code if getpwnam() issues one.
165 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
166 This is apparently needed in addition to the PH/07 change above to avoid
167 any possible encoding problems.
169 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
170 but not after initializing Perl.
172 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
173 output them only if debugging. By default they are written stderr,
174 apparently, which is not desirable.
176 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
179 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
182 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
184 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
185 authenticated or an ident call has been made. Suppress the default
186 values for $authenticated_id and $authenticated_sender (but permit -oMai
187 and -oMas) when testing with -bh.
189 PH/24 Re-jigged the order of the tests in the default configuration so that the
190 tests for valid domains and recipients precede the DNS black list and CSA
191 tests, on the grounds that those ones are more expensive.
193 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
194 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
195 This bug exists in every version of Exim that I still have, right back to
198 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
199 However, an attempt to turn on case-sensitivity in a regex key by
200 including (?-i) didn't work because the subject string was already
201 lowercased, and the effects were non-intuitive. It turns out that a
202 one-line patch can be used to allow (?-i) to work as expected.
208 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
209 other effects) broke the use of negated acl sub-conditions.
211 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
214 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
215 "Deny" causes Exim to reject the incoming connection with a 554 error.
216 Unfortunately, if there is a major crisis, such as a disk failure,
217 tcp-wrappers gives "deny", whereas what one would like would be some
218 kind of temporary error. A kludge has been added to help with this.
219 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
220 554 error is used if errno is still zero or contains ENOENT (which occurs
221 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
224 PH/03 Add -lutil to the default FreeBSD LIBS setting.
226 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
227 errors. Otherwise a message that provokes a temporary error (when other
228 messages do not) can cause a whole host to time out.
230 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
231 addresses were routed directly to files or pipes from a redirect router.
232 File deliveries just didn't batch; pipe deliveries might have suffered
235 PH/06 A failure to get a lock for a hints database would erroneously always say
236 "Failed to get write lock", even when it was really a read lock.
238 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
239 of 0600. This has been changed to use the value of the lockfile_mode
240 option (which defaults to 0600).
242 PH/08 Applied small patch from the Sieve maintainer.
244 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
245 folder from quota calculations, a direct delivery into this folder messed
246 up the contents of the maildirsize file. This was because the regex was
247 used only to exclude .Trash (or whatever) when the size of the mailbox
248 was calculated. There was no check that a delivery was happening into an
249 excluded directory. This bug has been fixed by ignoring all quota
250 processing for deliveries into excluded directories.
252 PH/10 Added the maildirfolder_create_regex option to appendfile.
258 PH/01 The code for finding all the local interface addresses on a FreeBSD
259 system running IPv6 was broken. This may well have applied to all BSD
260 systems, as well as to others that have similar system calls. The broken
261 code found IPv4 interfaces correctly, but gave incorrect values for the
262 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
263 that it would not match correctly against @[] and not recognize the IPv6
266 PH/02 The ipliteral router was not recognizing addresses of the form user@
267 [ipv6:....] because it didn't know about the "ipv6:" prefix.
269 PH/03 Added disable_ipv6.
271 PH/04 Changed $reply_address to use the raw form of the headers instead of the
272 decoded form, because it is most often used to construct To: headers
273 lines in autoreplies, and the decoded form may well be syntactically
274 invalid. However, $reply_address has leading white space removed, and all
275 newlines turned into spaces so that the autoreply transport does not
278 PH/05 If group was specified without a user on a router, and no group or user
279 was specified on a transport, the group from the router was ignored.
281 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
282 for visible compile-time settings that can be used to change these
283 numbers, for those that want even more. Backwards compatibility with old
284 spool files has been maintained. However, going back to a previous Exim
285 release will lost any variables that are in spool files.
287 PH/07 Two small changes when running in the test harness: increase delay when
288 passing a TCP/IP connection to a new process, in case the original
289 process has to generate a bounce, and remove special handling of
290 127.0.0.2 (sic), which is no longer necessary.
292 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
293 be the same on different OS.
295 PH/09 Moved a debug statement in filter processing to avoid a race problem when
298 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
299 whether --show-vars was specified or not
301 JJ/02 exipick: Added support for new ACL variable spool format introduced
304 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
305 syntactically invalid From: or Reply-to: line, and a filter used this to
306 generate an autoreply, and therefore failed to obtain an address for the
307 autoreply, Exim could try to deliver to a non-existent relative file
308 name, causing unrelated and misleading errors. What now happens is that
309 it logs this as a hard delivery error, but does not attempt to create a
312 PH/11 The exinext utility has a -C option for testing purposes, but although
313 the given file was scanned by exinext itself; it wasn't being passed on
314 when Exim was called.
316 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
317 an end-of-file indication when reading a command response.
319 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
320 compiled. In many other places in Exim, IPv6 addresses are always
321 recognized, so I have changed this. It also means that IPv4 domain
322 literals of the form [IPV4:n.n.n.n] are now always recognized.
324 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
325 used if the router is not running as root, for example, when verifying at
326 ACL time, or when using -bh. The debugging output from this situation was
327 non-existent - all you got was a failure to exec. I have made two
330 (a) Failures to set uid/gid, the current directory, or a process leader
331 in a subprocess such as that created by queryprogram now generate
332 suitable debugging ouput when -d is set.
334 (b) The queryprogram router detects when it is not running as root,
335 outputs suitable debugging information if -d is set, and then runs
336 the subprocess without attempting to change uid/gid.
338 PH/15 Minor change to Makefile for building test_host (undocumented testing
341 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
342 additional section of a DNS packet that returns MX or SRV records.
343 Instead, it always explicitly searches for A/AAAA records. This avoids
344 major problems that occur when a DNS server includes only records of one
345 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
346 fixed another bug: if SRV records were looked up and the corresponding
347 address records were *not* found in the additional section, the port
348 values from the SRV records were lost.
350 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
351 using the correct key (the original address) when searching the retry
352 rules in order to find which one to use for generating the retry hint.
354 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
355 adding the default one. Similarly, if it contains a Reply-To: header, the
356 errors_reply_to option, if set, is not used.
358 PH/19 When calculating a retry time, Exim used to measure the "time since
359 failure" by looking at the "first failed" field in the retry record. Now
360 it does not use this if it is later than than the arrival time of the
361 message. Instead it uses the arrival time. This makes for better
362 behaviour in cases where some deliveries succeed, thus re-setting the
363 "first failed" field. An example is a quota failure for a huge message
364 when small messages continue to be delivered. Without this change, the
365 "time since failure" will always be short, possible causing more frequent
366 delivery attempts for the huge message than are intended.
367 [Note: This change was subsequently modified - see PH/04 for 4.62.]
369 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
370 $1, $2, $3) because the numerical variables can be reset during some
371 expansion items (e.g. "match"), thereby losing the authentication data.
373 PH/21 Make -bV show the size of off_t variables so that the test suite can
374 decide whether to run tests for quotas > 2G.
376 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
377 mailbox_size, and mailbox_filecount in the appendfile transport. If a
378 filecount value is greater than 2G or if a quota value is greater than 2G
379 on a system where the size of off_t is not greater than 4, a panic error
382 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
383 never match. The debug and -bh output now contains an explicit error
384 message indicating a malformed IPv4 address or mask.
386 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
387 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
390 PH/25 Do not write to syslog when running in the test harness. The only
391 occasion when this arises is a failure to open the main or panic logs
392 (for which there is an explicit test).
394 PH/26 Added the /no_tell option to "control=freeze".
396 PH/27 If a host name lookup failed very early in a connection, for example, if
397 the IP address matched host_lookup and the reverse lookup yielded a name
398 that did not have a forward lookup, an error message of the form "no IP
399 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
400 could be logged. Now it outputs the IP address instead of "NULL".
402 PH/28 An enabling patch from MH: add new function child_open_exim2() which
403 allows the sender and the authenticated sender to be set when
404 submitting a message from within Exim. Since child_open_exim() is
405 documented for local_scan(), the new function should be too.
407 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
408 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
409 results in an empty string is now treated as unset.
411 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
413 PH/31 Added sender_verify_fail logging option.
415 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
416 needed by RFC 822 but not by RFC 2822 was commented out. I have now
417 tidied the source and removed it altogether.
419 PH/33 When a queue run was abandoned because the load average was too high, a
420 log line was always written; now it is written only if the queue_run log
421 selector is set. In addition, the log line for abandonment now contains
422 information about the queue run such as the pid. This is always present
423 in "start" and "stop" lines but was omitted from the "abandon" line.
425 PH/34 Omit spaces between a header name and the colon in the error message that
426 is given when verify = headers_syntax fails (if there are lots of them,
427 the message gets confusing).
429 PH/35 Change the default for dns_check_names_pattern to allow slashes within
430 names, as there are now some PTR records that contain slashes. This check
431 is only to protect against broken name servers that fall over on strange
432 characters, so the fact that it applies to all lookups doesn't matter.
434 PH/36 Now that the new test suite is complete, we can remove some of the
435 special code in Exim that was needed for the old test suite. For example,
436 sorting DNS records because real resolvers return them in an arbitrary
437 order. The new test suite's fake resolver always returns records in the
440 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
441 bounces) except when queue_only is set, to avoid logging races between
442 the different processes.
444 PH/38 Panic-die if .include specifies a non-absolute path.
446 PH/39 A tweak to the "H" retry rule from its user.
448 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
449 a label. They prevented compilation on older perls.
451 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
452 a warning to be raised on newish perls.
454 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
455 on queue. Changes to match documented behaviour of showing count of
456 messages matching specified criteria.
458 PH/40 Changed the default ident timeout from 30s to 5s.
460 PH/41 Added support for the use of login_cap features, on those BSD systems
461 that have them, for controlling the resources used by pipe deliveries.
463 PH/42 The content-scanning code uses fopen() to create files in which to put
464 message data. Previously it was not paying any attention to the mode of
465 the files. Exim runs with umask(0) because the rest of the code creates
466 files with open(), and sets the required mode explicitly. Thus, these
467 files were ending up world-writeable. This was not a big issue, because,
468 being within the spool directory, they were not world-accessible. I have
469 created a function called modefopen, which takes an additional mode
470 argument. It sets umask(777), creates the file, chmods it to the required
471 mode, then resets the umask. All the relevant calls to fopen() in the
472 content scanning code have been changed to use this function.
474 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
475 to 24 hours. This avoids potential overflow problems when processing G
476 and H retry rules. I suspect nobody ever tinkers with this value.
478 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
480 PH/45 When the plaintext authenticator is running as a client, the server's
481 challenges are checked to ensure they are valid base64 strings. By
482 default, the authentication attempt is cancelled if an invalid string is
483 received. Setting client_ignore_invalid_base64 true ignores these errors.
484 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
485 they are received. Thus, the responses can be made to depend on the
486 challenges. If an invalid string is ignored, an empty string is placed in
489 PH/46 Messages that are created by the autoreply transport now contains a
490 References: header, in accordance with RFCs 2822 and 3834.
492 PH/47 Added authenticated_sender_force to the smtp transport.
494 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
496 PH/49 Installed latest patch from the Sieve maintainer.
498 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
499 was also set, the appendfile transport was unnecessarily scanning a
500 directory of message files (e.g. for maildir delivery) to find the count
501 of files (along with the size), even though it did not need this
502 information. It now does the scan only if it needs to find either the
503 size of the count of files.
505 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
507 PH/52 Two bugs concerned with error handling when the smtp transport is
510 (i) Exim was not creating retry information for temporary errors given
511 for individual recipients after the DATA command when the smtp transport
512 was used in LMTP mode. This meant that they could be retried too
513 frequently, and not timed out correctly.
515 (ii) Exim was setting the flag that allows error details to be returned
516 for LMTP errors on RCPT commands, but not for LMTP errors for individual
517 recipients that were returned after the DATA command.
519 PH/53 This is related to PH/52, but is more general: for any failing address,
520 when detailed error information was permitted to be returned to the
521 sender, but the error was temporary, then after the final timeout, only
522 "retry timeout exceeded" was returned. Now it returns the full error as
523 well as "retry timeout exceeded".
525 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
526 do this, and (what is worse) MTAs that accept it.
528 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
529 will now be deprecated.
531 PH/56 New os.c-cygwin from the Cygwin maintainer.
533 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
534 formats (previously only available in exim formats via -bpr, -bpru,
535 and -bpra. Now also available in native and exiqgrep formats)
537 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
538 with very large, slow to parse queues
540 JJ/08 exipick: added ! as generic prefix to negate any criteria format
542 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
544 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
545 responses to authentication challenges, though it was showing the
546 challenges; (ii) I've removed the CR characters from the debug output for
549 PH/58 Allow for the insertion of a newline as well as a space when a string
550 is turned into more than one encoded-word during RFC 2047 encoding. The
551 Sieve code now uses this.
553 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
554 data_4xx, lost_connection, tls_required.
556 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
557 message was being sent as an SMTP response.
559 PH/61 Add -l and -k options to exicyclog.
561 PH/62 When verifying, if an address was redirected to one new address, so that
562 verification continued, and the new address failed or deferred after
563 having set something in $address_data, the value of $address_data was not
564 passed back to the ACL. This was different to the case when no
565 redirection occurred. The value is now passed back in both cases.
567 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
568 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
569 use login_cap.h, so on its own it isn't the distinguishing feature. The
570 new name refers directly to the setclassresources() function.
572 PH/65 Added configuration files for NetBSD3.
574 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
576 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
577 is preferred over IPv4.
579 PH/68 The bounce_return_message and bounce_return_body options were not being
580 honoured for bounces generated during the reception of non-SMTP messages.
581 In particular, this applied to messages rejected by the ACL. This bug has
582 been fixed. However, if bounce_return_message is true and bounce_return_
583 body is false, the headers that are returned for a non-SMTP message
584 include only those that have been read before the error was detected.
585 (In the case of an ACL rejection, they have all been read.)
587 PH/69 The HTML version of the specification is now built in a directory called
588 spec_html instead of spec.html, because the latter looks like a path with
589 a MIME-type, and this confuses some software.
591 PH/70 Catch two compiler warnings in sieve.c.
593 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
594 function verify_get_ident() calls ip_connect() to connect a socket, but
595 if the "connect()" function timed out, ip_connect() used to close the
596 socket. However, verify_get_ident() also closes the socket later, and in
597 between Exim writes to the log, which may get opened at this point. When
598 the socket was closed in ip_connect(), the log could get the same file
599 descriptor number as the socket. This naturally causes chaos. The fix is
600 not to close the socket in ip_connect(); the socket should be closed by
601 the function that creates it. There was only one place in the code where
602 this was missing, in the iplookup router, which I don't think anybody now
603 uses, but I've fixed it anyway.
605 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
606 well as to direct DNS lookups. Otherwise the handling of names in host
607 lists is inconsistent and therefore confusing.
613 PH/01 Two changes to the default runtime configuration:
615 (1) Move the checks for relay_from_hosts and authenticated clients from
616 after to before the (commented out) DNS black list checks.
618 (2) Add control=submission to the relay_from_hosts and authenticated
619 clients checks, on the grounds that messages accepted by these
620 statements are most likely to be submissions.
622 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
624 (1) Generate an error if the third argument for the ${prvs expansion is
627 (2) Treat a missing third argument of ${prvscheck as if it were an empty
630 (3) Reset the variables that are obtained from the first argument of
631 ${prvscheck and used in the second argument before leaving the code,
632 because their memory is reclaimed, so using them afterwards may do
635 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
636 one (it's much easier than Tom thought :-).
638 (5) Because of (4), we can now allow for the use of $prvscheck_result
639 inside the third argument.
641 PH/03 For some reason, the default setting of PATH when running a command from
642 a pipe transport was just "/usr/bin". I have changed it to
645 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
646 anything to be listed in the output from -bV.
648 PH/05 When a filter generated an autoreply, the entire To: header line was
649 quoted in the delivery log line, like this:
651 => >A.N.Other <ano@some.domain> <original@ddress> ...
653 This has been changed so that it extracts the operative address. There
654 may be more than one such address. If so, they are comma-separated, like
657 => >ano@some.domain,ona@other.domain <original@ddress> ...
659 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
660 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
661 not being looked up in the rDNS to get a host name, Exim was showing the
662 IP address twice in Received: lines, even though the IP addresses were
663 identical. For example:
665 Received: from [1.2.3.4] (helo=[1.2.3.4])
667 However, if the real host name was known, it was omitting the HELO data
668 if it matched the actual IP address. This has been tidied up so that it
669 doesn't show the same IP address twice.
671 PH/07 When both +timestamp and +memory debugging was on, the value given by
672 $tod_xxx expansions could be wrong, because the tod_stamp() function was
673 called by the debug printing, thereby overwriting the timestamp buffer.
674 Debugging no longer uses the tod_stamp() function when +timestamp is set.
676 PH/08 When the original message was included in an autoreply transport, it
677 always said "this is a copy of the message, including all the headers",
678 even if body_only or headers_only was set. It now gives an appropriate
681 PH/09 Applied a patch from the Sieve maintainer which:
683 o fixes some comments
684 o adds the (disabled) notify extension core
685 o adds some debug output for the result of if/elsif tests
686 o points to the current vacation draft in the documentation
687 and documents the missing references header update
691 o fixes a bug in processing the envelope test (when testing
692 multiple envelope elements, the last element determinted the
695 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
696 Electronic Mail") by including:
698 Auto-submitted: auto-generated
700 in the messages that it generates (bounce messages and others, such as
701 warnings). In the case of bounce messages for non-SMTP mesages, there was
702 also a typo: it was using "Auto_submitted" (underscore instead of
703 hyphen). Since every message generated by Exim is necessarily in response
704 to another message, thes have all been changed to:
706 Auto-Submitted: auto-replied
708 in accordance with these statements in the RFC:
710 The auto-replied keyword:
712 - SHOULD be used on messages sent in direct response to another
713 message by an automatic process,
715 - MUST NOT be used on manually-generated messages,
717 - MAY be used on Delivery Status Notifications (DSNs) and Message
718 Disposition Notifications (MDNs),
720 - MUST NOT be used on messages generated by automatic or periodic
721 processes, except for messages which are automatic responses to
724 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
725 to the default Received: header definition.
727 PH/12 Added log selector acl_warn_skipped (default on).
729 PH/13 After a successful wildlsearch lookup, discard the values of numeric
730 variables because (a) they are in the wrong storage pool and (b) even if
731 they were copied, it wouldn't work properly because of the caching.
733 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
734 checking when decoding. Apparently there are clients that generate
735 overlong encoded strings. Why am I not surprised?
737 PH/15 If the first argument of "${if match_address" was not empty, but did not
738 contain an "@" character, Exim crashed. Now it writes a panic log message
739 and treats the condition as false.
741 PH/16 In autoreply, treat an empty string for "once" the same as unset.
743 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
744 extension "envelope-auth". The code is finished and in agreement with
745 other implementations, but there is no documentation so far and in fact,
746 nobody wrote the draft yet. This extension is currently #undef'ed, thus
747 not changing the active code.
749 Print executed "if" and "elsif" statements when debugging is used. This
750 helps a great deal to understand what a filter does.
752 Document more things not specified clearly in RFC3028. I had all this
753 sorted out, when out of a sudden new issues came to my mind. Oops."
755 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
758 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
759 canonical form (as documented). However, after a host name lookup from
760 the IP address, check_host() was doing a simple string comparison with
761 addresses acquired from the DNS when checking that the found name did
762 have the original IP as one of its addresses. Since any found IPv6
763 addresses are likely to be in abbreviated form, the comparison could
764 fail. Luckily, there already exists a function for doing the comparison
765 by converting both addresses to binary, so now that is used instead of
768 PH/20 There was another similar case to PH/19, when a complete host name was
769 given in a host list; looking up its IP address could give an abbreviated
770 form, whereas the current host's name might or might not be abbreviated.
771 The same fix has been applied.
777 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
778 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
781 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
782 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
784 PH/03 Typo: missing ".o" in src/pcre/Makefile.
786 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
787 header line, restrict the check to what is listed in RFCs 2369 and 2929.
788 Also, for "Auto-Submitted", treat anything other than "no" as
789 non-personal, in accordance with RFC 3834. (Previously it treated
790 anything starting "auto-" as non-personal.)
792 TF/01 The control=submission/name=... option had a problem with syntax
793 errors if the name included a slash character. The /name= option
794 now slurps the rest of the string, so it can include any characters
795 but it must come last in the list of options (after /sender_retain
798 PH/05 Some modifications to the interface to the fake nameserver for the new
806 TK/01 Added the "success_on_redirect" address verification option. See
807 NewStuff for rationale and an example.
809 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
811 PH/02 Patch to exigrep to allow it to work on syslog lines.
813 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
814 fread() to skip over the body file's header line, because in Cygwin the
815 header line is locked and is inaccessible.
817 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
818 co-exist for some time) to make it clear that it is the Exim ID that is
819 referenced, not the Message-ID: header line.
821 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
822 string_format() function, because snprintf() does not exist on all
825 PH/06 The use of forbid_filter_existstest now also locks out the use of the
826 ${stat: expansion item.
828 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
829 protocol synchronization error", to keep the pedants happy.
831 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
832 well as for IRIX systems, when gcc is being used. See the host.c source
835 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
837 PH/10 Named domain lists were not working if used in a queue_smtp_domains
840 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
841 transport and to the smtp transport in LMTP mode.
843 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
845 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
846 run a filter in a subprocess. This could lead to confusion in subsequent
847 lookups in the parent process. There should also be a search_tidyup() at
848 the end of the subprocess.
850 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
851 only if the host matched helo_try_verify_hosts, which caused the
852 verification to occur when the EHLO/HELO command was issued. The ACL just
853 tested the remembered result. Now, if a previous verification attempt has
854 not happened, "verify = helo" does it there and then.
856 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
858 TK/03 Fix log output including CR from clamd.
860 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
861 exist provoked a memory error which could cause a segfault.
863 PH/15 Installed PCRE 6.2
865 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
867 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
868 of the problem. Specifically, suggested +O2 rather than +O1 for the
871 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
873 PH/20 If a delivery was routed to a non-standard port by means of an SRV
874 record, the port was not correctly logged when the outgoing_port log
875 selector was set (it logged the transort's default port).
877 PH/21 Added support for host-specific ports to manualroute, queryprogram,
878 fallback_hosts, and "hosts" in the smtp transport.
880 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
881 host errors such as "Connection refused".
883 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
884 authentication with radiusclient 0.4.9:
886 - Error returned from rc_read_config was caught wrongly
887 - Username/password not passed on to radius server due to wrong length.
889 The presumption is that some radiusclient API changes for 4.51/PH/17
890 were not taken care of correctly. The code is still untested by me (my
891 Linux distribution still has 0.3.2 of radiusclient), but it was
892 contributed by a Radius user.
894 PH/24 When doing a callout, the value of $domain wasn't set correctly when
895 expanding the "port" option of the smtp transport.
897 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
898 while reading a MIME header. Thanks to Tom Hughes for a patch.
900 PH/24 Include config.h inside local_scan.h so that configuration settings are
903 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
904 that in an ACL for RCPT (for example), you can examine exactly what was
907 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
908 commands, but it was not correctly comparing the address with the actual
909 client host address. Thus, it would show the EHLO address in Received:
910 header lines when this was not necessary.
912 PH/27 Added the % operator to ${eval:}.
914 PH/28 Exim tries to create and chdir to its spool directory when it starts;
915 it should be ignoring failures (because with -C, for example, it has lost
916 privilege). It wasn't ignoring creation failures other than "already
919 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
922 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
923 because an input line was too long, either on its own, or by virtue of
924 too many continuations, the temporary file was not being removed, and the
925 return code was incorrect.
927 PH/31 Missing "BOOL" in function definition in filtertest.c.
929 PH/32 Applied Sieve patches from the maintainer.
931 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
933 PH/33 Added "verify = not_blind".
935 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
936 Local/Makefile (with some defaults set). These are used in built scripts
937 such as exicyclog, but they have never been used in the exim_install
938 script (though there are many overriding facilities there). I have
939 arranged that the exim_install script now takes note of these two
942 PH/35 Installed configuration files for Dragonfly.
944 PH/36 When a locally submitted message by a trusted user did not contain a
945 From: header, and the sender address was obtained from -f or from an SMTP
946 MAIL command, and the trusted user did not use -F to supply a sender
947 name, $originator_name was incorrectly used when constructing a From:
948 header. Furthermore, $originator_name was used for submission mode
949 messages from external hosts without From: headers in a similar way,
950 which is clearly wrong.
952 PH/37 Added control=suppress_local_fixups.
954 PH/38 When log_selector = +received_sender was set, and the addition of the
955 sender made the log line's construction buffer exactly full, or one byte
956 less than full, an overflow happened when the terminating "\n" was
959 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
960 when the result of a list match is failure because a DNS lookup failed.
962 PH/40 RM_COMMAND is now used in the building process.
964 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
965 the "build-* directories that it finds.
967 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
968 address in a domain literal was a prefix of an interface address.
970 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
971 when verifying a sender address, unless rewrite_headers is false.
973 PH/44 Wrote a long comment about why errors_to addresses are verified as
974 recipients, not senders.
976 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
977 the ratelimit ACL was added.
979 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
981 PH/46 Added extra information about PostgreSQL errors to the error string.
983 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
984 avoiding the need to install special zones in a real server. This is
985 backwards compatible; if it can't find the fake resolver, it drops back.
986 Thus, both old and new test suites can be run.
988 TF/02 Added util/ratelimit.pl
990 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
991 clock is set back in time.
993 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
994 Candler <B.Candler@pobox.com>.
996 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
997 OK for addresses that are the result of redirections.
999 PH/48 A number of further additions for the benefit of the new test suite,
1000 including a fake gethostbyname() that interfaces to the fake DNS resolver
1003 TF/06 The fix for widen_domains has also been applied to qualify_single and
1004 search_parents which are the other dnslookup options that can cause
1007 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1010 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1012 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1013 service" if it encountered an MX record with an empty target hostname.
1014 The message is now "an MX or SRV record indicated no SMTP service".
1016 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1017 if the DNS of the sending site is misconfigured. This is quite a
1018 common situation. This change restores the behaviour of treating a
1019 helo verification defer as a failure.
1021 PH/51 If self=fail was set on a router, the bounce message did not include the
1022 actual error message.
1028 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1030 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1031 command itself ended in error, the underlying message about the transport
1032 filter timeout was being overwritten with the pipe command error. Now the
1033 underlying error message should be appended to the second error message.
1035 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1037 PH/02 Reduce the amount of output that "make" produces by default. Full output
1038 can still be requested.
1040 PH/03 The warning log line about a condition test deferring for a "warn" verb
1041 was being output only once per connection, rather than after each
1042 occurrence (because it was using the same function as for successful
1043 "warn" verbs). This seems wrong, so I have changed it.
1045 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1046 it should not have, which might have caused a crash in the right
1047 circumstances, but probably never did.
1049 PH/04 Installed a modified version of Tony Finch's patch to make submission
1050 mode fix the return path as well as the Sender: header line, and to
1051 add a /name= option so that you can make the user's friendly name appear
1054 TF/03 Added the control = fakedefer ACL modifier.
1056 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1057 Mark Lowes for thorough testing.
1059 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1061 TK/03 Merged latest SRS patch from Miles Wilton.
1063 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1064 with the definition in sysexits.h (which is #included earlier).
1065 Fortunately, Exim does not actually use EX_OK. The code used to try to
1066 preserve the sysexits.h value, by assumimg that macro definitions were
1067 scanned for macro replacements. I have been disabused of this notion,
1068 so now the code just undefines EX_OK before #including unistd.h.
1070 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1071 in the smtp transport. When a block could not be written in a single
1072 write() function, the timeout was being re-applied to each part-write.
1073 This seems wrong - if the receiver was accepting one byte at a time it
1074 would take for ever. The timeout is now adjusted when this happens. It
1075 doesn't have to be particularly precise.
1077 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1078 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1080 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1081 without a domain if the check to <postmaster@domain> fails.
1083 SC/01 Eximstats: added -xls and the ability to specify output files
1084 (patch written by Frank Heydlauf).
1086 SC/02 Eximstats: use FileHandles for outputing results.
1088 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1090 SC/04 Eximstats: fixed display of large numbers with -nvr option
1092 SC/05 Eximstats: fixed merging of reports with empty tables.
1094 SC/06 Eximstats: added the -include_original_destination flag
1096 SC/07 Eximstats: removed tabs and trailing whitespace.
1098 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1100 TK/06 MBOX spool code: Add real "From " MBOX separator line
1101 so the .eml file is really in mbox format (even though
1102 most programs do not really care). Patch from Alex Miller.
1104 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1105 The latter is generated from $received_to and is only set if the
1106 message has one envelope recipient. SA can use these headers,
1107 obviously out-of-the-box. Patch from Alex Miller.
1109 PH/08 The ${def test on a variable was returning false if the variable's
1110 value was "0", contrary to what the specification has always said!
1111 The result should be true unless the variable is empty.
1113 PH/09 The syntax error of a character other than { following "${if
1114 def:variable_name" (after optional whitespace) was not being diagnosed.
1115 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1116 accidental colon was present, for example, could give incorrect results.
1118 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1119 result is used (not including appendfile, where other changes are about
1122 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1123 This involved changing a lot of size variables from int to off_t. It
1124 should work with maildirs and everything.
1126 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1127 spamd dying while we are connected to it.
1129 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1132 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1133 function for BDB 4.3.
1135 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1137 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1138 the sources for PCRE have been re-arranged and more files are now
1141 PH/15 The code I had for printing potentially long long variables in PH/11
1142 above was not the best (it lost precision). The length of off_t variables
1143 is now inspected at build time, and an appropriate printing format (%ld
1144 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1145 to be "long long int" or "long int". This is needed for the internal
1146 formatting function string_vformat().
1148 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1149 the configuration file to be ":syslog", then the script "guesses" where
1150 the logs files are, rather than using the compiled in default. In our
1151 case the guess is not the same as the compiled default, so the script
1152 suddenly stopped working when I started to use syslog. The patch checks
1153 to see if log_file_path is "". If so, it attempts to read it from exim
1154 with no configuration file to get the compiled in version, before it
1155 falls back to the previous guessing code."
1157 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1158 implementing BATV in an Exim configuration. See NewStuff for the gory
1161 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1162 Makefile that are specific to HP-UX.
1164 PH/18 If the "use_postmaster" option was set for a recipient callout together
1165 with the "random" option, the postmaster address was used as the MAIL
1166 FROM address for the random test, but not for the subsequent recipient
1167 test. It is now used for both.
1169 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1170 patch removes a few documentation additions to RFC 3028, because the
1171 latest draft now contains them. It adds the new en;ascii-case comparator
1172 and a new error check for 8bit text in MIME parts. Comparator and
1173 require names are now matched exactly. I enabled the subaddress
1174 extension, but it is not well tested yet (read: it works for me)."
1176 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1177 rework some of the code of TK/09 above to avoid the hardwired use of
1178 "%lld" and "long long". Replaced the call to snprintf() with a call to
1181 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1182 records point to non-existent hosts", "retry timeout exceeded", and
1183 "retry time not reached for any host after a long failure period".
1185 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1186 experimental DomainKeys support:
1188 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1189 (2) On an error such as an illegally used "control", the wrong name for
1190 the control was given.
1192 These problems did NOT occur unless DomainKeys support was compiled.
1194 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1196 PH/24 Added ${if match_ip condition.
1198 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1199 databases so that it will be absolutely obvious if a crash occurs in the
1200 DB library. This is a regular occurrence (often caused by mis-matched
1203 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1204 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1205 picked up on a user's system that detects such things. There doesn't seem
1206 to be a gcc warning option for this - only an attribute that has to be
1207 put on the function's prototype. It seems that in Fedora Core 4 they have
1208 set this on a number of new functions. No doubt there will be more in due
1211 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1212 specify a transport. However, if an address that was verified by such a
1213 router was the subject of a callout, Exim crashed because it tried to
1214 read the rcpt_include_affixes from the non-existent transport. Now it
1215 just assumes that the setting of that option is false. This bug was
1216 introduced by 4.51/PH/31.
1218 PH/28 Changed -d+all to exclude +memory, because that information is very
1219 rarely of interest, but it makes the output a lot bigger. People tend to
1220 do -d+all out of habit.
1222 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1223 code in os-type was giving problems when libc.so lives in lib64, like on
1226 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1227 aren't the modern standard, and it seems that some systems' include files
1228 don't always have them. Exim was already checking for some of the newer
1229 ones like T_AAAA, and defining it itself. I've added checks for all the
1230 record types that Exim uses.
1232 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1233 not automatically generating a new one, as it is supposed to. This
1234 prevented TLS from working. If the file did exist, but contained invalid
1235 data, a new version was generated, as expected. It was only the case of a
1236 non-existent file that was broken.
1238 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1239 with a change in libdomainkeys > 0.64.
1241 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1242 from DNS. If the selector record carries the flag, it now has
1243 precedence over the domain-wide flag.
1245 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1247 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1248 the use of a port name that isn't defined in /etc/services) occurred, the
1249 message was deferred as in a normal delivery, and thus remained on the
1250 spool, instead of being failed because of the mua_wrapper setting. This
1251 is now fixed, and I tidied up some of the mua_wrapper messages at the
1254 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1255 the messages in a hash of arrays rather than using individual hashes.
1256 This is a bit cleaner and results in dramatic memory savings, albeit
1257 at a slight CPU cost.
1259 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1260 as requested by Marc Sherman.
1262 SC/10 Eximstats: added histograms for user specified patterns as requested
1265 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1267 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1268 fopen() in the content-scanning modules that did not already have it.
1274 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1275 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1277 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1279 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1281 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1282 to test Sieve filters that use "vacation".
1284 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1285 that changes the way the GnuTLS parameters are stored in the cache file.
1286 The new format can be generated externally. For backward compatibility,
1287 if the data in the cache doesn't make sense, Exim assumes it has read an
1288 old-format file, and it generates new data and writes a new file. This
1289 means that you can't go back to an older release without removing the
1292 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1293 work if there are any delivery delays because "one_time" forces the
1294 parent to be marked "delivered", so its unseen clone is never tried
1295 again. For this reason, Exim now forbids the simultaneous setting of
1298 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1299 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1300 entry if you want to know the details. The fix, however, neglected to
1301 consider the case when local delivery batching is involved. The test for
1302 "previously delivered" was not happening when checking to see if an
1303 address could be batched with a previous (undelivered) one; under
1304 certain circumstances this could lead to multiple deliveries to the same
1307 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1308 in its include files, and this causes problems building Exim.
1310 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1311 header_syntax) but Exim was just ignoring anything given after a slash.
1312 In particular, this caused confusion with an attempt to use "verify =
1313 reverse_host_lookup/defer_ok". An error is now given when options are
1314 supplied for verify items that do not have them. (Maybe reverse_host_
1315 lookup should have a defer_ok option, but that's a different point.)
1317 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1318 defined by RFC 821) to 2048, because there were problems with some AUTH
1319 commands, and RFC 1869 says the size should be increased for extended
1320 SMTP commands that take arguments.
1322 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1325 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1326 "unknown" error; now it says that the functionality isn't in the binary.
1328 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1329 an address' error message when a string expansion fails (syntax or
1330 whatever). Otherwise the password may appear in the log. Following change
1331 PH/42 below, there is no longer a chance of it appearing in a bounce
1334 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1336 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1337 of its addresses was ever tried. (Bugzilla bug #2.)
1339 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1340 the result incorrectly in the debug output. (It correctly added a newline
1341 to what was transported.)
1343 TF/01 Added $received_time.
1345 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1346 commented out examples of how to interface to a virus scanner and to
1347 SpamAssassin. Also added commented examples of av_scanner and
1348 spamd_address settings.
1350 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1351 and controls are allowed in which ACLs. There were a couple of minor
1352 errors. Some of the entries in the conditions table (which is a table of
1353 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1354 negation of where the condition IS allowed.
1356 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1358 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1359 header file does not have a version number, so I've had to invent a new
1360 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1361 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1362 radiusclient), but it was contributed by a Radius user.
1364 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1365 files or directories, for OpenSSL.
1367 PH/19 When an Exim process that is running as root has to create an Exim log
1368 file, it does so in a subprocess that runs as exim:exim so as to get the
1369 ownership right at creation (otherwise, other Exim processes might see
1370 the file with the wrong ownership). There was no test for failure of this
1371 fork() call, which would lead to the process getting stuck as it waited
1372 for a non-existent subprocess. Forks do occasionally fail when resources
1373 run out. I reviewed all the other calls to fork(); they all seem to check
1376 PH/20 When checking for unexpected SMTP input at connect time (before writing
1377 the banner), Exim was not dealing correctly with a non-positive return
1378 from the read() function. If the client had disconnected by this time,
1379 the result was a log entry for a synchronization error with an empty
1380 string after "input=" when read() returned zero. If read() returned -1
1381 (an event I could not check), uninitialized data bytes were printed.
1382 There were reports of junk text (parts of files, etc) appearing after
1385 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1387 PH/22 Added support for macro redefinition, and (re)definition in between
1388 driver and ACL definitions.
1390 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1391 forgetting to use the resulting value; it was using the unexpanded value.
1393 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1394 hadn't been configured. The fix is from Juergen Kreileder, who
1395 understands it better than I do:
1397 "Here's what I see happening with three configured cyrus_sasl
1398 authenticators configured (plain, login, cram-md5):
1400 On startup auth_cyrus_sasl_init() gets called for each of these.
1401 This means three calls to sasl_listmech() without a specified mech_list.
1402 => SASL tests which mechs of all available mechs actually work
1403 => three warnings about OTP not working
1404 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1406 With the patch, sasl_listmech() also gets called three times. But now
1407 SASL's mech_list option is set to the server_mech specified in the the
1408 authenticator. Or in other words, the answer from sasl_listmech()
1409 gets limited to just the mech you're testing for (which is different
1411 => the return list contains just 'plain' or 'login', 'cram-md5' or
1412 nothing depending on the value of ob->server_mech.
1414 I've just tested the patch: Authentication still works fine,
1415 unavailable mechs specified in the exim configuration are still
1416 caught, and the auth.log warnings about OTP are gone."
1418 PH/25 When debugging is enabled, the contents of the command line are added
1419 to the debugging output, even when log_selector=+arguments is not
1422 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1423 answer is "GNU", and only if the return is "GNU/something" is the answer
1426 PH/27 $acl_verify_message is now set immediately after the failure of a
1427 verification in an ACL, and so is available in subsequent modifiers. In
1428 particular, the message can be preserved by coding like this:
1430 warn !verify = sender
1431 set acl_m0 = $acl_verify_message
1433 Previously, $acl_verify_message was set only while expanding "message"
1434 and "log_message" when a very denied access.
1436 PH/28 Modified OS/os.c-Linux with
1438 -#ifndef OS_LOAD_AVERAGE
1439 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1441 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1442 nomenclature these days.)
1444 PH/29 Installed patch from the Sieve maintainer that adds the options
1445 sieve_useraddress and sieve_subaddress to the redirect router.
1447 PH/30 In these circumstances:
1448 . Two addresses routed to the same list of hosts;
1449 . First host does not offer TLS;
1450 . First host accepts first address;
1451 . First host gives temporary error to second address;
1452 . Second host offers TLS and a TLS session is established;
1453 . Second host accepts second address.
1454 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1455 and peerdn, if requested) that were in fact used only for the second
1458 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1459 attention to any local part prefix or suffix that was matched by the
1460 router that accepted the address. It now behaves in the same way as it
1461 does for delivery: the affixes are removed from the local part unless
1462 rcpt_include_affixes is set on the transport.
1464 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1465 timeout during the DATA phase of an incoming message.
1467 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1468 applied a patch sanctioned by the Sieve maintainer.
1470 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1471 the uid or gid is negative. A case of a negative gid caused this to be
1472 noticed. The fix allows for either to be negative.
1474 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1475 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1478 PH/36 The change PH/12 above was broken. Fixed it.
1480 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1481 the grounds that routing the same address twice would always produce the
1482 same answer. This might have been true once, but it is certainly no
1483 longer true now. Routing a child address may depend on the previous
1484 routing that produced that child. Some complicated redirection strategies
1485 went wrong when messages had multiple recipients, and made Exim's
1486 behaviour dependent on the order in which the addresses were given.
1488 I have moved the duplicate checking until after the routing is complete.
1489 Exim scans the addresses that are assigned to local and remote
1490 transports, and removes any duplicates. This means that more work will be
1491 done, as duplicates will always all be routed, but duplicates are
1492 presumably rare, so I don't expect this is of any significance.
1494 For deliveries to pipes, files, and autoreplies, the duplicate checking
1495 still happens during the routing process, since they are not going to be
1498 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1499 It corrects a timeout issue with spamd. This is Ian's comment: "The
1500 background is that sometimes spamd either never reads data from a
1501 connection it has accepted, or it never writes response data. The exiscan
1502 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1503 blindly assumes that writes won't block so it may never time out."
1505 PH/39 Allow G after quota size as well as K and M.
1507 PH/40 The value set for $authenticated_id in an authenticator may not contain
1508 binary zeroes or newlines because the value is written to log lines and
1509 to spool files. There was no check on this. Now the value is run through
1510 the string_printing() function so that such characters are converted to
1511 printable escape sequences.
1513 PH/41 $message_linecount is a new variable that contains the total number of
1514 lines in the message. Compare $body_linecount, which is the count for the
1517 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1518 bounce and delay warning messages, except in certain special cases, which
1521 (a) An SMTP error message from a remote host;
1522 (b) A message specified in a :fail: redirection;
1523 (c) A message specified in a "fail" command in a system filter;
1524 (d) A message specified in a FAIL return from the queryprogram router;
1525 (e) A message specified by the cannot_route_message router option.
1527 In these cases only, Exim does include the error details in bounce and
1528 warning messages. There are also a few cases where bland messages such
1529 as "unrouteable address" or "local delivery error" are given.
1531 PH/43 $value is now also set for the "else" part of a ${run expansion.
1533 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1534 being worked on, but at least Exim now implements the latest version to
1537 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1538 process to complete was treated as a delivery failure, a timeout while
1539 writing the message to the pipe was logged, but erroneously treated as a
1540 successful delivery. Such timeouts include transport filter timeouts. For
1541 consistency with the overall process timeout, these timeouts are now
1542 treated as errors, giving rise to delivery failures by default. However,
1543 there is now a new Boolean option for the pipe transport called
1544 timeout_defer, which, if set TRUE, converts the failures into defers for
1545 both kinds of timeout. A transport filter timeout is now identified in
1548 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1549 systems where "make" and "gmake" are different, calling "gmake" at top
1550 level broke things. I've arranged for the value of $(MAKE) to be passed
1551 from the Makefile to this script so that it can call the same version of
1555 A note about Exim versions 4.44 and 4.50
1556 ----------------------------------------
1558 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1559 changes of various kinds. As a consequence, a big documentation update was
1560 needed. This delayed the release for rather longer than seemed good, especially
1561 in the light of a couple of (minor) security issues. Therefore, the changes
1562 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1563 release. So 4.44 and 4.50 are in effect two different branches that both start
1566 I have left the 4.50 change log unchanged; it contains all the changes since
1567 4.43. The change log for 4.44 is below; many of its items are identical to
1568 those for 4.50. This seems to be the most sensible way to preserve the
1569 historical information.
1575 1. Minor wording change to the doc/README.SIEVE file.
1577 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1578 computation of the current number of files was incorrect.
1580 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1581 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1584 4. Give more explanation in the error message when the command for a transport
1585 filter fails to execute.
1587 5. There are several places where Exim runs a non-Exim command in a
1588 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1589 was being done only for the command run by the queryprogram router. It is
1590 now done for all such subprocesses. The other cases are: ${run, transport
1591 filters, and the commands run by the lmtp and pipe transports.
1593 6. Added CONFIGURE_GROUP build-time option.
1595 7. Some older OS have a limit of 256 on the maximum number of file
1596 descriptors. Exim was using setrlimit() to set 1000 as a large value
1597 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1598 systems. I've change it so that if it can't get 1000, it tries for 256.
1600 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1601 was an oversight, and furthermore, ever since the addition of extra
1602 controls (e.g. 4.43/32), the checks on when to allow different forms of
1603 "control" were broken. There should now be diagnostics for all cases when a
1604 control that does not make sense is encountered.
1606 9. Added the /retain_sender option to "control=submission".
1608 10. $recipients is now available in the predata ACL (oversight).
1610 11. Tidy the search cache before the fork to do a delivery from a message
1611 received from the command line. Otherwise the child will trigger a lookup
1612 failure and thereby defer the delivery if it tries to use (for example) a
1613 cached ldap connection that the parent has called unbind on.
1615 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1616 of $address_data from the recipient verification was clobbered by the
1617 sender verification.
1619 13. The value of address_data from a sender verification is now available in
1620 $sender_address_data in subsequent conditions in the ACL statement.
1622 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1624 15. Added a new option "connect=<time>" to callout options, to set a different
1627 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1628 was its contents. (It was OK if the option was not defined at all.)
1630 17. A "Completed" log line is now written for messages that are removed from
1631 the spool by the -Mrm option.
1633 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1634 information about exactly what failed.
1636 19. Added -dd to debug only the daemon process.
1638 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1639 handles timeouts, both on the server side and network timeouts. Renamed the
1640 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1642 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1643 It is now set to "smtps".
1645 22. $host_address is now set to the target address during the checking of
1646 ignore_target_hosts.
1648 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1649 being passed; this would have caused $sender_host_name to have been used if
1650 matching the list had actually called for a host name (not very likely,
1651 since this list is usually IP addresses). A host name is now passed as
1654 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1655 code that specifies a non-restarting handler (typically sigaction() in
1656 modern systems) in an attempt to fix a rare and obscure crash bug.
1658 25. Narrowed the window for a race in the daemon that could cause it to ignore
1659 SIGCHLD signals. This is not a major problem, because they are used only to
1660 wake it up if nothing else does.
1662 26. A malformed maildirsize file could cause Exim to calculate negative values
1663 for the mailbox size or file count. Odd effects could occur as a result.
1664 The maildirsize information is now recalculated if the size or filecount
1667 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1668 support for a long time. Removed HAVE_SYS_VFS_H.
1670 28. Installed the latest version of exipick from John Jetmore.
1672 29. In an address list, if the pattern was not a regular expression, an empty
1673 subject address (from a bounce message) matched only if the pattern was an
1674 empty string. Non-empty patterns were not even tested. This was the wrong
1675 because it is perfectly reasonable to use an empty address as part of a
1676 database query. An empty address is now tested by patterns that are
1677 lookups. However, all the other forms of pattern expect the subject to
1678 contain a local part and a domain, and therefore, for them, an empty
1679 address still always fails if the pattern is not itself empty.
1681 30. Exim went into a mad DNS loop when attempting to do a callout where the
1682 host was specified on an smtp transport, and looking it up yielded more
1683 than one IP address.
1685 31. Re-factored the code for checking spool and log partition space into a
1686 function that finds that data and another that does the check. The former
1687 is then used to implement four new variables: $spool_space, $log_space,
1688 $spool_inodes, and $log_inodes.
1690 32. The RFC2047 encoding function was originally intended for short strings
1691 such as real names; it was not keeping to the 75-character limit for
1692 encoded words that the RFC imposes. It now respects the limit, and
1693 generates multiple encoded words if necessary. To be on the safe side, I
1694 have increased the buffer size for the ${rfc2047: expansion operator from
1697 33. It is now permitted to omit both strings after an "if" condition; if the
1698 condition is true, the result is "true". As before, when the second string
1699 is omitted, a false condition yields an empty string. This makes it less
1700 cumbersome to write custom ACL and router conditions.
1702 34. Failure to deliver a bounce message always caused it to be frozen, even if
1703 there was an errors_to setting on the router. The errors_to setting is now
1706 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1707 canonical form (fully expanded) before being placed in
1708 $sender_host_address.
1710 36. The table in the code that translates DNS record types into text (T_A to
1711 "A" for instance) was missing entries for NS and CNAME. It is just possible
1712 that this could have caused confusion if both these types were looked up
1713 for the same domain, because the text type is used as part of Exim's
1714 per-process caching. But the chance of anyone hitting this buglet seems
1717 37. The dnsdb lookup has been extended in a number of ways.
1719 (1) There is a new type, "zns", which walks up the domain tree until it
1720 finds some nameserver records. It should be used with care.
1722 (2) There is a new type, "mxh", which is like "mx" except that it returns
1723 just the host names, not the priorities.
1725 (3) It is now possible to give a list of domains (or IP addresses) to be
1726 looked up. The behaviour when one of the lookups defers can be
1727 controlled by a keyword.
1729 (4) It is now possible to specify the separator character for use when
1730 multiple records are returned.
1732 38. The dnslists ACL condition has been extended: it is now possible to supply
1733 a list of IP addresses and/or domains to be looked up in a particular DNS
1736 39. Added log_selector=+queue_time_overall.
1738 40. When running the queue in the test harness, wait just a tad after forking a
1739 delivery process, to get repeatability of debugging output.
1741 41. Include certificate and key file names in error message when GnuTLS fails
1742 to set them up, because the GnuTLS error message doesn't include the name
1743 of the failing file when there is a problem reading it.
1745 42. Allow both -bf and -bF in the same test run.
1747 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1749 44. The "Exiscan patch" is now merged into the mainline Exim source.
1751 45. Sometimes the final signoff response after QUIT could fail to get
1752 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1753 < 0 before doing a fflush(). This bug looks as though it goes back to the
1754 introduction of TLS in release 3.20, but "sometimes" must have been rare
1755 because the tests only now provoked it.
1757 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1758 (this can affect the format of dates).
1760 47. exim_tidydb, when checking for the continued existence of a message for
1761 which it has found a message-specific retry record, was not finding
1762 messages that were in split spool directories. Consequently, it was
1763 deleting retry records that should have stayed in existence.
1765 48. Steve fixed some bugs in eximstats.
1767 49. The SPA authentication driver was not abandoning authentication and moving
1768 on to the next authenticator when an expansion was forced to fail,
1769 contradicting the general specification for all authenticators. Instead it
1770 was generating a temporary error. It now behaves as specified.
1772 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1773 (the order specifies the preference for clients). The order is now AES256,
1774 AES128, 3DES, ARCFOUR128.
1776 51. Small patch to Sieve code - explicitly set From: when generating an
1779 52. Exim crashed if a remote delivery caused a very long error message to be
1780 recorded - for instance if somebody sent an entire SpamAssassin report back
1781 as a large number of 550 error lines. This bug was coincidentally fixed by
1782 increasing the size of one of Exim's internal buffers (big_buffer) that
1783 happened as part of the Exiscan merge. However, to be on the safe side, I
1784 have made the code more robust (and fixed the comments that describe what
1787 53. Now that there can be additional text after "Completed" in log lines (if
1788 the queue_time_overall log selector is set), a one-byte patch to exigrep
1789 was needed to allow it to recognize "Completed" as not the last thing in
1792 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1793 patch that reportedly fixes this has been added. I am not expert enough to
1794 create a test for it. This is what the patch creator wrote:
1796 "I found a little strange behaviour of ldap code when working with
1797 Windows 2003 AD Domain, where users was placed in more than one
1798 Organization Units. When I tried to give exim partial DN, the exit code
1799 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1800 But simultaneously result of request was absolutely normal ldap result,
1801 so I produce this patch..."
1803 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1804 REFERENCE, so I have modified the code to exclude the patch when that macro
1807 55. Some experimental protocols are using DNS PTR records for new purposes. The
1808 keys for these records are domain names, not reversed IP addresses. The
1809 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1810 leaves it alone. Component reversal etc. now happens only for IP addresses.
1813 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1815 57. Double the size of the debug message buffer (to 2048) so that more of very
1816 long debug lines gets shown.
1818 58. The exicyclog utility now does better if the number of log files to keep
1819 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1821 59. Two changes related to the smtp_active_hostname option:
1823 (1) $smtp_active_hostname is now available as a variable.
1824 (2) The default for smtp_banner uses $smtp_active_hostname instead
1825 of $primary_hostname.
1827 60. The host_aton() function is supposed to be passed a string that is known
1828 to be a valid IP address. However, in the case of IPv6 addresses, it was
1829 not checking this. This is a hostage to fortune. Exim now panics and dies
1830 if the condition is not met. A case was found where this could be provoked
1831 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1832 components; fortuitously, this particular loophole had already been fixed
1833 by change 4.50/55 above.
1835 If there are any other similar loopholes, the new check in host_aton()
1836 itself should stop them being exploited. The report I received stated that
1837 data on the command line could provoke the exploit when Exim was running as
1838 exim, but did not say which command line option was involved. All I could
1839 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1840 running as the user.
1843 61. There was a buffer overflow vulnerability in the SPA authentication code
1844 (which came originally from the Samba project). I have added a test to the
1845 spa_base64_to_bits() function which I hope fixes it.
1848 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1849 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1851 63. The daemon start-up calls getloadavg() while still root for those OS that
1852 need the first call to be done as root, but it missed one case: when
1853 deliver_queue_load_max is set with deliver_drop_privilege. This is
1854 necessary for the benefit of the queue runner, because there is no re-exec
1855 when deliver_drop_privilege is set.
1857 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1858 This has been fixed.
1860 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1861 was in use, was not putting the data itself into the right store pool;
1862 consequently, it could be overwritten for a subsequent message in the same
1863 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1866 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1868 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1869 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1870 the function were treating the return as a boolean value, which happened to
1871 work because 0=false and not-0=true, but is not correct code.
1873 68. The host_aton() function was not handling scoped IPv6 addresses (those
1874 with, for example, "%eth0" on the end) correctly.
1876 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1877 negated items (that is, ~something) in unsigned ints. Some compilers
1878 apparently mutter when there is no cast.
1880 70. If an address verification called from an ACL failed, and did not produce a
1881 user-specific message (i.e. there was only a "system" message), nothing was
1882 put in $acl_verify_message. In this situation, it now puts the system
1885 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1886 session; change 4.31/43 added the unwanted input to the log line - except
1887 that it did not do this in the start of session case. It now does.
1889 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1890 This is wrong and can cause the other end to generate a synchronization
1891 error if it is another Exim or anything else that does the synchronization
1892 check. A QUIT command is no longer sent after a timeout.
1894 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1895 during host lookups.
1897 74. The defer_ok option of callout verification was not working if it was used
1898 when verifying addresses in header lines, that is, for this case:
1900 verify = header_sender/callout=defer_ok
1902 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1903 those file descriptors could be used for SMTP connections. If anything
1904 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1905 could be sent to the SMTP client, causing chaos. The daemon now opens
1906 stdin, stdout, and stderr to /dev/null when it puts itself into the
1909 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1910 log by default. The user can override this with suitable Perl magic.
1912 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1913 the log message when discard triggers, was not working for the DATA ACL or
1914 for the non-SMTP ACL.
1916 78. Error message wording change in sieve.c.
1918 79. If smtp_accept_max_per_host was set, the number of connections could be
1919 restricted to fewer than expected, because the daemon was trying to set up
1920 a new connection before checking whether the processes handling previous
1921 connections had finished. The check for completed processes is now done
1922 earlier. On busy systems, this bug wouldn't be noticed because something
1923 else would have woken the daemon, and it would have reaped the completed
1926 80. If a message was submitted locally by a user whose login name contained one
1927 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1928 It caused a spool format error. I have fixed the spool reading code. A
1929 related problem was that the "from" clause in the Received: line became
1930 illegal because of the space(s). It is now covered by ${quote_local_part.
1932 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1935 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1937 83. Patch from Sieve maintainer to fix unterminated string problem in
1938 "vacation" handling.
1940 84. Some minor changes to the Linux configuration files to help with other
1941 OS variants using glibc.
1943 85. One more patch for Sieve to update vacation handling to latest spec.
1946 ----------------------------------------------------
1947 See the note above about the 4.44 and 4.50 releases.
1948 ----------------------------------------------------
1954 1. Change 4.43/35 introduced a bug that caused file counts to be
1955 incorrectly computed when quota_filecount was set in an appendfile
1958 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1959 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1962 3. Give more explanation in the error message when the command for a transport
1963 filter fails to execute.
1965 4. There are several places where Exim runs a non-Exim command in a
1966 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1967 was being done only for the command run by the queryprogram router. It is
1968 now done for all such subprocesses. The other cases are: ${run, transport
1969 filters, and the commands run by the lmtp and pipe transports.
1971 5. Some older OS have a limit of 256 on the maximum number of file
1972 descriptors. Exim was using setrlimit() to set 1000 as a large value
1973 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1974 systems. I've change it so that if it can't get 1000, it tries for 256.
1976 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1977 was an oversight, and furthermore, ever since the addition of extra
1978 controls (e.g. 4.43/32), the checks on when to allow different forms of
1979 "control" were broken. There should now be diagnostics for all cases when a
1980 control that does not make sense is encountered.
1982 7. $recipients is now available in the predata ACL (oversight).
1984 8. Tidy the search cache before the fork to do a delivery from a message
1985 received from the command line. Otherwise the child will trigger a lookup
1986 failure and thereby defer the delivery if it tries to use (for example) a
1987 cached ldap connection that the parent has called unbind on.
1989 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1990 of $address_data from the recipient verification was clobbered by the
1991 sender verification.
1993 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1994 was its contents. (It was OK if the option was not defined at all.)
1996 11. A "Completed" log line is now written for messages that are removed from
1997 the spool by the -Mrm option.
1999 12. $host_address is now set to the target address during the checking of
2000 ignore_target_hosts.
2002 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2003 being passed; this would have caused $sender_host_name to have been used if
2004 matching the list had actually called for a host name (not very likely,
2005 since this list is usually IP addresses). A host name is now passed as
2008 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2009 code that specifies a non-restarting handler (typically sigaction() in
2010 modern systems) in an attempt to fix a rare and obscure crash bug.
2012 15. Narrowed the window for a race in the daemon that could cause it to ignore
2013 SIGCHLD signals. This is not a major problem, because they are used only to
2014 wake it up if nothing else does.
2016 16. A malformed maildirsize file could cause Exim to calculate negative values
2017 for the mailbox size or file count. Odd effects could occur as a result.
2018 The maildirsize information is now recalculated if the size or filecount
2021 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2022 support for a long time. Removed HAVE_SYS_VFS_H.
2024 18. Updated exipick to current release from John Jetmore.
2026 19. Allow an empty sender to be matched against a lookup in an address list.
2027 Previously the only cases considered were a regular expression, or an
2030 20. Exim went into a mad DNS lookup loop when doing a callout where the
2031 host was specified on the transport, if the DNS lookup yielded more than
2034 21. The RFC2047 encoding function was originally intended for short strings
2035 such as real names; it was not keeping to the 75-character limit for
2036 encoded words that the RFC imposes. It now respects the limit, and
2037 generates multiple encoded words if necessary. To be on the safe side, I
2038 have increased the buffer size for the ${rfc2047: expansion operator from
2041 22. Failure to deliver a bounce message always caused it to be frozen, even if
2042 there was an errors_to setting on the router. The errors_to setting is now
2045 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2046 canonical form (fully expanded) before being placed in
2047 $sender_host_address.
2049 24. Updated eximstats to version 1.33
2051 25. Include certificate and key file names in error message when GnuTLS fails
2052 to set them up, because the GnuTLS error message doesn't include the name
2053 of the failing file when there is a problem reading it.
2055 26. Expand error message when OpenSSL has problems setting up cert/key files.
2058 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2059 (this can affect the format of dates).
2061 28. exim_tidydb, when checking for the continued existence of a message for
2062 which it has found a message-specific retry record, was not finding
2063 messages that were in split spool directories. Consequently, it was
2064 deleting retry records that should have stayed in existence.
2066 29. eximstats updated to version 1.35
2067 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2068 1.35 - bugfix such that pie charts by volume are generated correctly
2070 30. The SPA authentication driver was not abandoning authentication and moving
2071 on to the next authenticator when an expansion was forced to fail,
2072 contradicting the general specification for all authenticators. Instead it
2073 was generating a temporary error. It now behaves as specified.
2075 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2076 (the order specifies the preference for clients). The order is now AES256,
2077 AES128, 3DES, ARCFOUR128.
2079 31. Small patch to Sieve code - explicitly set From: when generating an
2082 32. Exim crashed if a remote delivery caused a very long error message to be
2083 recorded - for instance if somebody sent an entire SpamAssassin report back
2084 as a large number of 550 error lines. This bug was coincidentally fixed by
2085 increasing the size of one of Exim's internal buffers (big_buffer) that
2086 happened as part of the Exiscan merge. However, to be on the safe side, I
2087 have made the code more robust (and fixed the comments that describe what
2090 33. Some experimental protocols are using DNS PTR records for new purposes. The
2091 keys for these records are domain names, not reversed IP addresses. The
2092 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2093 leaves it alone. Component reversal etc. now happens only for IP addresses.
2096 34. The host_aton() function is supposed to be passed a string that is known
2097 to be a valid IP address. However, in the case of IPv6 addresses, it was
2098 not checking this. This is a hostage to fortune. Exim now panics and dies
2099 if the condition is not met. A case was found where this could be provoked
2100 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2101 components; fortuitously, this particular loophole had already been fixed
2102 by change 4.50/55 or 4.44/33 above.
2104 If there are any other similar loopholes, the new check in host_aton()
2105 itself should stop them being exploited. The report I received stated that
2106 data on the command line could provoke the exploit when Exim was running as
2107 exim, but did not say which command line option was involved. All I could
2108 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2109 running as the user.
2112 35. There was a buffer overflow vulnerability in the SPA authentication code
2113 (which came originally from the Samba project). I have added a test to the
2114 spa_base64_to_bits() function which I hope fixes it.
2117 36. The daemon start-up calls getloadavg() while still root for those OS that
2118 need the first call to be done as root, but it missed one case: when
2119 deliver_queue_load_max is set with deliver_drop_privilege. This is
2120 necessary for the benefit of the queue runner, because there is no re-exec
2121 when deliver_drop_privilege is set.
2123 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2124 was in use, was not putting the data itself into the right store pool;
2125 consequently, it could be overwritten for a subsequent message in the same
2126 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2129 38. Sometimes the final signoff response after QUIT could fail to get
2130 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2131 < 0 before doing a fflush(). This bug looks as though it goes back to the
2132 introduction of TLS in release 3.20, but "sometimes" must have been rare
2133 because the tests only now provoked it.
2139 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2140 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2141 Now it returns an integer. A number of calls were still expecting a T/F
2142 return. Fortuitously, in all cases, the tests worked in OK situations,
2143 which is the norm. However, things would have gone wrong on any write
2144 failures on the smtp file descriptor. This function is used when sending
2145 messages over SMTP and also when doing verify callouts.
2147 2. When Exim is called to do synchronous delivery of a locally submitted
2148 message (the -odf or -odi options), it no longer closes stderr before doing
2151 3. Implemented the mua_wrapper option.
2153 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2155 5. Implemented the functions header_remove(), header_testname(),
2156 header_add_at_position(), and receive_remove_recipient(), and exported them
2159 6. If an ACL "warn" statement specified the addition of headers, Exim already
2160 inserted X-ACL-Warn: at the start if there was no header name. However, it
2161 was not making this test for the second and subsequent header lines if
2162 there were newlines in the string. This meant that an invalid header could
2163 be inserted if Exim was badly configured.
2165 7. Allow an ACL "warn" statement to add header lines at the start or after all
2166 the Received: headers, as well as at the end.
2168 8. Added the rcpt_4xx retry error code.
2170 9. Added postmaster_mailfrom=xxx to callout verification option.
2172 10. Added mailfrom=xxxx to the callout verification option, for verify=
2175 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2176 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2178 12. Inserted some casts to stop certain compilers warning when using pointer
2179 differences as field lengths or precisions in printf-type calls (mostly
2180 affecting debugging statements).
2182 13. Added optional readline() support for -be (dynamically loaded).
2184 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2185 same clock tick as a message's arrival, so that its received time was the
2186 same as the "first fail" time on the retry record, and that message
2187 remained on the queue past the ultimate address timeout, every queue runner
2188 would try a delivery (because it was past the ultimate address timeout) but
2189 after another failure, the ultimate address timeout, which should have then
2190 bounced the address, did not kick in. This was a "< instead of <=" error;
2191 in most cases the first failure would have been in the next clock tick
2192 after the received time, and all would be well.
2194 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2195 being recognized when the domain list was tested by the match_domain
2196 condition in an expansion string.
2198 16. Added the ${str2b64: operator.
2200 17. Exim was always calling setrlimit() to set a large limit for the number of
2201 processes, without checking whether the existing limit was already
2202 adequate. (It did check for the limit on file descriptors.) Furthermore,
2203 errors from getrlimit() and setrlimit() were being ignored. Now they are
2204 logged to the main and panic logs, but Exim does carry on, to try to do its
2205 job under whatever limits there are.
2207 18. Imported PCRE 5.0.
2209 19. Trivial typo in log message " temporarily refused connection" (the leading
2212 20. If the log selector return_path_on_delivery was set and an address was
2213 redirected to /dev/null, the delivery process crashed because it assumed
2214 that a return path would always be set for a "successful" delivery. In this
2215 case, the whole delivery is bypassed as an optimization, and therefore no
2218 21. Internal re-arrangement: the function for sending a challenge and reading
2219 a response while authentication was assuming a zero-terminated challenge
2220 string. It's now changed to take a pointer and a length, to allow for
2221 binary data in such strings.
2223 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2225 23. Exim was not respecting finduser_retries when seeking the login of the
2226 uid under which it was called; it was always trying 10 times. (The default
2227 setting of finduser_retries is zero.) Also, it was sleeping after the final
2228 failure, which is pointless.
2230 24. Implemented tls_on_connect_ports.
2232 25. Implemented acl_smtp_predata.
2234 26. If the domain in control=submission is set empty, Exim assumes that the
2235 authenticated id is a complete email address when it generates From: or
2236 Sender: header lines.
2238 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2239 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2240 chown and chgrp in /bin and hostname in /usr/bin.
2242 28. Exim was keeping the "process log" file open after each use, just as it
2243 does for the main log. This opens the possibility of it remaining open for
2244 long periods when the USR1 signal hits a daemon. Occasional processlog
2245 errors were reported, that could have been caused by this. Anyway, it seems
2246 much more sensible not to leave this file open at all, so that is what now
2249 29. The long-running daemon process does not normally write to the log once it
2250 has entered its main loop, and it closes the log before doing so. This is
2251 so that log files can straightforwardly be renamed and moved. However,
2252 there are a couple of unusual error situations where the daemon does write
2253 log entries, and I had neglected to close the log afterwards.
2255 30. The text of an SMTP error response that was received during a remote
2256 delivery was being truncated at 512 bytes. This is too short for some of
2257 the long messages that one sometimes sees. I've increased the limit to
2260 31. It is now possible to make retry rules that apply only when a message has a
2261 specific sender, in particular, an empty sender.
2263 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2264 it possible to be selective about when SMTP synchronization is enforced.
2266 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2268 32. Implemented hosts_connection_nolog.
2270 33. Added an ACL for QUIT.
2272 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2275 35. Added mailbox_size and mailbox_filecount to appendfile.
2277 36. Added control = no_multiline_responses to ACLs.
2279 37. There was a bug in the logic of the code that waits for the clock to tick
2280 in the case where the clock went backwards by a substantial amount such
2281 that the microsecond fraction of "now" was more than the microsecond
2282 fraction of "then" (but the whole seconds number was less).
2284 38. Added support for the libradius Radius client library this is found on
2285 FreeBSD (previously only the radiusclient library was supported).
2291 1. When certain lookups returned multiple values in the form name=value, the
2292 quoting of the values was not always being done properly. Specifically:
2293 (a) If the value started with a double quote, but contained no whitespace,
2295 (b) If the value contained whitespace other than a space character (i.e.
2296 tabs or newlines or carriage returns) it was not quoted.
2297 This fix has been applied to the mysql and pgsql lookups by writing a
2298 separate quoting function and calling it from the lookup code. The fix
2299 should probably also be applied to nisplus, ibase and oracle lookups, but
2300 since I cannot test any of those, I have not disturbed their existing code.
2302 2. A hit in the callout cache for a specific address caused a log line with no
2303 reason for rejecting RCPT. Now it says "Previous (cached) callout
2304 verification failure".
2306 3. There was an off-by-one bug in the queryprogram router. An over-long
2307 return line was truncated at 256 instead of 255 characters, thereby
2308 overflowing its buffer with the terminating zero. As well as fixing this, I
2309 have increased the buffer size to 1024 (and made a note to document this).
2311 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2312 when Exim is waiting for an SMTP response from a remote server, Exim
2313 restarts its select() call on the socket, thereby resetting its timeout.
2314 This is not a problem when such interrupts are rare. Somebody set up a cron
2315 job to run exiwhat every 2 minutes, which is less than the normal select()
2316 timeout (5 or 10 minutes). This meant that the select() timeout never
2317 kicked in because it was always reset. I have fixed this by comparing the
2318 time when an interrupt arrives with the time at the start of the first call
2319 to select(). If more time than the timeout has elapsed, the interrupt is
2320 treated as a timeout.
2322 5. Some internal re-factoring in preparation for the addition of Sieve
2323 extensions (by MH). In particular, the "personal" test is moved to a
2324 separate function, and given an option for scanning Cc: and Bcc: (which is
2325 not set for Exim filters).
2327 6. When Exim created an email address using the login of the caller as the
2328 local part (e.g. when creating a From: or Sender: header line), it was not
2329 quoting the local part when it contained special characters such as @.
2331 7. Installed new OpenBSD configuration files.
2333 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2334 try to make them clearer.
2336 9. Callout options, other than the timeout value, were being ignored when
2337 verifying sender addresses in header lines. For example, when using
2339 verify = header_sender/callout=no_cache
2341 the cache was (incorrectly) being used.
2343 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2344 only the Cygwin environment.
2346 11. When return_path_on_delivery was set as a log selector, if different remote
2347 addresses in the same message used different return paths and parallel
2348 remote delivery occurred, the wrong values would sometimes be logged.
2349 (Whenever a remote delivery process finished, the return path value from
2350 the most recently started remote delivery process was logged.)
2352 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2353 lines when AUTH and/or TLS are in use. This is the "received protocol"
2354 field. Exim used to use "asmtp" for authenticated SMTP, without any
2355 indication (in the protocol name) for TLS use. Now it follows the RFC and
2356 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2357 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2358 names appear in log lines as well as in Received: header lines.
2360 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2361 extensions, and comparison tests, and to fix some bugs.
2363 14. Changes to the "personal" filter test:
2365 (1) The test was buggy in that it was just doing the equivalent of
2366 "contains" tests on header lines. For example, if a user's address was
2367 anne@some.where, the "personal" test would incorrectly be true for
2369 To: susanne@some.where
2371 This test is now done by extracting each address from the header in turn,
2372 and checking the entire address. Other tests that are part of "personal"
2373 are now done using regular expressions (for example, to check local parts
2374 of addresses in From: header lines).
2376 (2) The list of non-personal local parts in From: addresses has been
2377 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2378 taken from the Sieve specification recommendations.
2380 (3) If the message contains any header line starting with "List-" it is
2381 treated as non-personal.
2383 (4) The test for "circular" in the Subject: header line has been removed
2384 because it now seems ill-conceived.
2386 15. Minor typos in src/EDITME comments corrected.
2388 16. Installed latest exipick from John Jetmore.
2390 17. If headers_add on a router specified a text string that was too long for
2391 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2392 of string_sprintf() is now avoided.
2394 18. $message_body_size was not set (it was always zero) when running the DATA
2395 ACL and the local_scan() function.
2397 19. For the "mail" command in an Exim filter, no default was being set for
2398 the once_repeat time, causing a random time value to be used if "once" was
2399 specified. (If the value happened to be <= 0, no repeat happened.) The
2400 default is now 0s, meaning "never repeat". The "vacation" command was OK
2401 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2402 (I found it when inspecting the code).
2404 20. There is now an overall timeout for performing a callout verification. It
2405 defaults to 4 times the callout timeout, which applies to individual SMTP
2406 commands during the callout. The overall timeout applies when there is more
2407 than one host that can be tried. The timeout is checked before trying the
2408 next host. This prevents very long delays if there are a large number of
2409 hosts and all are timing out (e.g. when the network connections are timing
2410 out). The value of the overall timeout can be changed by specifying an
2411 additional sub-option for "callout", called "maxwait". For example:
2413 verify = sender/callout=5s,maxwait=20s
2415 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2416 to the end before writing, but this should make it even safer).
2418 22. Exim was forgetting that it had advertised PIPELINING for the second and
2419 subsequent messages on an SMTP connection. It was also not resetting its
2420 memory on STARTTLS and an internal HELO.
2422 23. When Exim logs an SMTP synchronization error within a session, it now
2423 records whether PIPELINING has been advertised or not.
2425 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2426 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2427 rather than long int.
2429 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2431 26. Added the never_mail option to autoreply.
2437 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2438 crash if the getsockname() call failed; this can happen if a connection is
2439 closed very soon after it is established. The problem was simply in the
2440 order in which certain operations were done, causing Exim to try to write
2441 to the SMTP stream before it had set up the file descriptor. The bug has
2442 been fixed by making things happen in the correct order.
2448 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2449 before the connection was closed, thus losing the rejection response.
2451 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2452 some early Solaris releases, but causes trouble in current releases where
2453 socklen_t is defined.
2455 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2458 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2461 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2462 that releasing the top of it at the end releases what was used for sub-
2463 expansions (unless the block got too big). However, discard this block if
2464 the first thing is a variable or header, so that we can use its block when
2465 it is dynamic (useful for very large $message_headers, for example).
2467 6. Lookups now cache *every* query, not just the most recent. A new, separate
2468 store pool is used for this. It can be recovered when all lookup caches are
2469 flushed. Lookups now release memory at the end of their result strings.
2470 This has involved some general refactoring of the lookup sources.
2472 7. Some code has been added to the store_xxx() functions to reduce the amount
2473 of flapping under certain conditions.
2475 8. log_incoming_interface used to affect only the <= reception log lines. Now
2476 it causes the local interface and port to be added to several more SMTP log
2477 lines, for example "SMTP connection from", and rejection lines.
2479 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2481 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2483 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2484 could be overwritten at the end of the current message (or the start of a
2485 new message if it was set in a HELO ACL). The value is now preserved for
2486 the duration of the SMTP connection.
2488 12. If a transport had a headers_rewrite setting, and a matching header line
2489 contained an unqualified address, that address was qualified, even if it
2490 did not match any rewriting rules. The underlying bug was that the values
2491 of the flags that permit the existence of unqualified sender and recipient
2492 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2493 non-local messages, and by -bnq for local messages) were not being
2494 preserved with the message after it was received.
2496 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2497 "next input=" as part of the text comprising the host identity instead of
2498 the correct text. The code was using the same buffer for two different
2499 strings. However, depending on which order the printing function evaluated
2500 its arguments, the bug did not always show up. Under Linux, for example, my
2501 test suite worked just fine.
2503 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2504 doesn't work with some older versions of Perl. It has been changed to "my",
2505 which in any case is probably the better facility to use.
2507 15. A really picky compiler found some instances of statements for creating
2508 error messages that either had too many or two few arguments for the format
2511 16. The size of the buffer for calls to the DNS resolver has been increased
2512 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2513 for addresses that have a lot of PTR records. This alleviates a problem; it
2514 does not fully solve it.
2516 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2517 buffer now truncates the list and logs the incident, which is the same
2518 action as happens when Exim is looking up a host name and its aliases.
2519 Previously in this situation something unpredictable would happen;
2520 sometimes it was "internal error: store_reset failed".
2522 18. If a server dropped the connection unexpectedly when an Exim client was
2523 using GnuTLS and trying to read a response, the client delivery process
2524 crashed while trying to generate an error log message.
2526 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2527 string, for example:
2529 warn message = H1: something\nH2: something
2531 the text was added as a single header line from Exim's point of view
2532 though it ended up OK in the delivered message. However, searching for the
2533 second and subsequent header lines using $h_h2: did not work. This has been
2534 fixed. Similarly, if a system filter added multiple headers in this way,
2535 the routers could not see them.
2537 20. Expanded the error message when iplsearch is called with an invalid key to
2538 suggest using net-iplsearch in a host list.
2540 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2541 ACLs are no longer actually imposed (and a message to that effect is
2544 22. If a "gecos" field in a passwd entry contained escaped characters, in
2545 particular, if it contained a \" sequence, Exim got it wrong when building
2546 a From: or a Sender: header from that name. A second bug also caused
2547 incorrect handling when an unquoted " was present following a character
2548 that needed quoting.
2550 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2551 was not being matched caselessly.
2553 24. Arranged for all hyphens in the exim.8 source to be escaped with
2556 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2557 properly. Recipient callout cache records were still being keyed to include
2558 the sender, even when use_sender was set false. This led to far more
2559 callouts that were necessary. The sender is no longer included in the key
2560 when use_sender is false.
2562 26. Added "control = submission" modifier to ACLs.
2564 27. Added the ${base62d: operator to decode base 62 numbers.
2566 28. dnsdb lookups can now access SRV records.
2568 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2569 the configuration file.
2571 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2572 (-v) mode. This makes the output for a verbose queue run more intelligible.
2574 31. Added a use_postmaster feature to recipient callouts.
2576 32. Added the $body_zerocount variable, containing the number of binary zero
2577 bytes in the message body.
2579 33. The time of last modification of the "new" subdirectory is now used as the
2580 "mailbox time last read" when there is a quota error for a maildir
2583 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2585 35. Added +ignore_unknown as a special item in host lists.
2587 36. Code for decoding IPv6 addresses in host lists is now included, even if
2588 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2589 address was recognized as an IP address, but was then not correctly decoded
2590 into binary, causing unexpected and incorrect effects when compared with
2597 1. Very minor rewording of debugging text in manualroute to say "list of
2598 hosts" instead of "hostlist".
2600 2. If verify=header_syntax was set, and a header line with an unqualified
2601 address (no domain) and a large number of spaces between the end of the
2602 name and the colon was received, the reception process suffered a buffer
2603 overflow, and (when I tested it) crashed. This was caused by some obsolete
2604 code that should have been removed. The fix is to remove it!
2606 3. When running in the test harness, delay a bit after writing a bounce
2607 message to get a bit more predictability in the log output.
2609 4. Added a call to search_tidyup() just before forking a reception process. In
2610 theory, someone could use a lookup in the expansion of smtp_accept_max_
2611 per_host which, without the tidyup, could leave open a database connection.
2613 5. Added the variables $recipient_data and $sender_data which get set from a
2614 lookup success in an ACL "recipients" or "senders" condition, or a router
2615 "senders" option, similar to $domain_data and $local_part_data.
2617 6. Moved the writing of debug_print from before to after the "senders" test
2620 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2621 problems for message scanning, either using a data ACL, or using
2622 local_scan() because the Received: header was not generated till after they
2623 were called (in order to set the time as the time of reception completion).
2624 I have revised the way this works. The header is now generated after the
2625 body is received, but before the ACL or local_scan() are called. After they
2626 are run, the timestamp in the header is updated.
2632 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2633 before starting a queue runner without re-exec. This happened only when
2634 deliver_drop_privilege was set or when the Exim user was set to root. The
2635 effect of the bug was that timeouts during subsequent deliveries caused
2636 crashes instead of being properly handled. The handler is now left at its
2637 default (and expected) setting.
2639 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2640 message, again when deliver_drop_privilege is set or Exim is run as root.
2641 The bug described in (1) was not present in this case, but the tidying up
2642 of the other signals was missing. I have made the two cases consistent.
2644 3. The ignore_target_hosts setting on a manualroute router was being ignored
2645 for hosts that were looked up using the /MX notation.
2647 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2650 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2651 operated on the sender address. After changing the $sender_address to <>
2652 for the sender address verify, Exim was re-instated it as the original
2653 (before rewriting) address, but remembering that it had rewritten it, so it
2654 wasn't rewriting it again. This bug also had the effect of breaking the
2655 sender address verification caching when the sender address was rewritten.
2657 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2658 This has been changed so that if the ip literal address matches
2659 ignore_target_hosts, the router declines.
2661 7. Added expansion conditions match_domain, match_address, and match_local_
2662 part (NOT match_host).
2664 8. The placeholder for the Received: header didn't have a length field set.
2666 9. Added code to Exim itself and to exim_lock to test for a specific race
2667 condition that could lead to file corruption when using MBX delivery. The
2668 issue is with the lockfile that is created in /tmp. If this file is removed
2669 after a process has opened it but before that process has acquired a lock,
2670 there is the potential for a second process to recreate the file and also
2671 acquire a lock. This could lead to two Exim processes writing to the file
2672 at the same time. The added code performs the same test as UW imapd; it
2673 checks after acquiring the lock that its file descriptor still refers to
2674 the same named file.
2676 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2677 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2680 11. Added the smtp_active_hostname option. If used, this will typically be made
2681 to depend on the incoming interface address. Because $interface_address is
2682 not set up until the daemon has forked a reception process, error responses
2683 that can happen earlier (such as "too many connections") no longer contain
2686 12. If an expansion in a condition on a "warn" statement fails because a lookup
2687 defers, the "warn" statement is abandoned, and the next ACL statement is
2688 processed. Previously this caused the whole ACL to be aborted.
2690 13. Added the iplsearch lookup type.
2692 14. Added ident_timeout as a log selector.
2694 15. Added tls_certificate_verified as a log selector.
2696 16. Added a global option tls_require_ciphers (compare the smtp transport
2697 option of the same name). This controls incoming TLS connections.
2699 17. I finally figured out how to make tls_require_ciphers do a similar thing
2700 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2701 before starting the TLS session.
2703 18. Tabs are now shown as \t in -bP output.
2705 19. If the log selector return_path_on_delivery was set, Exim crashed when
2706 bouncing a message because it had too many Received: header lines.
2708 20. If two routers both had headers_remove settings, and the first one included
2709 a superfluous trailing colon, the final name in the first list and the
2710 first name in the second list were incorrectly joined into one item (with a
2711 colon in the middle).
2717 1. Added -C and -D options to the exinext utility, mainly to make it easier
2718 to include in the automated testing, but these could be helpful when
2719 multiple configurations are in use.
2721 2. The exinext utility was not formatting the output nicely when there was
2722 an alternate port involved in the retry record key, nor when there was a
2723 message id as well (for retries that were specific to a specific message
2724 and a specific host). It was also confused by IPv6 addresses, because of
2725 the additional colons they contain. I have fixed the IPv4 problem, and
2726 patched it up to do a reasonable job for IPv6.
2728 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2729 delivery, the log line now contains "pipelined" if PIPELINING was used.
2731 4. An SMTP transport process used to panic and die if the bind() call to set
2732 an explicit outgoing interface failed. This has been changed; it is now
2733 treated in the same way as a connect() failure.
2735 5. A reference to $sender_host_name in the part of a conditional expansion
2736 that was being skipped was still causing a DNS lookup. This no longer
2739 6. The def: expansion condition was not recognizing references to header lines
2740 that used bh_ and bheader_.
2742 7. Added the _cache feature to named lists.
2744 8. The code for checking quota_filecount in the appendfile transport was
2745 allowing one more file than it should have been.
2747 9. For compatibility with Sendmail, the command line option
2755 and sets the incoming protocol and host name (for trusted callers). The
2756 host name and its colon can be omitted when only the protocol is to be set.
2757 Note the Exim already has two private options, -pd and -ps, that refer to
2758 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2759 "s", but I don't think that's a major issue.
2761 10. A number of refactoring changes to the code, none of which should affect
2764 (a) The number of logging options was getting close to filling up the
2765 32-bit word that was used as a bit map. I have split them into two classes:
2766 those that are passed in the argument to log_write(), and those that are
2767 only ever tested independently outside of that function. These are now in
2768 separate 32-bit words, so there is plenty of room for expansion again.
2769 There is no change in the user interface or the logging behaviour.
2771 (b) When building, for example, log lines, the code previously used a
2772 macro that called string_cat() twice, in order to add two strings. This is
2773 not really sufficiently general. Furthermore, there was one instance where
2774 it was actually wrong because one of the argument was used twice, and in
2775 one call a function was used. (As it happened, calling the function twice
2776 did not affect the overall behaviour.) The macro has been replaced by a
2777 function that can join an arbitrary number of extra strings onto a growing
2780 (c) The code for expansion conditions now uses a table and a binary chop
2781 instead of a serial search (which was left over from when there were very
2782 few conditions). Also, it now recognizes conditions like "pam" even when
2783 the relevant support is not compiled in: a suitably worded error message is
2784 given if an attempt is made to use such a condition.
2786 11. Added ${time_interval:xxxxx}.
2788 12. A bug was causing one of the ddress fields not to be passed back correctly
2789 from remote delivery subprocesses. The field in question was not being
2790 subsequently used, so this caused to problems in practice.
2792 13. Added new log selectors queue_time and deliver_time.
2794 14. Might have fixed a bug in maildirsizefile handling that threw up
2795 "unexpected character" debug warnings, and recalculated the data
2796 unnecessarily. In any case, I expanded the warning message to give more
2799 15. Added the message "Restricted characters in address" to the statements in
2800 the default ACL that block characters like @ and % in local parts.
2802 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2803 Three changes have been made:
2805 (a) There was a serious bug; a negative response to MAIL caused the whole
2806 recipient domain to be cached as invalid, thereby blocking all messages
2807 to all local parts at the same domain, from all senders. This bug has
2808 been fixed. The domain is no longer cached after a negative response to
2809 MAIL if the sender used is not empty.
2811 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2814 (c) A new callout option, "use_sender" has been added for people who want
2815 the modified behaviour.
2821 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2824 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2825 indeed breaks things for older releases.
2827 3. Added additional logging to the case where there is a problem reading data
2828 from a filter that is running in a subprocess using a pipe, in order to
2829 try to track down a specific problem.
2831 4. Testing facility fudge: when running in the test harness and attempting
2832 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2833 getting "No route to host". Convert this to a timeout.
2835 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2838 6. Some OS don't have socklen_t but use size_t instead. This affects the
2839 fifth argument of getsockopt() amongst other things. This is now
2840 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2841 can be set for individual OS. I have set it for SunOS5, OSF1, and
2842 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2843 some earlier ones do not.
2845 7. Change 4.30/15 was not doing the test caselessly.
2847 8. The standard form for an IPv6 address literal was being rejected by address
2848 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2849 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2850 this, as well as the form without the "IPv6" on the front (but only when
2851 address literals are enabled, of course).
2853 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2855 10. Exim crashed if a message with an empty sender address specified by -f
2856 encountered a router with an errors_to setting. This could be provoked only
2857 by a command such as
2861 where an empty string was supplied; "<>" did not hit this bug.
2863 11. Installed PCRE release 4.5.
2865 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2866 remained set. It is now erased.
2868 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2869 times from message ids (which are base 36 rather than the normal 62).
2871 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2872 were being counted as actual protocol errors, and logged if the log
2873 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2874 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2875 and DATA following a set of rejected RCPTs do not count as protocol errors.
2876 In other words, Exim assumes they were pipelined, though this may not
2877 actually be the case. Of course, in all cases the client gets an
2878 appropriate error code.
2880 15. If a lookup fails in an ACL condition, a message about the failure may
2881 be available; it is used if testing the ACL cannot continue, because most
2882 such messages specify what the cause of the deferral is. However, some
2883 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2884 that caused an old message to be retained and used if a later statement
2885 caused a defer, replacing the real cause of the deferral.
2887 16. If an IP address had so many PTR records that the DNS lookup buffer
2888 was not large enough to hold them, Exim could crash while trying to process
2889 the truncated data. It now detects and logs this case.
2891 17. Further to 4.21/58, another change has been made: if (and only if) the
2892 first line of a message (the first header line) ends with CRLF, a bare LF
2893 in a subsequent header line has a space inserted after it, so as not to
2894 terminate the header.
2896 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2897 unnecessarily, used atoi() instead of strtol(), and didn't check the
2898 termination when getting file sizes from file names by regex.
2900 19. Completely re-implemented the support for maildirsize files, in the light
2901 of a number of problems with the previous contributed implementation
2902 (4.30/29). In particular:
2904 . If the quota is zero, the maildirsize file is maintained, but no quota is
2907 . If the maildir directory does not exist, it is created before any attempt
2908 to write a maildirsize file.
2910 . The quota value in the file is just a cache; if the quota is changed in
2911 the transport, the new value overrides.
2913 . A regular expression is available for excluding directories from the
2916 20. The autoreply transport checks the characters in options that define the
2917 message's headers; it allows continued headers, but it was checking with
2918 isspace() after an embedded newline instead of explicitly looking for a
2921 21. If all the "regular" hosts to which an address was routed had passed their
2922 expiry times, and had not reached their retry times, the address was
2923 bounced, even if fallback hosts were defined. Now Exim should go on to try
2926 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2927 equivalent code in the SMTP transport. Some hosts send humungous responses
2928 to HELO/EHLO, more than 1024 it seems.
2930 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2931 gives compiler warnings in some environments. I've now done it "properly",
2934 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2935 (because of problems with the built-in one) was declared to return uschar *
2936 instead of char *, causing compiler failure.
2938 25. Fixed a file descriptor leak when processing alias/forward files.
2940 26. Fixed a minor format string issue in dbfn.c.
2942 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2944 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2945 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2948 29. When a sender address is verified, it is cached, to save repeating the test
2949 when there is more than one recipient in a message. However, when the
2950 verification involves a callout, it is possible for different callout
2951 options to be set for different recipients. It is too complicated to keep
2952 track of this in the cache, so now Exim always runs a verification when a
2953 callout is required, relying on the callout cache for the optimization.
2954 The overhead is duplication of the address routing, but this should not be
2957 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2958 to be verified with callout=postmaster, and the main callout worked but the
2959 postmaster check failed, the verification correctly failed. However, if a
2960 subsequent RCPT command asked for sender verification *without* the
2961 postmaster check, incorrect caching caused this verification also to fail,
2964 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2965 it was not caching the DNS options (qualify_single, search_parents) that
2966 were used when the lookup failed. A subsequent lookup with different
2967 options therefore always gave the same answer, though there were cases
2968 where it should not have. (Example: a "domains = !$mx_any" option on a
2969 dnslookup router: the "domains" option is always processed without any
2970 widening, but the router might have qualify_single set.) Now Exim uses the
2971 cached value only when the same options are set.
2973 32. Added John Jetmore's "exipick" utility to the distribution.
2975 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2976 than a timeout (e.g. a certificate is required, and is not provided), an
2977 Exim server now closes the connection immediately. Previously it waited for
2978 the client to close - but if the client is SSL, it seems that they each
2979 wait for each other, leading to a delay before one of them times out.
2981 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2982 maintained 0.8.x compatibility because I don't think many are using it, and
2983 it is clearly obsolete.
2985 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2988 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2989 client certificate was expired. A simple patch fixes this, though I don't
2990 understand the full logic of why the verify callback is called multiple
2993 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2994 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2995 which causes problems with some clients (such as the Certicom SSL Plus
2996 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2997 disables the coutermeasure allowing Eudora to connect."
2999 38. Exim was not checking that a write() to a log file succeeded. This could
3000 lead to Bad Things if a log got too big, in particular if it hit a file
3001 size limit. Exim now panics and dies if it cannot write to a log file, just
3002 as it does if it cannot open a log file.
3004 39. Modified OS/Makefile-Linux so that it now contains
3006 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3008 The two -D definitions ensure that Exim is compiled with large file
3009 support, which makes it possible to handle log files that are bigger than
3012 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3013 instance) a domain was checked against a named list that involved a lookup,
3014 causing $domain_data to be set, then another domain was checked against the
3015 same list, then the first domain was re-checked, the value of $domain_data
3016 after the final check could be wrong. In particular, if the second check
3017 failed, it could be set empty. This bug probably also applied to
3020 41. The strip_trailing_dot option was not being applied to the address given
3021 with the -f command-line option.
3023 42. The code for reading a message's header from the spool was incrementing
3024 $received_count, but never initializing it. This meant that the value was
3025 incorrect (doubled) while delivering a message in the same process in which
3026 it was received. In the most common configuration of Exim, this never
3027 happens - a fresh exec is done - but it can happen when
3028 deliver_drop_privilege is set.
3030 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3031 it now includes up to 150 characters of the unexpected data in the log
3034 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3035 and building data strings. The size of both of these buffers was 10 000
3036 bytes - far larger than anybody would *ever* want, thought I. Needless to
3037 say, somebody hit the limit. I have increased the maximum line length to
3038 20 000 and the maximum data length of concatenated lines to 100 000. I have
3039 also fixed two bugs, because there was no checking on these buffers. Tsk,
3040 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3041 buffer is too small.
3043 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3044 lsearch lookups. Now it does.
3046 46. When parsing a route_list item in a manualroute router, a fixed-length
3047 buffer was used for the list of hosts. I made this 1024 bytes long,
3048 thinking that nobody would ever have a list of hosts that long. Wrong.
3049 Somebody had a whole pile of complicated expansion conditions, and the
3050 string was silently truncated, leading to an expansion error. It turns out
3051 that it is easier to change to an unlimited length (owing to other changes
3052 that have happened since this code was originally written) than to build
3053 structure for giving a limitation error. The length of the item that
3054 expands into the list of hosts is now unlimited.
3056 47. The lsearch lookup could not handle data where the length of text line was
3057 more than 4095 characters. Such lines were truncated, leading to shortened
3058 data being returned. It should now handle lines of any length.
3060 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3061 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3064 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3065 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3066 no longer happens in comment lines. A list of replacements is now placed
3067 at the head of all of the source files, except those whose only change is
3068 to replace PERL_COMMAND in the very first #! line.
3070 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3071 messages on the queue, with a bottom-up merge sort, using code contributed
3072 by Michael Haardt. This should make operations like -bp somewhat faster on
3073 large queues. It won't affect queue runners, except when queue_run_in_order
3076 51. Installed eximstats 1.31 in the distribution.
3078 52. Added support for SRV lookups to the dnslookup router.
3080 53. If an ACL referred to $message_body or $message_body_end, the value was not
3081 reset for any messages that followed in the same SMTP session.
3083 54. The store-handling optimization for building very long strings was not
3084 differentiating between the different store pools. I don't think this
3085 actually made any difference in practice, but I've tidied it.
3087 55. While running the routers to verify a sender address, $sender_address
3088 was still set to the sender address. This is wrong, because when routing to
3089 send a bounce to the sender, it would be empty. Therefore, I have changed
3090 it so that, while verifying a sender address, $sender_address is set to <>.
3091 (There is no change to what happens when verifying a recipient address.)
3093 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3094 target A or AAAA records (if not already returned) without resetting the
3095 qualify_single or search_parents options of the DNS resolver. These are
3096 inappropriate in this case because the targets of MX and SRV records must
3097 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3098 target that, when qualified, matched something in the local domain. These
3099 two options are now turned off when doing these lookups.
3101 57. It seems that at least some releases of Reiserfs (which does not have the
3102 concept of a fixed number of inodes) returns zero and not -1 for the
3103 number of available inodes. This interacted badly with check_spool_inodes,
3104 which assumed that -1 was the "no such thing" setting. What I have done is
3105 to check that the total number of inodes is greater than zero before doing
3106 the test of how many are available.
3108 58. When a "warn" ACL statement has a log_message modifier, the message is
3109 remembered, and not repeated. This is to avoid a lot of repetition when a
3110 message has many recipients that cause the same warning to be written.
3111 Howewer, Exim was preserving the list of already written lines for an
3112 entire SMTP session, which doesn't seem right. The memory is now reset if a
3113 new message is started.
3115 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3116 debugging output unless log_rewrite was also set.
3118 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3119 of (int)(handle) when we know that handle contains (void *)(-1).
3121 61. The Exim daemon panic-logs an error return when it closes the incoming
3122 connection. However "connection reset by peer" seems to be common, and
3123 isn't really an error worthy of noting specially, so that particular error
3126 62. When Exim is trying to find all the local interfaces, it used to panic and
3127 die if the ioctl to get the interface flags failed. However, it seems that
3128 on at least one OS (Solaris 9) it is possible to have an interface that is
3129 included in the list of interfaces, but for which you get a failure error
3130 for this call. This happens when the interface is not "plumbed" into a
3131 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3132 failure of the "get flags" call assumes that the interface is down.
3134 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3135 makes life easier for people who are doing arithmetic on fields extracted
3136 from dates, where you often get leading zeros that should not be
3137 interpreted as octal.
3139 64. Added qualify_domain to the redirect router, to override the global
3142 65. If a pathologically long header line contained very many addresses (the
3143 report of this problem mentioned 10 000) and each of them was rewritten,
3144 Exim could use up a very large amount of memory. (It kept on making new
3145 copies of the header line as it rewrote, and never released the old ones.)
3146 At the expense of a bit more processing, the header rewriting function has
3147 been changed so that it no longer eats memory in this way.
3149 66. The generation of the Received: header has been moved from the time that a
3150 message starts to be received, to the time that it finishes. The timestamp
3151 in the Received: header should now be very close to that of the <= log
3152 line. There are two side-effects of this change:
3154 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3155 logged header lines no longer include the local Received: line, because
3156 it has not yet been created. The same applies to a copy of the message
3157 that is returned to a non-SMTP sender when a message is rejected.
3159 (b) When a filter file is tested using -bf, no additional Received: header
3160 is added to the test message. After some thought, I decided that this
3163 This change does not affect the value of $received_for. It is still set
3164 after address rewriting, but before local_scan() is called.
3166 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3168 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3169 gave an unhelpful panic error message, and a defer error. I have managed to
3170 change this behaviour so that it now rejects any supplied certificate,
3171 which seems right, as the list of acceptable certificates is empty.
3173 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3174 gave an unhelpful defer error. I have not managed to make this reject any
3175 supplied certificates, but the error message it gives is "no certificate
3176 supplied", which is not helpful.
3178 70. exigrep's output now also includes lines that are not associated with any
3179 message, but which match the given pattern. Implemented by a patch from
3180 Martin Sluka, which also tidied up the Perl a bit.
3182 71. Recipient callout verification, like sender verification, was using <> in
3183 the MAIL FROM command. This isn't really the right thing, since the actual
3184 sender may affect whether the remote host accepts the recipient or not. I
3185 have changed it to use the actual sender in the callout; this means that
3186 the cache record is now keyed on a recipient/sender pair, not just the
3187 recipient address. There doesn't seem to be a real danger of callout loops,
3188 since a callout by the remote host to check the sender would use <>.
3189 [SEE ABOVE: changed after hitting problems.]
3191 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3192 temporary errors. However, in the case of such a code being given after
3193 the end of a data transmission (i.e. after ".") Exim was failing to write
3194 a retry record for the message. (Yes, there was some broken host that was
3195 actually sending 8xx at this point.)
3197 73. An unknown lookup type in a host list could cause Exim to panic-die when
3198 the list was checked. (An example that provoked this was putting <; in the
3199 middle of a list instead of at the start.) If this happened during a DATA
3200 ACL check, a -D file could be left lying around. This kind of configuration
3201 error no longer causes Exim to die; instead it causes a defer errror. The
3202 incident is still logged to the main and panic logs.
3204 74. Buglet left over from Exim 3 conversion. The message "too many messages
3205 in one connection" was written to the rejectlog but not the mainlog, except
3206 when address rewriting (yes!) was being logged.
3208 75. Added write_rejectlog option.
3210 76. When a system filter was run not as root (that is, when system_filter_user
3211 was set), the values of the $n variables were not being returned to the
3212 main process; thus, they were not subsequently available in the $sn
3215 77. Added +return_path_on_delivery log selector.
3217 78. A connection timeout was being treated differently from recipients deferred
3218 when testing hosts_max_try with a message that was older than the host's
3219 retry timeout. (The host should not be counted, thus allowing all hosts to
3220 be tried at least once before bouncing.) This may have been the cause of an
3221 occasionally reported bug whereby a message would remain on the queue
3222 longer than the retry timeout, but would be bounced if a delivery was
3223 forced. I say "may" because I never totally pinned down the problem;
3224 setting up timeout/retry tests is difficult. See also the next item.
3226 79. The ultimate address timeout was not being applied to errors that involved
3227 a combination of host plus message (for example, a timeout on a MAIL
3228 command). When an address resolved to a number of possible hosts, and they
3229 were not all tried for each delivery (e.g. because of hosts_max_try), a
3230 message could remain on the queue longer than the retry timeout.
3232 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3235 81. Fixed an obscure SMTP outgoing bug which required at least the following
3236 conditions: (a) there was another message waiting for the same server;
3237 (b) the server returned 5xx to all RCPT commands in the first message so
3238 that the message was not completed; (c) the server dropped the connection
3239 or gave a negative response to the RSET that Exim sends to abort the
3240 transaction. The observed case was a dropped connection after DATA that had
3241 been sent in pipelining mode. That is, the server had advertised PIPELINING
3242 but was not implementing it correctly. The effect of the bug was incorrect
3243 behaviour, such as trying another host, and this could lead to a crash.
3249 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3250 and daemon.c were passed as pointers to ints; they should have been
3251 pointers to socklen_t variables (which are typically unsigned ints).
3253 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3256 3. Fixed a really odd bug that affected only the testing scheme; patching a
3257 certain fixed string in the binary changed the value of another string that
3258 happened to be identical to the end of the original first string.
3260 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3261 name", it returns that address as the IP address. On some operating
3262 systems (e.g. Solaris), it also passes back the IP address string as the
3263 "host name". However, on others (e.g. Linux), it passes back an empty
3264 string. Exim wasn't checking for this, and was changing the host name to an
3265 empty string, assuming it had been canonicized.
3267 5. Although rare, it is permitted to have more than one PTR record for a given
3268 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3269 all the names associated with an address, because they do in Solaris.
3270 However, it seems that they do not in Linux for data that comes from the
3271 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3272 I found this out when I moved to a new Linux workstation and tried to run
3273 the Exim test suite.
3275 To get round this problem I have changed the code so that it now does its
3276 own call to the DNS to look up PTR records when searching for a host name.
3277 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3278 addresses that are only in /etc/hosts are still found.
3280 This behaviour is, however, controlled by an option called host_lookup_
3281 order, which defaults to "bydns:byaddr". If people want to use the other
3282 order, or indeed, just use one or the other means of lookup, they can
3283 specify it in this variable.
3285 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3286 some operating systems, this comes back from gethostbyaddr() as an empty
3287 string, and this is what Exim used to test for. However, it seems that in
3288 other systems, "." is yielded. Exim now tests for this case too.
3290 7. The values of check_spool_space and check_log_space are now held internally
3291 as a number of kilobytes instead of an absolute number of bytes. If a
3292 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3293 kilobyte. This means that much larger values can be stored.
3295 8. Exim monitor: an attempt to get the action menu when not actually pointing
3296 at a message produces an empty menu entitled "No message selected". This
3297 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3298 no entries in it ("Shell widget menu has zero width and/or height"). So I
3299 have added a single, blank menu entry in this case.
3301 9. Added ${quote_local_part.
3303 10. MIME decoding is now applied to the contents of Subject: header lines when
3306 11. Now that a reference to $sender_host_address automatically causes a reverse
3307 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3308 host lookup before query-style lookups in lists that might use this
3309 variable. This has therefore been abolished, and the "net-" prefix is no
3310 longer necessary for query-style lookups.
3312 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3313 have been a typo for LFLAGS, so it has been changed.
3315 13. The install script calls Exim with "-C /dev/null" in order to find the
3316 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3317 to be output. Howeve, since Exim outputs its version number before the
3318 error, it didn't break the script. It just looked ugly. I fixed this by
3319 always allowing "-C /dev/null" if the caller is root.
3321 14. Ignore overlarge ACL variable number when reading spool file - insurance
3322 against a later release with more variables having written the file.
3324 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3325 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3326 this, as well as the form without the "IPv6" on the front.
3328 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3329 OS/Makefile-Darwin file.
3331 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3332 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3334 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3335 variable. However, some OS use ints for inodes. I've added cast to long int
3336 to get rid of the compiler warning.
3338 19. I had forgotten to lock out "/../" in configuration file names when
3339 ALT_CONFIG_PREFIX was set.
3341 20. Routers used for verification do not need to specify transports. However,
3342 if such a router generated a host list, and callout was configured, Exim
3343 crashed, because it could not find a port number from the (non-existent)
3344 transport. It now assumes port 25 in this circumstance.
3346 21. Added the -t option to exigrep.
3348 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3349 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3350 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3351 them is set, LOOKUP_LSEARCH is forced.
3353 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3355 24. Added sender and host information to the "rejected by local_scan()" log
3356 line; previously there was no indication of these.
3358 25. Added .include_if_exists.
3360 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3361 Linux. It turns out that not all file systems support this. Apparently some
3362 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3363 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3364 not supported on the file descriptor, is now ignored when Exim is trying to
3365 sync a directory. This applies only to Linux.
3367 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3369 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3370 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3373 29. Added support for maildirsize files from supplied patch (modified a bit).
3375 30. The use of :fail: followed by an empty string could lead Exim to respond to
3376 sender verification failures with (e.g.):
3378 550 Verification failed for <xxx>
3379 550 Sender verify failed
3381 where the first response line was missing the '-' that indicates it is not
3382 the final line of the response.
3384 31. The loop for finding the name of the user that called Exim had a hardwired
3385 limit of 10; it now uses the value of finduser_retries, which is used for
3386 all other user lookups.
3388 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3391 33. Exim was neglecting to zero errno before one call of strtol() when
3392 expanding a string and expecting an integer value. On some systems this
3393 resulted in spurious "integer overflow" errors. Also, it was casting the
3394 result into an int without checking.
3396 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3397 did not work. The code looks as if it has *never* worked, though it appears
3398 to have been documented since at least releast 1.62. I have made it work.
3400 35. The "timeout_DNS" error in retry rules, also documented since at least
3401 1.62, also never worked. As it isn't clear exactly what this means, and
3402 clearly it isn't a major issue, I have abolished the feature by treating it
3403 as "timeout", and writing a warning to the main and panic logs.
3405 36. The display of retry rules for -brt wasn't always showing the error code
3408 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3409 timeout_connect_A, timeout_connect_MX.
3411 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3412 to the empty sender.
3414 39. The daemon was not analysing the content of -oX till after it had closed
3415 stderr and disconnected from the controlling terminal. This meant that any
3416 syntax errors were only noted on the panic log, and the return code from
3417 the command was 0. By re-arranging the code a little, I've made the
3418 decoding happen first, so such errors now appear on stderr, and the return
3419 code is 1. However, the actual setting up of the sockets still happens in
3420 the disconnected process, so errors there are still only recorded on the
3423 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3424 connections (as happens on some IP stacks) was logged at start up time as
3425 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3426 it from "IPv6 and IPv4", which means that two separate sockets are being
3429 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3430 says whether AF_INET or AF_INET6 was passed as an argument.
3432 42. Exiwhat output was messed up when time zones were included in log
3435 43. Exiwhat now gives more information about the daemon's listening ports,
3436 and whether -tls-on-connect was used.
3438 44. The "port" option of the smtp transport is now expanded.
3440 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3441 silently ignored. Now an error message is written to the main and panic
3444 46. There's a new ACL modifier called "logwrite" which writes to a log file
3445 as soon as it is encountered.
3447 47. Added $local_user_uid and $local_user_gid at routing time.
3449 48. Exim crashed when trying to verify a sender address that was being
3452 49. Exim was recognizing only a space character after ".include". It now also
3453 recognizes a tab character.
3455 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3456 extracting the relevant information from the specification. The man page no
3457 longer contains scrambled data for the -d option, and I've added a section
3458 at the front about calling Exim under different names.
3460 51. Added "extra_headers" argument to the "mail" command in filter files.
3462 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3465 53. Installed eximstats 1.29.
3467 54. Added transport_filter_timeout as a generic transport option.
3469 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3470 Cc: header lines. This was required by RFC 822, but it not required by RFC
3473 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3474 incoming messages that did not have them. Now it does so only if the
3475 message originates locally, that is, if there is no associated remote host
3476 address. When Resent- header lines are present, this applies to the Resent-
3477 lines rather than the non-Resent- lines.
3479 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3480 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3482 58. Messages for configuration errors now include the name of the main
3483 configuration file - useful now that there may be more than one file in a
3484 list (.included file names were always shown).
3486 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3487 for those rare installations that do not start the daemon as root or run it
3488 setuid root. I've cut out the call to initgroups() if the daemon is not
3491 60. The Exim user and group can now be bound into the binary as text strings
3492 that are looked up at the start of Exim's processing.
3494 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3496 62. Added $mailstore_basename variable.
3498 63. Installed patch to sieve.c from Michael Haardt.
3500 64. When Exim failed to open the panic log after failing to open the main log,
3501 the original message it was trying to log was written to stderr and debug
3502 output, but if they were not available (the usual case in production), it
3503 was lost. Now it is written to syslog before the two lines that record the
3504 failures to open the logs.
3506 65. Users' Exim filters run in subprocesses under the user's uid. It is
3507 possible for a "deliver" command or an alias in a "personal" command to
3508 provoke an address rewrite. If logging of address rewriting is configured,
3509 this fails because the process is not running as root or exim. There may be
3510 a better way of dealing with this, but for the moment (because 4.30 needs
3511 to be released), I have disabled address rewrite logging when running a
3512 filter in a non-root, non-exim process.
3518 1. The buildconfig auxiliary program wasn't quoting the value set for
3519 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3520 not defined. This bug was masked in 4.22 by the effect that was fixed in
3523 2. Some messages that were rejected after a message id was allocated were
3524 shown as "incomplete" by exigrep. It no longer does this for messages that
3525 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3527 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3528 have allow_domain_literals set, the ID did not get logged in the <= line.
3529 Domain literals are now always recognized in Message-ID: header lines.
3531 4. The first argument for a ${extract expansion item is the key name or field
3532 number. Leading and trailing spaces in this item were not being ignored,
3533 causing some misleading effects.
3535 5. When deliver_drop_privilege was set, single queue runner processes started
3536 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3537 same command in the process it spins off) were not dropping privilege.
3539 6. When the daemon running as "exim" started a queue runner, it always
3540 re-executed Exim in the spun-off process. This is a waste of effort when
3541 deliver_drop_privilege is set. The new process now just calls the
3542 queue-runner function directly.
3548 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3551 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3552 ignored. Though the use of -oP was forcing the writing of a pid file, it
3553 was always written to the default place.
3555 3. If the message "no IP address found for host xxxx" is generated during
3556 incoming verification, it is now followed by identification of the incoming
3557 connection (so you can more easily find what provoked it).
3559 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3561 5. Added some features to "harden" Exim a bit more against certain attacks:
3563 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3564 be put in Local/Makefile. This is like the never_users runtime option,
3565 but it cannot be overridden. The default setting is "root".
3567 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3568 prefix string with which any file named in a -C command line option
3571 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3572 is retained for -C and -D only if the caller of Exim is root. Without
3573 it, the exim user may also use -C and -D and retain privilege.
3575 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3576 command line option is disabled.
3578 6. Macro names set by the -D option must start with an upper case letter, just
3579 like macro names defined in the configuration file.
3581 7. Added "dereference=" facility to LDAP.
3583 8. Two instances of the typo "uknown" in the source files are fixed.
3585 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3586 the Configure-Makefile script screwed up while processing it.
3588 10. Incorporated PCRE 4.4.
3590 11. The SMTP synchronization check was not operating right at the start of an
3591 SMTP session. For example, it could not catch a HELO sent before the client
3592 waited for the greeting. There is now a check for outstanding input at the
3593 point when the greeting is written. Because of the duplex, asynchronous
3594 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3595 way, but not yet received, when the check is performed.
3597 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3598 on TCP/IP sockets, because this apparently causes some broken clients to
3601 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3602 unchanged) from the Cygwin maintainer.
3604 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3605 was supported instead of testing for mbx. Effectively a typo.
3607 15. The spa authenticator server code was not checking that the input it
3608 received was valid base64.
3610 16. The debug output line for the "set" modifier in ACLs was not showing the
3611 name of the variable that was being set.
3613 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3615 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3616 lookup on its own. Something else was needed to trigger the lookup. For
3617 example, a match in host_lookup or the need for a host name in a host list.
3618 Now, if $sender_host_name is referenced and the host name has not yet been
3619 looked up, a lookup is performed. If the lookup fails, the variable remains
3620 empty, and $host_lookup_failed is set to "1".
3622 19. Added "eqi" as a case-independent comparison operator.
3624 20. The saslauthd authentication condition could segfault if neither service
3625 nor realm was specified.
3627 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3628 error message that was logged was misleading, and incoming SMTP
3629 connections were dropped. The message is now more accurate, and temporary
3630 errors are given to SMTP connections.
3632 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3633 (except RSET) with a 421 error, until QUIT is received. However, it was
3634 failing to send a response to QUIT.
3636 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3637 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3638 ACL is now run after the helo_try_verify_hosts code.
3640 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3641 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3642 case-independent, so other case variants are also recognized). Apparently
3643 some systems use these upper case variants.
3645 25. If more than two messages were waiting for the same host, and a transport
3646 filter was specified for the transport, Exim sent two messages over the
3647 same TCP/IP connection, and then failed with "socket operation on non-
3648 socket" when it tried to send the third.
3650 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3652 27. The extern definition of crypt16() in expand.c was not being excluded when
3653 the OS had its own crypt16() function.
3655 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3656 as a preferred synonym for return_size_limit, both as an option and as an
3659 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3661 30. Changed the default configuration ACL to relax the local part checking rule
3662 for addresses that are not in any local domains. For these addresses,
3663 slashes and pipe symbols are allowed within local parts, but the sequence
3664 /../ is explicitly forbidden.
3666 31. SPA server authentication was not clearing the challenge buffer before
3669 32. log_message in a "warn" ACL statement was writing to the reject log as
3670 well as to the main log, which contradicts the documentation and doesn't
3671 seem right (because no rejection is happening). So I have stopped it.
3673 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3674 However, I am unable to do any testing of this.
3676 34. Fixed an infelicity in the appendfile transport. When checking directories
3677 for a mailbox, to see if any needed to be created, it was accidentally
3678 using path names with one or more superfluous leading slashes; tracing
3679 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3681 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3682 contents are added to the log line that is written for every discarded
3683 recipient. (Previously a log_message setting was ignored.)
3685 36. The ${quote: operator now quotes the string if it is empty.
3687 37. The install script runs exim in order to find its version number. If for
3688 some reason other than non-existence or emptiness, which it checks, it
3689 could not run './exim', it was installing it with an empty version number,
3690 i.e. as "exim-". This error state is now caught, and the installation is
3693 38. An argument was missing from the function that creates an error message
3694 when Exim fails to connect to the socket for saslauthd authentication.
3695 This could cause Exim to crash, or give a corrupted message.
3697 39. Added isip, isip4, and isip6 to ${if conditions.
3699 40. The ACL variables $acl_xx are now saved with the message, and can be
3700 accessed later in routers, transports, and filters.
3702 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3703 strings in the file are not string-expanded.
3705 42. If a MAIL command specified a SIZE value that was too large to fit into an
3706 int variable, the check against message_size_limit failed. Such values are
3707 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3708 day this will have to be increased, but I don't think I want to be around
3709 when emails are that large.
3716 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3717 iconv() is not standard in FreeBSD.
3719 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3720 IPv6 enabled. The observed symptom was a segmentation fault on return from
3721 the function os_common_find_running_interfaces() in src/os.c.
3723 3. In the check_special_case() function in daemon.c I had used "errno" as an
3724 argument name, which causes warnings on some systems. This was basically a
3725 typo, since it was named "eno" in the comments!
3727 4. The code that waits for the clock to tick (at a resolution of some fraction
3728 of a second) so as to ensure message-id uniqueness was always waiting for
3729 at least one whole tick, when it could have waited for less. [This is
3730 almost certainly not relevant at current processor speeds, where it is
3731 unlikely to ever wait at all. But we try to future-proof.]
3733 5. The function that sleeps for a time interval that includes fractions of a
3734 second contained a race. It did not block SIGALRM between setting the
3735 timer, and suspending (a couple of lines later). If the interval was short
3736 and the sigsuspend() was delayed until after it had expired, the suspension
3737 never ended. On busy systems this could lead to processes getting stuck for
3740 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3741 process, before it forks any delivery processes. The open lookup caching
3742 mechanism meant that the open file or database connection was passed into
3743 the delivery process. The problem was that delivery processes always tidy
3744 up cached lookup data. This could cause a problem for the next delivery
3745 process started by the queue runner, because the external queue runner
3746 process does not know about the closure. So the next delivery process
3747 still has data in the lookup cache. In the case of a file lookup, there was
3748 no problem because closing a file descriptor in a subprocess doesn't affect
3749 the parent. However, if the lookup was caching a connection to a database,
3750 the connection was closed, and the second delivery process was likely to
3751 see errors such as "PGSQL: query failed: server closed the connection
3752 unexpectedly". The problem has been fixed by closing all cached lookups
3753 in a queue runner before running a delivery process.
3755 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3756 seem to have the "const" qualifier which it has on other OS. I've
3759 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3760 *of the same type* (client or server) with the same public name that an
3761 error should be diagnosed.
3763 9. When Exim looked up a host name for an IP address, but failed to find the
3764 original IP address when looking up the host name (a safety check), it
3765 output the message "<ip address> does not match any IP for NULL", which was
3766 confusing, to say the least. The bug was that the host name should have
3767 appeared instead of "NULL".
3769 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3770 user that is built into the binary, and the -C or -D options is used, root
3771 privilege is dropped before the configuration file is read. In addition,
3772 logging is switched to stderr instead of the normal log files. If the
3773 configuration then re-defines the Exim user, the unprivileged environment
3774 is probably not what is expected, so Exim logs a panic warning message (but
3777 However, if deliver_drop_privilege is set, the unprivileged state may well
3778 be exactly what is intended, so the warning has been cut out in that case,
3779 and Exim is allowed to try to write to its normal log files.
3785 1. smtp_return_error_details was not giving details for temporary sender
3786 or receiver verification errors.
3788 2. Diagnose a configuration error if two authenticators have the same public
3791 3. Exim used not to create the message log file for a message until the first
3792 delivery attempt. This could be confusing when incoming messages were held
3793 for policy or load reasons. The message log file is now created at the time
3794 the message is received, and an initial "Received" line is written to it.
3796 4. The automatically generated man page for command line options had a minor
3797 bug that caused no ill effects; however, a more serious problem was that
3798 the procedure for building the man page automatically didn't always
3799 operate. Consequently, release 4.20 contains an out-of-date version. This
3800 shouldn't happen again.
3802 5. When building Exim with embedded Perl support, the script that builds the
3803 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3804 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3806 6. The freeze_tell option was not being used for messages that were frozen on
3807 arrival, either by an ACL or by local_scan().
3809 7. Added the smtp_incomplete_transaction log selector.
3811 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3812 was accepting AUTH without a new EHLO.
3814 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3815 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3818 10. Logging of TCP/IP connections (when configured) now happens in the main
3819 daemon process instead of the child process, so that the TCP/IP connection
3820 count is more accurate (but it can never be perfect).
3822 11. The use of "drop" in a nested ACL was not being handled correctly in the
3823 outer ACL. Now, if condition failure induced by the nested "drop" causes
3824 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3825 or "require"), the connection is dropped.
3827 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3828 that yield "discard" can now be used with an "accept" or a "discard" verb,
3829 but an error is generated for any others (because I can't see a useful way
3830 to define what should happen).
3832 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3833 are now processed in the same way as lines in the Exim configuration file.
3834 In particular, continuation lines are supported.
3836 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3838 15. Added -ti meaning -t -i.
3840 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3841 domains, and warn by logging if others are found.
3843 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3844 returned from GIFCONF when Exim is trying to find the list of interfaces on
3845 a host. The code in os.c has been modified to copy each ifreq to an aligned
3846 structure in all cases.
3848 Also, in some cases, the returned ifreq's were being copied to a 'struct
3849 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3850 means the last couple of bytes of an IPv6 address could be chopped if the
3851 ifreq contained only a normal sockaddr (14 bytes storage).
3853 18. Named domain lists were not supported in the hosts_treat_as_local option.
3854 An entry such as +xxxx was not recognized, and was treated as a literal
3857 19. Ensure that header lines added by a DATA ACL are included in the reject log
3858 if the ACL subsequently rejects the message.
3860 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3861 MD5 (which is deprecated).
3863 21. When testing a filter file using -bf, Exim was writing a message when it
3864 took the sender from a "From " line in the message, but it was not doing so
3865 when it took $return_path from a Return-Path: header line. It now does.
3867 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3868 with a valid header line field name (a series of printing characters
3869 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3871 23. Changed "disc" in the source to "disk" to conform to the documentation and
3872 the book and for uniformity.
3874 24. Ignore Sendmail's -Ooption=value command line item.
3876 25. When execve() failed while trying to run a command in a pipe transport,
3877 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3878 could be confused with a return value of 69 from the command itself. This
3879 has been changed to 127, the value the shell returns if it is asked to run
3880 a non-existent command. The wording for the related log line suggests a
3881 non-existent command as the problem.
3883 26. If received_header_text expands to an empty string, do not add a Received:
3884 header line to the message. (Well, it adds a token one on the spool, but
3885 marks it "old" so that it doesn't get used or transmitted.)
3887 27. Installed eximstats 1.28 (addition of -nt option).
3889 28. There was no check for failure on the call to getsockname() in the daemon
3890 code. This can fail if there is a shortage of resources on the system, with
3891 ENOMEM, for example. A temporary error is now given on failure.
3893 29. Contrary to the C standard, it seems that in some environments, the
3894 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3895 program. Exim now does this explicitly; it affects the formatting of
3896 timestamps using strftime().
3898 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3899 complaints. I've now initialized all the variables, to avoid this.
3901 32. Header lines added by a system filter were not being "seen" during
3902 transport-time rewrites.
3904 33. The info_callback() function passed to OpenSSL is set up with type void
3905 (*)(SSL *, int, int), as described somewhere. However, when calling the
3906 function (actually a macro) that sets it up, the type void(*)() is
3907 expected. I've put in a cast to prevent warnings from picky compilers.
3909 34. If a DNS black list lookup found a CNAME record, but there were no A
3910 records associated with the domain it pointed at, Exim crashed.
3912 35. If a DNS black list lookup returned more than one A record, Exim ignored
3913 all but the first. It now scans all returned addresses if a particular IP
3914 value is being sought. In this situation, the contents of the
3915 $dnslist_value variable are a list of all the addresses, separated by a
3918 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3919 to accept a host name and all its aliases if the forward lookup for any of
3920 them yielded the IP address of the incoming connection. Now it accepts only
3921 those names whose forward lookup yields the correct IP address. Any other
3922 names are discarded. This closes a loophole whereby a rogue DNS
3923 administrator could create reverse DNS records to break through a
3924 wildcarded host restriction in an ACL.
3926 37. If a user filter or a system filter that ran in a subprocess used any of
3927 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3928 the wrong values were passed to the pipe command ($thisaddress had the
3929 value of $0, $0 had the value of $1, etc). This bug was introduced by
3930 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3932 38. Improved the line breaking for long SMTP error messages from ACLs.
3933 Previously, if there was no break point between 40 and 75 characters, Exim
3934 left the rest of the message alone. Two changes have been made: (a) I've
3935 reduced the minimum length to 35 characters; (b) if it can't find a break
3936 point between 35 and 75 characters, it looks ahead and uses the first one
3937 that it finds. This may give the occasional overlong line, but at least the
3938 remaining text gets split now.
3940 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3941 file descriptors might be low, and that setting 1000 would always raise it.
3942 It turns out that in some environments, the limit is already over 1000 and
3943 that lowering it causes trouble. So now Exim takes care not to decrease it.
3945 40. When delivering a message, the value of $return_path is set to $sender_
3946 address at the start of routing (routers may change the value). By an
3947 oversight, this default was not being set up when an address was tested by
3948 -bt or -bv, which affected the outcome if any router or filter referred to
3951 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3952 log only when "message" or "log_message" are set. However, if one of the
3953 conditions was an address verification, or a call to a nested ACL, the
3954 messages generated by the underlying test were being passed through. This
3955 no longer happens. The underlying message is available in $acl_verify_
3956 message for both "message" and "log_message" expansions, so it can be
3957 passed through if needed.
3959 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3960 new expansion $bh_ to give the encoded byte string without charset
3961 translation. Translation happens only if iconv() is available; HAVE_ICONV
3962 indicates this at build time. HEADERS_CHARSET gives the charset to
3963 translate to; headers_charset can change it in the configuration, and
3964 "headers charset" can change it in an individual filter file.
3966 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3967 that creates RFC 2047 encoded "words" labels them as that charset instead
3968 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3969 expansion operator; (ii) when Exim creates a From: line for a local
3970 message; (iii) when a header line is rewritten to include a "phrase" part.
3972 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3973 buggy, causing it to skip the first lines of messages whose message ID
3974 ended in 'D'. This would not have bitten before Exim release 4.14, because
3975 message IDs were unlikely to end in 'D' before then. The effect was to have
3976 incorrect size information for certain domains.
3978 45. #include "config.h" was missing at the start of the crypt16.c module. This
3979 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3982 46. If there was a timeout during a "random" callout check, Exim treated it as
3983 a failure of the random address, and carried on sending RSET and the real
3984 address. If the delay was just some slowness somewhere, the response to the
3985 original RCPT would be taken as a response to RSET and so on, causing
3986 mayhem of various kinds.
3988 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3989 when I implemented it. It didn't allow for the fact that some option values
3990 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3991 the right test for positive values.
3993 48. Domain names in DNS records are case-independent. Exim always looks them up
3994 in lower case. Some resolvers return domain names in exactly the case they
3995 appear in the zone file, that is, they may contain uppercase letters. Not
3996 all resolvers do this - some return always lower case. Exim was treating a
3997 change of case by a resolver as a change of domain, similar to a widening
3998 of a domain abbreviation. This triggered its re-routing code and so it was
3999 trying to route what was effectively the same domain again. This normally
4000 caused routing to fail (because the router wouldn't handle the domain
4001 twice). Now Exim checks for this case specially, and just changes the
4002 casing of the domain that it ultimately uses when it transmits the message
4005 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4008 50. If a filter generated a file delivery with a non-absolute name (possible if
4009 no home directory exists for the router), the forbid_file option was not
4012 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4013 the existing equality matching.
4015 52. Exim was using ints instead of ino_t variables in some places where it was
4016 dealing with inode numbers.
4018 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4019 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4020 called TMPDIR, and if it finds it is different, it changes its value.
4022 54. The smtp_printf() function is now made available to local_scan() so
4023 additional output lines can be written before returning. There is also an
4024 smtp_fflush() function to enable the detection of a dropped connection.
4025 The variables smtp_input and smtp_batched_input are exported to
4028 55. Changed the default runtime configuration: the message "Unknown user"
4029 has been removed from the ACL, and instead placed on the localuser router,
4030 using the cannot_route_message feature. This means that any verification
4031 failures that generate their own messages won't get overridden. Similarly,
4032 the "Unrouteable address" message that was in the ACL for unverifiable
4033 relay addresses has also been removed.
4035 56. Added hosts_avoid_esmtp to the smtp transport.
4037 57. The exicyclog script was not checking for the esoteric option
4038 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4039 will work only if exicyclog is run under the appropriate euid.
4041 58. Following a discussion on the list, the rules by which Exim recognises line
4042 endings on incoming messages have been changed. The -dropcr and drop_cr
4043 options are now no-ops, retained only for backwards compatibility. The
4044 following line terminators are recognized: LF CRLF CR. However, special
4045 processing applies to CR:
4047 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4048 nor a local message in the state where . is a terminator.
4050 (ii) If a bare CR is encountered in a header line, an extra space is added
4051 after the line terminator so as not to end the header. The reasoning
4052 behind this is that bare CRs in header lines are most likely either
4053 to be mistakes, or people trying to play silly games.
4055 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4056 was being incorrectly given as 18 bytes larger than it should have been.
4057 This is a VOB (very old bug).
4059 60. This may never have affected anything current, but just in case it has:
4060 When the local host is found other than at the start of a list of hosts,
4061 the local host, those with the same MX, and any that follow, are discarded.
4062 When the list in question was part of a longer list of hosts, the following
4063 hosts (not currently being processed) were also being discarded. This no
4064 longer happens. I'm not sure if this situation could ever has previously
4067 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4070 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4071 header. This is something that is recommended in a new Internet Draft, and
4072 is something that is documented as being done by Sendmail. There are two
4073 possible values. For messages generated by the autoreply transport, Exim
4076 Auto-Submitted: auto-replied
4078 whereas for all other generated messages (e.g. bounces) it adds
4080 Auto-Submitted: auto-generated
4082 63. The "personal" condition in filters now includes a test for the
4083 Auto-Submitted: header. If it contains the string "auto-" the message it
4084 not considered personal.
4086 64. Added rcpt_include_affixes as a generic transport option.
4088 65. Added queue_only_override (default true).
4090 66. Added the syslog_duplication option.
4092 67. If what should have been the first header line of a message consisted of
4093 a space followed by a colon, Exim was mis-interpreting it as a header line.
4094 It isn't of course - it is syntactically invalid and should therefore be
4095 treated as the start of the message body. The misbehaviour could have
4096 caused a number of strange effects, including loss of data in subsequent
4097 header lines, and spool format errors.
4099 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4100 client host had authenticated. This control can now be exercised by an ACL
4101 for more flexibility.
4103 69. By default, callouts do not happen when testing with -bh. There is now a
4104 variant, -bhc, which does actually run the callout code, including
4105 consulting and updating the callout cache.
4107 70. Added support for saslauthd authentication, courtesy of Alexander
4110 71. If statvfs() failed on the spool or log directories while checking their
4111 size for availability, Exim confusingly gave the error "space shortage".
4112 Furthermore, in debugging mode it crashed with a floating point exception.
4113 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4114 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4115 serious problem, Exim now writes to the main and panic logs when this
4116 happens, with details of the failure. It then refuses to accept the
4117 incoming message, giving the message "spool directory problem" or "log
4118 directory problem" with a 421 code for SMTP messages.
4120 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4121 0, 1, and 2 exist, because some OS complain for execs without them (see
4122 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4123 descriptors. However, the code omitted to check that the open succeeded,
4124 causing mysterious errors if for some reason the permissions on /dev/null
4125 got screwed. Now Exim writes a message to the main and panic logs, and
4126 bombs out if it can't open /dev/null.
4128 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4129 interact so that it is all more flexible. It is supposed to remain
4130 backwards compatible. Also added extra_local_interfaces.
4132 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4133 to bomb out with an assertion failure - to the client this appears as a
4134 connection drop. This problem occurs in the part of the code that was taken
4135 from the Samba project. Fortunately, the assertion is in a very simple
4136 function, so I have fixed this by reproducing the function inline in the
4137 one place where it is called, and arranging for authentication to fail
4138 instead of killing the process with assert().
4140 75. The SPA client code was not working when the server requested OEM rather
4141 than Unicode encoding.
4143 76. Added code to make require_files with a specific uid setting more usable in
4144 the case where statting the file as root fails - usually a non-root-mounted
4145 NFS file system. When this happens and the failure is EACCES, Exim now
4146 forks a subprocess and does the per-uid checking as the relevant uid.
4148 77. Added process_log_path.
4150 78. If log_file_path was not explicitly set, a setting of check_log_space or
4151 check_log_inodes was ignored.
4153 79. If a space check for the spool or log partitions fails, the incident is now
4154 logged. Of course, in the latter case the data may get lost...
4156 80. Added the %p formatting code to string_format() so that it can be used to
4157 print addresses in debug_print(). Adjusted all the address printing in the
4158 debugging in store.c to use %p rather than %d.
4160 81. There was a concern that a line of code in smtp_in.c could overflow a
4161 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4162 initially expressed, the concern was not well-founded, because trailing
4163 spaces are removed early. However, if the trailing spaces were followed by
4164 a NULL, they did not get removed, so the overflow was possible. Two fixes
4167 (a) I re-wrote the offending code in a cleaner fashion.
4168 (b) If an incoming SMTP command contains a NULL character, it is rejected
4171 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4172 runs initgroups(), so that if the Exim user is in any additional groups,
4173 they will be used during message reception.
4179 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob