1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.399 2006/09/26 13:51:57 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 Erfurth'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
58 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
59 hostname. This is wrong, because it relates to the incoming message (and
60 probably the interface on which it is arriving) and not to the outgoing
61 callout (which could be using a different interface). This has been
62 changed to use the value of the helo_data option from the smtp transport
63 instead - this is what is used when a message is actually being sent. If
64 there is no remote transport (possible with a router that sets up host
65 addresses), $smtp_active_hostname is used.
71 SC/01 Use a glob alias rather than an array ref in eximstats generated
72 parser. This improves both readability and performance.
74 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
75 Don't display local sender or destination tables in eximstats unless
76 there is data to show.
77 Added average volumes into the eximstats top table text output.
79 SC/03 Collect data on the number of addresses (recipients) as well
80 as the number of messages in eximstats.
82 TF/01 Correct an error in the documentation for the redirect router. Exim
83 does (usually) call initgroups() when daemonizing.
85 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
86 with consistent privilege compared to when running as a daemon.
88 TF/03 Note in the spec that $authenticated_id is not set for local
89 submissions from trusted users.
91 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
92 Thanks to Dean Brooks <dean@iglou.com> for the patch.
94 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
95 by adding some example configuration directives to the default
96 configuration file. A little bit of work is required to uncomment the
97 directives and define how usernames and passwords are checked, but
98 there is now a framework to start from.
100 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
101 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
102 without this. I don't know how relevant this is to other LDAP libraries.
104 PH/02 Add the verb name to the "unknown ACL verb" error.
106 PH/03 Magnus Holmgren's patch for filter_prepend_home.
108 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
110 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
111 directory not expanded when it should be if an expanded home directory
112 was set for the address (which is overridden by the transport).
114 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
117 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
118 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
119 because it is too late at that time, and has no effect.
121 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
122 security issue with \' (bugzilla #107). I could not use the
123 PQescapeStringConn() function, because it needs a PGconn value as one of
126 PH/08 When testing addresses using -bt, indicate those final addresses that
127 are duplicates that would not cause an additional delivery. At least one
128 person was confused, thinking that -bt output corresponded to deliveries.
129 (Suppressing duplicates isn't a good idea as you lose the information
130 about possibly different redirections that led to the duplicates.)
132 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
133 systems where poll() doesn't work, in particular OS X.
135 PH/10 Added more information to debugging output for retry time not reached.
137 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
138 operations in malware.c.
140 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
143 PH/13 If write_rejectlog was set false when logging was sent to syslog with
144 syslog_duplication set false, log lines that would normally be written
145 both the the main log and to the reject log were not written to syslog at
148 PH/14 In the default configuration, change the use of "message" in ACL warn
149 statements to "add_header".
151 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
152 not followed by a command (e.g. "seen endif").
154 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
155 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
158 PH/17 Added extra conditions to the default value of delay_warning_condition
162 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
163 { match{$h_precedence:}{(?i)bulk|list|junk} } \
164 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
167 The Auto-Submitted: and various List- headers are standardised, whereas I
168 don't think Precedence: ever was.
170 PH/18 Refactored debugging code in route_finduser() to show more information,
171 in particular, the error code if getpwnam() issues one.
173 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
174 This is apparently needed in addition to the PH/07 change above to avoid
175 any possible encoding problems.
177 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
178 but not after initializing Perl.
180 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
181 output them only if debugging. By default they are written stderr,
182 apparently, which is not desirable.
184 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
187 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
190 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
192 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
193 authenticated or an ident call has been made. Suppress the default
194 values for $authenticated_id and $authenticated_sender (but permit -oMai
195 and -oMas) when testing with -bh.
197 PH/24 Re-jigged the order of the tests in the default configuration so that the
198 tests for valid domains and recipients precede the DNS black list and CSA
199 tests, on the grounds that those ones are more expensive.
201 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
202 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
203 This bug exists in every version of Exim that I still have, right back to
206 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
207 However, an attempt to turn on case-sensitivity in a regex key by
208 including (?-i) didn't work because the subject string was already
209 lowercased, and the effects were non-intuitive. It turns out that a
210 one-line patch can be used to allow (?-i) to work as expected.
216 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
217 other effects) broke the use of negated acl sub-conditions.
219 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
222 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
223 "Deny" causes Exim to reject the incoming connection with a 554 error.
224 Unfortunately, if there is a major crisis, such as a disk failure,
225 tcp-wrappers gives "deny", whereas what one would like would be some
226 kind of temporary error. A kludge has been added to help with this.
227 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
228 554 error is used if errno is still zero or contains ENOENT (which occurs
229 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
232 PH/03 Add -lutil to the default FreeBSD LIBS setting.
234 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
235 errors. Otherwise a message that provokes a temporary error (when other
236 messages do not) can cause a whole host to time out.
238 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
239 addresses were routed directly to files or pipes from a redirect router.
240 File deliveries just didn't batch; pipe deliveries might have suffered
243 PH/06 A failure to get a lock for a hints database would erroneously always say
244 "Failed to get write lock", even when it was really a read lock.
246 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
247 of 0600. This has been changed to use the value of the lockfile_mode
248 option (which defaults to 0600).
250 PH/08 Applied small patch from the Sieve maintainer.
252 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
253 folder from quota calculations, a direct delivery into this folder messed
254 up the contents of the maildirsize file. This was because the regex was
255 used only to exclude .Trash (or whatever) when the size of the mailbox
256 was calculated. There was no check that a delivery was happening into an
257 excluded directory. This bug has been fixed by ignoring all quota
258 processing for deliveries into excluded directories.
260 PH/10 Added the maildirfolder_create_regex option to appendfile.
266 PH/01 The code for finding all the local interface addresses on a FreeBSD
267 system running IPv6 was broken. This may well have applied to all BSD
268 systems, as well as to others that have similar system calls. The broken
269 code found IPv4 interfaces correctly, but gave incorrect values for the
270 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
271 that it would not match correctly against @[] and not recognize the IPv6
274 PH/02 The ipliteral router was not recognizing addresses of the form user@
275 [ipv6:....] because it didn't know about the "ipv6:" prefix.
277 PH/03 Added disable_ipv6.
279 PH/04 Changed $reply_address to use the raw form of the headers instead of the
280 decoded form, because it is most often used to construct To: headers
281 lines in autoreplies, and the decoded form may well be syntactically
282 invalid. However, $reply_address has leading white space removed, and all
283 newlines turned into spaces so that the autoreply transport does not
286 PH/05 If group was specified without a user on a router, and no group or user
287 was specified on a transport, the group from the router was ignored.
289 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
290 for visible compile-time settings that can be used to change these
291 numbers, for those that want even more. Backwards compatibility with old
292 spool files has been maintained. However, going back to a previous Exim
293 release will lost any variables that are in spool files.
295 PH/07 Two small changes when running in the test harness: increase delay when
296 passing a TCP/IP connection to a new process, in case the original
297 process has to generate a bounce, and remove special handling of
298 127.0.0.2 (sic), which is no longer necessary.
300 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
301 be the same on different OS.
303 PH/09 Moved a debug statement in filter processing to avoid a race problem when
306 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
307 whether --show-vars was specified or not
309 JJ/02 exipick: Added support for new ACL variable spool format introduced
312 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
313 syntactically invalid From: or Reply-to: line, and a filter used this to
314 generate an autoreply, and therefore failed to obtain an address for the
315 autoreply, Exim could try to deliver to a non-existent relative file
316 name, causing unrelated and misleading errors. What now happens is that
317 it logs this as a hard delivery error, but does not attempt to create a
320 PH/11 The exinext utility has a -C option for testing purposes, but although
321 the given file was scanned by exinext itself; it wasn't being passed on
322 when Exim was called.
324 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
325 an end-of-file indication when reading a command response.
327 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
328 compiled. In many other places in Exim, IPv6 addresses are always
329 recognized, so I have changed this. It also means that IPv4 domain
330 literals of the form [IPV4:n.n.n.n] are now always recognized.
332 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
333 used if the router is not running as root, for example, when verifying at
334 ACL time, or when using -bh. The debugging output from this situation was
335 non-existent - all you got was a failure to exec. I have made two
338 (a) Failures to set uid/gid, the current directory, or a process leader
339 in a subprocess such as that created by queryprogram now generate
340 suitable debugging ouput when -d is set.
342 (b) The queryprogram router detects when it is not running as root,
343 outputs suitable debugging information if -d is set, and then runs
344 the subprocess without attempting to change uid/gid.
346 PH/15 Minor change to Makefile for building test_host (undocumented testing
349 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
350 additional section of a DNS packet that returns MX or SRV records.
351 Instead, it always explicitly searches for A/AAAA records. This avoids
352 major problems that occur when a DNS server includes only records of one
353 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
354 fixed another bug: if SRV records were looked up and the corresponding
355 address records were *not* found in the additional section, the port
356 values from the SRV records were lost.
358 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
359 using the correct key (the original address) when searching the retry
360 rules in order to find which one to use for generating the retry hint.
362 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
363 adding the default one. Similarly, if it contains a Reply-To: header, the
364 errors_reply_to option, if set, is not used.
366 PH/19 When calculating a retry time, Exim used to measure the "time since
367 failure" by looking at the "first failed" field in the retry record. Now
368 it does not use this if it is later than than the arrival time of the
369 message. Instead it uses the arrival time. This makes for better
370 behaviour in cases where some deliveries succeed, thus re-setting the
371 "first failed" field. An example is a quota failure for a huge message
372 when small messages continue to be delivered. Without this change, the
373 "time since failure" will always be short, possible causing more frequent
374 delivery attempts for the huge message than are intended.
375 [Note: This change was subsequently modified - see PH/04 for 4.62.]
377 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
378 $1, $2, $3) because the numerical variables can be reset during some
379 expansion items (e.g. "match"), thereby losing the authentication data.
381 PH/21 Make -bV show the size of off_t variables so that the test suite can
382 decide whether to run tests for quotas > 2G.
384 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
385 mailbox_size, and mailbox_filecount in the appendfile transport. If a
386 filecount value is greater than 2G or if a quota value is greater than 2G
387 on a system where the size of off_t is not greater than 4, a panic error
390 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
391 never match. The debug and -bh output now contains an explicit error
392 message indicating a malformed IPv4 address or mask.
394 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
395 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
398 PH/25 Do not write to syslog when running in the test harness. The only
399 occasion when this arises is a failure to open the main or panic logs
400 (for which there is an explicit test).
402 PH/26 Added the /no_tell option to "control=freeze".
404 PH/27 If a host name lookup failed very early in a connection, for example, if
405 the IP address matched host_lookup and the reverse lookup yielded a name
406 that did not have a forward lookup, an error message of the form "no IP
407 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
408 could be logged. Now it outputs the IP address instead of "NULL".
410 PH/28 An enabling patch from MH: add new function child_open_exim2() which
411 allows the sender and the authenticated sender to be set when
412 submitting a message from within Exim. Since child_open_exim() is
413 documented for local_scan(), the new function should be too.
415 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
416 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
417 results in an empty string is now treated as unset.
419 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
421 PH/31 Added sender_verify_fail logging option.
423 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
424 needed by RFC 822 but not by RFC 2822 was commented out. I have now
425 tidied the source and removed it altogether.
427 PH/33 When a queue run was abandoned because the load average was too high, a
428 log line was always written; now it is written only if the queue_run log
429 selector is set. In addition, the log line for abandonment now contains
430 information about the queue run such as the pid. This is always present
431 in "start" and "stop" lines but was omitted from the "abandon" line.
433 PH/34 Omit spaces between a header name and the colon in the error message that
434 is given when verify = headers_syntax fails (if there are lots of them,
435 the message gets confusing).
437 PH/35 Change the default for dns_check_names_pattern to allow slashes within
438 names, as there are now some PTR records that contain slashes. This check
439 is only to protect against broken name servers that fall over on strange
440 characters, so the fact that it applies to all lookups doesn't matter.
442 PH/36 Now that the new test suite is complete, we can remove some of the
443 special code in Exim that was needed for the old test suite. For example,
444 sorting DNS records because real resolvers return them in an arbitrary
445 order. The new test suite's fake resolver always returns records in the
448 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
449 bounces) except when queue_only is set, to avoid logging races between
450 the different processes.
452 PH/38 Panic-die if .include specifies a non-absolute path.
454 PH/39 A tweak to the "H" retry rule from its user.
456 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
457 a label. They prevented compilation on older perls.
459 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
460 a warning to be raised on newish perls.
462 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
463 on queue. Changes to match documented behaviour of showing count of
464 messages matching specified criteria.
466 PH/40 Changed the default ident timeout from 30s to 5s.
468 PH/41 Added support for the use of login_cap features, on those BSD systems
469 that have them, for controlling the resources used by pipe deliveries.
471 PH/42 The content-scanning code uses fopen() to create files in which to put
472 message data. Previously it was not paying any attention to the mode of
473 the files. Exim runs with umask(0) because the rest of the code creates
474 files with open(), and sets the required mode explicitly. Thus, these
475 files were ending up world-writeable. This was not a big issue, because,
476 being within the spool directory, they were not world-accessible. I have
477 created a function called modefopen, which takes an additional mode
478 argument. It sets umask(777), creates the file, chmods it to the required
479 mode, then resets the umask. All the relevant calls to fopen() in the
480 content scanning code have been changed to use this function.
482 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
483 to 24 hours. This avoids potential overflow problems when processing G
484 and H retry rules. I suspect nobody ever tinkers with this value.
486 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
488 PH/45 When the plaintext authenticator is running as a client, the server's
489 challenges are checked to ensure they are valid base64 strings. By
490 default, the authentication attempt is cancelled if an invalid string is
491 received. Setting client_ignore_invalid_base64 true ignores these errors.
492 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
493 they are received. Thus, the responses can be made to depend on the
494 challenges. If an invalid string is ignored, an empty string is placed in
497 PH/46 Messages that are created by the autoreply transport now contains a
498 References: header, in accordance with RFCs 2822 and 3834.
500 PH/47 Added authenticated_sender_force to the smtp transport.
502 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
504 PH/49 Installed latest patch from the Sieve maintainer.
506 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
507 was also set, the appendfile transport was unnecessarily scanning a
508 directory of message files (e.g. for maildir delivery) to find the count
509 of files (along with the size), even though it did not need this
510 information. It now does the scan only if it needs to find either the
511 size of the count of files.
513 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
515 PH/52 Two bugs concerned with error handling when the smtp transport is
518 (i) Exim was not creating retry information for temporary errors given
519 for individual recipients after the DATA command when the smtp transport
520 was used in LMTP mode. This meant that they could be retried too
521 frequently, and not timed out correctly.
523 (ii) Exim was setting the flag that allows error details to be returned
524 for LMTP errors on RCPT commands, but not for LMTP errors for individual
525 recipients that were returned after the DATA command.
527 PH/53 This is related to PH/52, but is more general: for any failing address,
528 when detailed error information was permitted to be returned to the
529 sender, but the error was temporary, then after the final timeout, only
530 "retry timeout exceeded" was returned. Now it returns the full error as
531 well as "retry timeout exceeded".
533 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
534 do this, and (what is worse) MTAs that accept it.
536 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
537 will now be deprecated.
539 PH/56 New os.c-cygwin from the Cygwin maintainer.
541 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
542 formats (previously only available in exim formats via -bpr, -bpru,
543 and -bpra. Now also available in native and exiqgrep formats)
545 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
546 with very large, slow to parse queues
548 JJ/08 exipick: added ! as generic prefix to negate any criteria format
550 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
552 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
553 responses to authentication challenges, though it was showing the
554 challenges; (ii) I've removed the CR characters from the debug output for
557 PH/58 Allow for the insertion of a newline as well as a space when a string
558 is turned into more than one encoded-word during RFC 2047 encoding. The
559 Sieve code now uses this.
561 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
562 data_4xx, lost_connection, tls_required.
564 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
565 message was being sent as an SMTP response.
567 PH/61 Add -l and -k options to exicyclog.
569 PH/62 When verifying, if an address was redirected to one new address, so that
570 verification continued, and the new address failed or deferred after
571 having set something in $address_data, the value of $address_data was not
572 passed back to the ACL. This was different to the case when no
573 redirection occurred. The value is now passed back in both cases.
575 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
576 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
577 use login_cap.h, so on its own it isn't the distinguishing feature. The
578 new name refers directly to the setclassresources() function.
580 PH/65 Added configuration files for NetBSD3.
582 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
584 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
585 is preferred over IPv4.
587 PH/68 The bounce_return_message and bounce_return_body options were not being
588 honoured for bounces generated during the reception of non-SMTP messages.
589 In particular, this applied to messages rejected by the ACL. This bug has
590 been fixed. However, if bounce_return_message is true and bounce_return_
591 body is false, the headers that are returned for a non-SMTP message
592 include only those that have been read before the error was detected.
593 (In the case of an ACL rejection, they have all been read.)
595 PH/69 The HTML version of the specification is now built in a directory called
596 spec_html instead of spec.html, because the latter looks like a path with
597 a MIME-type, and this confuses some software.
599 PH/70 Catch two compiler warnings in sieve.c.
601 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
602 function verify_get_ident() calls ip_connect() to connect a socket, but
603 if the "connect()" function timed out, ip_connect() used to close the
604 socket. However, verify_get_ident() also closes the socket later, and in
605 between Exim writes to the log, which may get opened at this point. When
606 the socket was closed in ip_connect(), the log could get the same file
607 descriptor number as the socket. This naturally causes chaos. The fix is
608 not to close the socket in ip_connect(); the socket should be closed by
609 the function that creates it. There was only one place in the code where
610 this was missing, in the iplookup router, which I don't think anybody now
611 uses, but I've fixed it anyway.
613 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
614 well as to direct DNS lookups. Otherwise the handling of names in host
615 lists is inconsistent and therefore confusing.
621 PH/01 Two changes to the default runtime configuration:
623 (1) Move the checks for relay_from_hosts and authenticated clients from
624 after to before the (commented out) DNS black list checks.
626 (2) Add control=submission to the relay_from_hosts and authenticated
627 clients checks, on the grounds that messages accepted by these
628 statements are most likely to be submissions.
630 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
632 (1) Generate an error if the third argument for the ${prvs expansion is
635 (2) Treat a missing third argument of ${prvscheck as if it were an empty
638 (3) Reset the variables that are obtained from the first argument of
639 ${prvscheck and used in the second argument before leaving the code,
640 because their memory is reclaimed, so using them afterwards may do
643 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
644 one (it's much easier than Tom thought :-).
646 (5) Because of (4), we can now allow for the use of $prvscheck_result
647 inside the third argument.
649 PH/03 For some reason, the default setting of PATH when running a command from
650 a pipe transport was just "/usr/bin". I have changed it to
653 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
654 anything to be listed in the output from -bV.
656 PH/05 When a filter generated an autoreply, the entire To: header line was
657 quoted in the delivery log line, like this:
659 => >A.N.Other <ano@some.domain> <original@ddress> ...
661 This has been changed so that it extracts the operative address. There
662 may be more than one such address. If so, they are comma-separated, like
665 => >ano@some.domain,ona@other.domain <original@ddress> ...
667 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
668 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
669 not being looked up in the rDNS to get a host name, Exim was showing the
670 IP address twice in Received: lines, even though the IP addresses were
671 identical. For example:
673 Received: from [1.2.3.4] (helo=[1.2.3.4])
675 However, if the real host name was known, it was omitting the HELO data
676 if it matched the actual IP address. This has been tidied up so that it
677 doesn't show the same IP address twice.
679 PH/07 When both +timestamp and +memory debugging was on, the value given by
680 $tod_xxx expansions could be wrong, because the tod_stamp() function was
681 called by the debug printing, thereby overwriting the timestamp buffer.
682 Debugging no longer uses the tod_stamp() function when +timestamp is set.
684 PH/08 When the original message was included in an autoreply transport, it
685 always said "this is a copy of the message, including all the headers",
686 even if body_only or headers_only was set. It now gives an appropriate
689 PH/09 Applied a patch from the Sieve maintainer which:
691 o fixes some comments
692 o adds the (disabled) notify extension core
693 o adds some debug output for the result of if/elsif tests
694 o points to the current vacation draft in the documentation
695 and documents the missing references header update
699 o fixes a bug in processing the envelope test (when testing
700 multiple envelope elements, the last element determinted the
703 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
704 Electronic Mail") by including:
706 Auto-submitted: auto-generated
708 in the messages that it generates (bounce messages and others, such as
709 warnings). In the case of bounce messages for non-SMTP mesages, there was
710 also a typo: it was using "Auto_submitted" (underscore instead of
711 hyphen). Since every message generated by Exim is necessarily in response
712 to another message, thes have all been changed to:
714 Auto-Submitted: auto-replied
716 in accordance with these statements in the RFC:
718 The auto-replied keyword:
720 - SHOULD be used on messages sent in direct response to another
721 message by an automatic process,
723 - MUST NOT be used on manually-generated messages,
725 - MAY be used on Delivery Status Notifications (DSNs) and Message
726 Disposition Notifications (MDNs),
728 - MUST NOT be used on messages generated by automatic or periodic
729 processes, except for messages which are automatic responses to
732 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
733 to the default Received: header definition.
735 PH/12 Added log selector acl_warn_skipped (default on).
737 PH/13 After a successful wildlsearch lookup, discard the values of numeric
738 variables because (a) they are in the wrong storage pool and (b) even if
739 they were copied, it wouldn't work properly because of the caching.
741 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
742 checking when decoding. Apparently there are clients that generate
743 overlong encoded strings. Why am I not surprised?
745 PH/15 If the first argument of "${if match_address" was not empty, but did not
746 contain an "@" character, Exim crashed. Now it writes a panic log message
747 and treats the condition as false.
749 PH/16 In autoreply, treat an empty string for "once" the same as unset.
751 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
752 extension "envelope-auth". The code is finished and in agreement with
753 other implementations, but there is no documentation so far and in fact,
754 nobody wrote the draft yet. This extension is currently #undef'ed, thus
755 not changing the active code.
757 Print executed "if" and "elsif" statements when debugging is used. This
758 helps a great deal to understand what a filter does.
760 Document more things not specified clearly in RFC3028. I had all this
761 sorted out, when out of a sudden new issues came to my mind. Oops."
763 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
766 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
767 canonical form (as documented). However, after a host name lookup from
768 the IP address, check_host() was doing a simple string comparison with
769 addresses acquired from the DNS when checking that the found name did
770 have the original IP as one of its addresses. Since any found IPv6
771 addresses are likely to be in abbreviated form, the comparison could
772 fail. Luckily, there already exists a function for doing the comparison
773 by converting both addresses to binary, so now that is used instead of
776 PH/20 There was another similar case to PH/19, when a complete host name was
777 given in a host list; looking up its IP address could give an abbreviated
778 form, whereas the current host's name might or might not be abbreviated.
779 The same fix has been applied.
785 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
786 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
789 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
790 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
792 PH/03 Typo: missing ".o" in src/pcre/Makefile.
794 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
795 header line, restrict the check to what is listed in RFCs 2369 and 2929.
796 Also, for "Auto-Submitted", treat anything other than "no" as
797 non-personal, in accordance with RFC 3834. (Previously it treated
798 anything starting "auto-" as non-personal.)
800 TF/01 The control=submission/name=... option had a problem with syntax
801 errors if the name included a slash character. The /name= option
802 now slurps the rest of the string, so it can include any characters
803 but it must come last in the list of options (after /sender_retain
806 PH/05 Some modifications to the interface to the fake nameserver for the new
814 TK/01 Added the "success_on_redirect" address verification option. See
815 NewStuff for rationale and an example.
817 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
819 PH/02 Patch to exigrep to allow it to work on syslog lines.
821 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
822 fread() to skip over the body file's header line, because in Cygwin the
823 header line is locked and is inaccessible.
825 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
826 co-exist for some time) to make it clear that it is the Exim ID that is
827 referenced, not the Message-ID: header line.
829 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
830 string_format() function, because snprintf() does not exist on all
833 PH/06 The use of forbid_filter_existstest now also locks out the use of the
834 ${stat: expansion item.
836 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
837 protocol synchronization error", to keep the pedants happy.
839 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
840 well as for IRIX systems, when gcc is being used. See the host.c source
843 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
845 PH/10 Named domain lists were not working if used in a queue_smtp_domains
848 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
849 transport and to the smtp transport in LMTP mode.
851 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
853 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
854 run a filter in a subprocess. This could lead to confusion in subsequent
855 lookups in the parent process. There should also be a search_tidyup() at
856 the end of the subprocess.
858 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
859 only if the host matched helo_try_verify_hosts, which caused the
860 verification to occur when the EHLO/HELO command was issued. The ACL just
861 tested the remembered result. Now, if a previous verification attempt has
862 not happened, "verify = helo" does it there and then.
864 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
866 TK/03 Fix log output including CR from clamd.
868 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
869 exist provoked a memory error which could cause a segfault.
871 PH/15 Installed PCRE 6.2
873 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
875 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
876 of the problem. Specifically, suggested +O2 rather than +O1 for the
879 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
881 PH/20 If a delivery was routed to a non-standard port by means of an SRV
882 record, the port was not correctly logged when the outgoing_port log
883 selector was set (it logged the transort's default port).
885 PH/21 Added support for host-specific ports to manualroute, queryprogram,
886 fallback_hosts, and "hosts" in the smtp transport.
888 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
889 host errors such as "Connection refused".
891 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
892 authentication with radiusclient 0.4.9:
894 - Error returned from rc_read_config was caught wrongly
895 - Username/password not passed on to radius server due to wrong length.
897 The presumption is that some radiusclient API changes for 4.51/PH/17
898 were not taken care of correctly. The code is still untested by me (my
899 Linux distribution still has 0.3.2 of radiusclient), but it was
900 contributed by a Radius user.
902 PH/24 When doing a callout, the value of $domain wasn't set correctly when
903 expanding the "port" option of the smtp transport.
905 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
906 while reading a MIME header. Thanks to Tom Hughes for a patch.
908 PH/24 Include config.h inside local_scan.h so that configuration settings are
911 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
912 that in an ACL for RCPT (for example), you can examine exactly what was
915 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
916 commands, but it was not correctly comparing the address with the actual
917 client host address. Thus, it would show the EHLO address in Received:
918 header lines when this was not necessary.
920 PH/27 Added the % operator to ${eval:}.
922 PH/28 Exim tries to create and chdir to its spool directory when it starts;
923 it should be ignoring failures (because with -C, for example, it has lost
924 privilege). It wasn't ignoring creation failures other than "already
927 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
930 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
931 because an input line was too long, either on its own, or by virtue of
932 too many continuations, the temporary file was not being removed, and the
933 return code was incorrect.
935 PH/31 Missing "BOOL" in function definition in filtertest.c.
937 PH/32 Applied Sieve patches from the maintainer.
939 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
941 PH/33 Added "verify = not_blind".
943 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
944 Local/Makefile (with some defaults set). These are used in built scripts
945 such as exicyclog, but they have never been used in the exim_install
946 script (though there are many overriding facilities there). I have
947 arranged that the exim_install script now takes note of these two
950 PH/35 Installed configuration files for Dragonfly.
952 PH/36 When a locally submitted message by a trusted user did not contain a
953 From: header, and the sender address was obtained from -f or from an SMTP
954 MAIL command, and the trusted user did not use -F to supply a sender
955 name, $originator_name was incorrectly used when constructing a From:
956 header. Furthermore, $originator_name was used for submission mode
957 messages from external hosts without From: headers in a similar way,
958 which is clearly wrong.
960 PH/37 Added control=suppress_local_fixups.
962 PH/38 When log_selector = +received_sender was set, and the addition of the
963 sender made the log line's construction buffer exactly full, or one byte
964 less than full, an overflow happened when the terminating "\n" was
967 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
968 when the result of a list match is failure because a DNS lookup failed.
970 PH/40 RM_COMMAND is now used in the building process.
972 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
973 the "build-* directories that it finds.
975 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
976 address in a domain literal was a prefix of an interface address.
978 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
979 when verifying a sender address, unless rewrite_headers is false.
981 PH/44 Wrote a long comment about why errors_to addresses are verified as
982 recipients, not senders.
984 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
985 the ratelimit ACL was added.
987 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
989 PH/46 Added extra information about PostgreSQL errors to the error string.
991 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
992 avoiding the need to install special zones in a real server. This is
993 backwards compatible; if it can't find the fake resolver, it drops back.
994 Thus, both old and new test suites can be run.
996 TF/02 Added util/ratelimit.pl
998 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
999 clock is set back in time.
1001 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1002 Candler <B.Candler@pobox.com>.
1004 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1005 OK for addresses that are the result of redirections.
1007 PH/48 A number of further additions for the benefit of the new test suite,
1008 including a fake gethostbyname() that interfaces to the fake DNS resolver
1011 TF/06 The fix for widen_domains has also been applied to qualify_single and
1012 search_parents which are the other dnslookup options that can cause
1015 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1018 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1020 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1021 service" if it encountered an MX record with an empty target hostname.
1022 The message is now "an MX or SRV record indicated no SMTP service".
1024 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1025 if the DNS of the sending site is misconfigured. This is quite a
1026 common situation. This change restores the behaviour of treating a
1027 helo verification defer as a failure.
1029 PH/51 If self=fail was set on a router, the bounce message did not include the
1030 actual error message.
1036 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1038 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1039 command itself ended in error, the underlying message about the transport
1040 filter timeout was being overwritten with the pipe command error. Now the
1041 underlying error message should be appended to the second error message.
1043 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1045 PH/02 Reduce the amount of output that "make" produces by default. Full output
1046 can still be requested.
1048 PH/03 The warning log line about a condition test deferring for a "warn" verb
1049 was being output only once per connection, rather than after each
1050 occurrence (because it was using the same function as for successful
1051 "warn" verbs). This seems wrong, so I have changed it.
1053 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1054 it should not have, which might have caused a crash in the right
1055 circumstances, but probably never did.
1057 PH/04 Installed a modified version of Tony Finch's patch to make submission
1058 mode fix the return path as well as the Sender: header line, and to
1059 add a /name= option so that you can make the user's friendly name appear
1062 TF/03 Added the control = fakedefer ACL modifier.
1064 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1065 Mark Lowes for thorough testing.
1067 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1069 TK/03 Merged latest SRS patch from Miles Wilton.
1071 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1072 with the definition in sysexits.h (which is #included earlier).
1073 Fortunately, Exim does not actually use EX_OK. The code used to try to
1074 preserve the sysexits.h value, by assumimg that macro definitions were
1075 scanned for macro replacements. I have been disabused of this notion,
1076 so now the code just undefines EX_OK before #including unistd.h.
1078 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1079 in the smtp transport. When a block could not be written in a single
1080 write() function, the timeout was being re-applied to each part-write.
1081 This seems wrong - if the receiver was accepting one byte at a time it
1082 would take for ever. The timeout is now adjusted when this happens. It
1083 doesn't have to be particularly precise.
1085 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1086 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1088 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1089 without a domain if the check to <postmaster@domain> fails.
1091 SC/01 Eximstats: added -xls and the ability to specify output files
1092 (patch written by Frank Heydlauf).
1094 SC/02 Eximstats: use FileHandles for outputing results.
1096 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1098 SC/04 Eximstats: fixed display of large numbers with -nvr option
1100 SC/05 Eximstats: fixed merging of reports with empty tables.
1102 SC/06 Eximstats: added the -include_original_destination flag
1104 SC/07 Eximstats: removed tabs and trailing whitespace.
1106 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1108 TK/06 MBOX spool code: Add real "From " MBOX separator line
1109 so the .eml file is really in mbox format (even though
1110 most programs do not really care). Patch from Alex Miller.
1112 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1113 The latter is generated from $received_to and is only set if the
1114 message has one envelope recipient. SA can use these headers,
1115 obviously out-of-the-box. Patch from Alex Miller.
1117 PH/08 The ${def test on a variable was returning false if the variable's
1118 value was "0", contrary to what the specification has always said!
1119 The result should be true unless the variable is empty.
1121 PH/09 The syntax error of a character other than { following "${if
1122 def:variable_name" (after optional whitespace) was not being diagnosed.
1123 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1124 accidental colon was present, for example, could give incorrect results.
1126 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1127 result is used (not including appendfile, where other changes are about
1130 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1131 This involved changing a lot of size variables from int to off_t. It
1132 should work with maildirs and everything.
1134 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1135 spamd dying while we are connected to it.
1137 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1140 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1141 function for BDB 4.3.
1143 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1145 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1146 the sources for PCRE have been re-arranged and more files are now
1149 PH/15 The code I had for printing potentially long long variables in PH/11
1150 above was not the best (it lost precision). The length of off_t variables
1151 is now inspected at build time, and an appropriate printing format (%ld
1152 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1153 to be "long long int" or "long int". This is needed for the internal
1154 formatting function string_vformat().
1156 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1157 the configuration file to be ":syslog", then the script "guesses" where
1158 the logs files are, rather than using the compiled in default. In our
1159 case the guess is not the same as the compiled default, so the script
1160 suddenly stopped working when I started to use syslog. The patch checks
1161 to see if log_file_path is "". If so, it attempts to read it from exim
1162 with no configuration file to get the compiled in version, before it
1163 falls back to the previous guessing code."
1165 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1166 implementing BATV in an Exim configuration. See NewStuff for the gory
1169 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1170 Makefile that are specific to HP-UX.
1172 PH/18 If the "use_postmaster" option was set for a recipient callout together
1173 with the "random" option, the postmaster address was used as the MAIL
1174 FROM address for the random test, but not for the subsequent recipient
1175 test. It is now used for both.
1177 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1178 patch removes a few documentation additions to RFC 3028, because the
1179 latest draft now contains them. It adds the new en;ascii-case comparator
1180 and a new error check for 8bit text in MIME parts. Comparator and
1181 require names are now matched exactly. I enabled the subaddress
1182 extension, but it is not well tested yet (read: it works for me)."
1184 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1185 rework some of the code of TK/09 above to avoid the hardwired use of
1186 "%lld" and "long long". Replaced the call to snprintf() with a call to
1189 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1190 records point to non-existent hosts", "retry timeout exceeded", and
1191 "retry time not reached for any host after a long failure period".
1193 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1194 experimental DomainKeys support:
1196 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1197 (2) On an error such as an illegally used "control", the wrong name for
1198 the control was given.
1200 These problems did NOT occur unless DomainKeys support was compiled.
1202 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1204 PH/24 Added ${if match_ip condition.
1206 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1207 databases so that it will be absolutely obvious if a crash occurs in the
1208 DB library. This is a regular occurrence (often caused by mis-matched
1211 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1212 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1213 picked up on a user's system that detects such things. There doesn't seem
1214 to be a gcc warning option for this - only an attribute that has to be
1215 put on the function's prototype. It seems that in Fedora Core 4 they have
1216 set this on a number of new functions. No doubt there will be more in due
1219 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1220 specify a transport. However, if an address that was verified by such a
1221 router was the subject of a callout, Exim crashed because it tried to
1222 read the rcpt_include_affixes from the non-existent transport. Now it
1223 just assumes that the setting of that option is false. This bug was
1224 introduced by 4.51/PH/31.
1226 PH/28 Changed -d+all to exclude +memory, because that information is very
1227 rarely of interest, but it makes the output a lot bigger. People tend to
1228 do -d+all out of habit.
1230 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1231 code in os-type was giving problems when libc.so lives in lib64, like on
1234 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1235 aren't the modern standard, and it seems that some systems' include files
1236 don't always have them. Exim was already checking for some of the newer
1237 ones like T_AAAA, and defining it itself. I've added checks for all the
1238 record types that Exim uses.
1240 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1241 not automatically generating a new one, as it is supposed to. This
1242 prevented TLS from working. If the file did exist, but contained invalid
1243 data, a new version was generated, as expected. It was only the case of a
1244 non-existent file that was broken.
1246 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1247 with a change in libdomainkeys > 0.64.
1249 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1250 from DNS. If the selector record carries the flag, it now has
1251 precedence over the domain-wide flag.
1253 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1255 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1256 the use of a port name that isn't defined in /etc/services) occurred, the
1257 message was deferred as in a normal delivery, and thus remained on the
1258 spool, instead of being failed because of the mua_wrapper setting. This
1259 is now fixed, and I tidied up some of the mua_wrapper messages at the
1262 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1263 the messages in a hash of arrays rather than using individual hashes.
1264 This is a bit cleaner and results in dramatic memory savings, albeit
1265 at a slight CPU cost.
1267 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1268 as requested by Marc Sherman.
1270 SC/10 Eximstats: added histograms for user specified patterns as requested
1273 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1275 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1276 fopen() in the content-scanning modules that did not already have it.
1282 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1283 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1285 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1287 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1289 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1290 to test Sieve filters that use "vacation".
1292 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1293 that changes the way the GnuTLS parameters are stored in the cache file.
1294 The new format can be generated externally. For backward compatibility,
1295 if the data in the cache doesn't make sense, Exim assumes it has read an
1296 old-format file, and it generates new data and writes a new file. This
1297 means that you can't go back to an older release without removing the
1300 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1301 work if there are any delivery delays because "one_time" forces the
1302 parent to be marked "delivered", so its unseen clone is never tried
1303 again. For this reason, Exim now forbids the simultaneous setting of
1306 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1307 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1308 entry if you want to know the details. The fix, however, neglected to
1309 consider the case when local delivery batching is involved. The test for
1310 "previously delivered" was not happening when checking to see if an
1311 address could be batched with a previous (undelivered) one; under
1312 certain circumstances this could lead to multiple deliveries to the same
1315 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1316 in its include files, and this causes problems building Exim.
1318 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1319 header_syntax) but Exim was just ignoring anything given after a slash.
1320 In particular, this caused confusion with an attempt to use "verify =
1321 reverse_host_lookup/defer_ok". An error is now given when options are
1322 supplied for verify items that do not have them. (Maybe reverse_host_
1323 lookup should have a defer_ok option, but that's a different point.)
1325 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1326 defined by RFC 821) to 2048, because there were problems with some AUTH
1327 commands, and RFC 1869 says the size should be increased for extended
1328 SMTP commands that take arguments.
1330 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1333 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1334 "unknown" error; now it says that the functionality isn't in the binary.
1336 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1337 an address' error message when a string expansion fails (syntax or
1338 whatever). Otherwise the password may appear in the log. Following change
1339 PH/42 below, there is no longer a chance of it appearing in a bounce
1342 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1344 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1345 of its addresses was ever tried. (Bugzilla bug #2.)
1347 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1348 the result incorrectly in the debug output. (It correctly added a newline
1349 to what was transported.)
1351 TF/01 Added $received_time.
1353 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1354 commented out examples of how to interface to a virus scanner and to
1355 SpamAssassin. Also added commented examples of av_scanner and
1356 spamd_address settings.
1358 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1359 and controls are allowed in which ACLs. There were a couple of minor
1360 errors. Some of the entries in the conditions table (which is a table of
1361 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1362 negation of where the condition IS allowed.
1364 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1366 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1367 header file does not have a version number, so I've had to invent a new
1368 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1369 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1370 radiusclient), but it was contributed by a Radius user.
1372 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1373 files or directories, for OpenSSL.
1375 PH/19 When an Exim process that is running as root has to create an Exim log
1376 file, it does so in a subprocess that runs as exim:exim so as to get the
1377 ownership right at creation (otherwise, other Exim processes might see
1378 the file with the wrong ownership). There was no test for failure of this
1379 fork() call, which would lead to the process getting stuck as it waited
1380 for a non-existent subprocess. Forks do occasionally fail when resources
1381 run out. I reviewed all the other calls to fork(); they all seem to check
1384 PH/20 When checking for unexpected SMTP input at connect time (before writing
1385 the banner), Exim was not dealing correctly with a non-positive return
1386 from the read() function. If the client had disconnected by this time,
1387 the result was a log entry for a synchronization error with an empty
1388 string after "input=" when read() returned zero. If read() returned -1
1389 (an event I could not check), uninitialized data bytes were printed.
1390 There were reports of junk text (parts of files, etc) appearing after
1393 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1395 PH/22 Added support for macro redefinition, and (re)definition in between
1396 driver and ACL definitions.
1398 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1399 forgetting to use the resulting value; it was using the unexpanded value.
1401 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1402 hadn't been configured. The fix is from Juergen Kreileder, who
1403 understands it better than I do:
1405 "Here's what I see happening with three configured cyrus_sasl
1406 authenticators configured (plain, login, cram-md5):
1408 On startup auth_cyrus_sasl_init() gets called for each of these.
1409 This means three calls to sasl_listmech() without a specified mech_list.
1410 => SASL tests which mechs of all available mechs actually work
1411 => three warnings about OTP not working
1412 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1414 With the patch, sasl_listmech() also gets called three times. But now
1415 SASL's mech_list option is set to the server_mech specified in the the
1416 authenticator. Or in other words, the answer from sasl_listmech()
1417 gets limited to just the mech you're testing for (which is different
1419 => the return list contains just 'plain' or 'login', 'cram-md5' or
1420 nothing depending on the value of ob->server_mech.
1422 I've just tested the patch: Authentication still works fine,
1423 unavailable mechs specified in the exim configuration are still
1424 caught, and the auth.log warnings about OTP are gone."
1426 PH/25 When debugging is enabled, the contents of the command line are added
1427 to the debugging output, even when log_selector=+arguments is not
1430 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1431 answer is "GNU", and only if the return is "GNU/something" is the answer
1434 PH/27 $acl_verify_message is now set immediately after the failure of a
1435 verification in an ACL, and so is available in subsequent modifiers. In
1436 particular, the message can be preserved by coding like this:
1438 warn !verify = sender
1439 set acl_m0 = $acl_verify_message
1441 Previously, $acl_verify_message was set only while expanding "message"
1442 and "log_message" when a very denied access.
1444 PH/28 Modified OS/os.c-Linux with
1446 -#ifndef OS_LOAD_AVERAGE
1447 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1449 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1450 nomenclature these days.)
1452 PH/29 Installed patch from the Sieve maintainer that adds the options
1453 sieve_useraddress and sieve_subaddress to the redirect router.
1455 PH/30 In these circumstances:
1456 . Two addresses routed to the same list of hosts;
1457 . First host does not offer TLS;
1458 . First host accepts first address;
1459 . First host gives temporary error to second address;
1460 . Second host offers TLS and a TLS session is established;
1461 . Second host accepts second address.
1462 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1463 and peerdn, if requested) that were in fact used only for the second
1466 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1467 attention to any local part prefix or suffix that was matched by the
1468 router that accepted the address. It now behaves in the same way as it
1469 does for delivery: the affixes are removed from the local part unless
1470 rcpt_include_affixes is set on the transport.
1472 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1473 timeout during the DATA phase of an incoming message.
1475 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1476 applied a patch sanctioned by the Sieve maintainer.
1478 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1479 the uid or gid is negative. A case of a negative gid caused this to be
1480 noticed. The fix allows for either to be negative.
1482 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1483 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1486 PH/36 The change PH/12 above was broken. Fixed it.
1488 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1489 the grounds that routing the same address twice would always produce the
1490 same answer. This might have been true once, but it is certainly no
1491 longer true now. Routing a child address may depend on the previous
1492 routing that produced that child. Some complicated redirection strategies
1493 went wrong when messages had multiple recipients, and made Exim's
1494 behaviour dependent on the order in which the addresses were given.
1496 I have moved the duplicate checking until after the routing is complete.
1497 Exim scans the addresses that are assigned to local and remote
1498 transports, and removes any duplicates. This means that more work will be
1499 done, as duplicates will always all be routed, but duplicates are
1500 presumably rare, so I don't expect this is of any significance.
1502 For deliveries to pipes, files, and autoreplies, the duplicate checking
1503 still happens during the routing process, since they are not going to be
1506 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1507 It corrects a timeout issue with spamd. This is Ian's comment: "The
1508 background is that sometimes spamd either never reads data from a
1509 connection it has accepted, or it never writes response data. The exiscan
1510 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1511 blindly assumes that writes won't block so it may never time out."
1513 PH/39 Allow G after quota size as well as K and M.
1515 PH/40 The value set for $authenticated_id in an authenticator may not contain
1516 binary zeroes or newlines because the value is written to log lines and
1517 to spool files. There was no check on this. Now the value is run through
1518 the string_printing() function so that such characters are converted to
1519 printable escape sequences.
1521 PH/41 $message_linecount is a new variable that contains the total number of
1522 lines in the message. Compare $body_linecount, which is the count for the
1525 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1526 bounce and delay warning messages, except in certain special cases, which
1529 (a) An SMTP error message from a remote host;
1530 (b) A message specified in a :fail: redirection;
1531 (c) A message specified in a "fail" command in a system filter;
1532 (d) A message specified in a FAIL return from the queryprogram router;
1533 (e) A message specified by the cannot_route_message router option.
1535 In these cases only, Exim does include the error details in bounce and
1536 warning messages. There are also a few cases where bland messages such
1537 as "unrouteable address" or "local delivery error" are given.
1539 PH/43 $value is now also set for the "else" part of a ${run expansion.
1541 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1542 being worked on, but at least Exim now implements the latest version to
1545 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1546 process to complete was treated as a delivery failure, a timeout while
1547 writing the message to the pipe was logged, but erroneously treated as a
1548 successful delivery. Such timeouts include transport filter timeouts. For
1549 consistency with the overall process timeout, these timeouts are now
1550 treated as errors, giving rise to delivery failures by default. However,
1551 there is now a new Boolean option for the pipe transport called
1552 timeout_defer, which, if set TRUE, converts the failures into defers for
1553 both kinds of timeout. A transport filter timeout is now identified in
1556 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1557 systems where "make" and "gmake" are different, calling "gmake" at top
1558 level broke things. I've arranged for the value of $(MAKE) to be passed
1559 from the Makefile to this script so that it can call the same version of
1563 A note about Exim versions 4.44 and 4.50
1564 ----------------------------------------
1566 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1567 changes of various kinds. As a consequence, a big documentation update was
1568 needed. This delayed the release for rather longer than seemed good, especially
1569 in the light of a couple of (minor) security issues. Therefore, the changes
1570 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1571 release. So 4.44 and 4.50 are in effect two different branches that both start
1574 I have left the 4.50 change log unchanged; it contains all the changes since
1575 4.43. The change log for 4.44 is below; many of its items are identical to
1576 those for 4.50. This seems to be the most sensible way to preserve the
1577 historical information.
1583 1. Minor wording change to the doc/README.SIEVE file.
1585 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1586 computation of the current number of files was incorrect.
1588 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1589 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1592 4. Give more explanation in the error message when the command for a transport
1593 filter fails to execute.
1595 5. There are several places where Exim runs a non-Exim command in a
1596 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1597 was being done only for the command run by the queryprogram router. It is
1598 now done for all such subprocesses. The other cases are: ${run, transport
1599 filters, and the commands run by the lmtp and pipe transports.
1601 6. Added CONFIGURE_GROUP build-time option.
1603 7. Some older OS have a limit of 256 on the maximum number of file
1604 descriptors. Exim was using setrlimit() to set 1000 as a large value
1605 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1606 systems. I've change it so that if it can't get 1000, it tries for 256.
1608 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1609 was an oversight, and furthermore, ever since the addition of extra
1610 controls (e.g. 4.43/32), the checks on when to allow different forms of
1611 "control" were broken. There should now be diagnostics for all cases when a
1612 control that does not make sense is encountered.
1614 9. Added the /retain_sender option to "control=submission".
1616 10. $recipients is now available in the predata ACL (oversight).
1618 11. Tidy the search cache before the fork to do a delivery from a message
1619 received from the command line. Otherwise the child will trigger a lookup
1620 failure and thereby defer the delivery if it tries to use (for example) a
1621 cached ldap connection that the parent has called unbind on.
1623 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1624 of $address_data from the recipient verification was clobbered by the
1625 sender verification.
1627 13. The value of address_data from a sender verification is now available in
1628 $sender_address_data in subsequent conditions in the ACL statement.
1630 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1632 15. Added a new option "connect=<time>" to callout options, to set a different
1635 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1636 was its contents. (It was OK if the option was not defined at all.)
1638 17. A "Completed" log line is now written for messages that are removed from
1639 the spool by the -Mrm option.
1641 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1642 information about exactly what failed.
1644 19. Added -dd to debug only the daemon process.
1646 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1647 handles timeouts, both on the server side and network timeouts. Renamed the
1648 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1650 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1651 It is now set to "smtps".
1653 22. $host_address is now set to the target address during the checking of
1654 ignore_target_hosts.
1656 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1657 being passed; this would have caused $sender_host_name to have been used if
1658 matching the list had actually called for a host name (not very likely,
1659 since this list is usually IP addresses). A host name is now passed as
1662 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1663 code that specifies a non-restarting handler (typically sigaction() in
1664 modern systems) in an attempt to fix a rare and obscure crash bug.
1666 25. Narrowed the window for a race in the daemon that could cause it to ignore
1667 SIGCHLD signals. This is not a major problem, because they are used only to
1668 wake it up if nothing else does.
1670 26. A malformed maildirsize file could cause Exim to calculate negative values
1671 for the mailbox size or file count. Odd effects could occur as a result.
1672 The maildirsize information is now recalculated if the size or filecount
1675 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1676 support for a long time. Removed HAVE_SYS_VFS_H.
1678 28. Installed the latest version of exipick from John Jetmore.
1680 29. In an address list, if the pattern was not a regular expression, an empty
1681 subject address (from a bounce message) matched only if the pattern was an
1682 empty string. Non-empty patterns were not even tested. This was the wrong
1683 because it is perfectly reasonable to use an empty address as part of a
1684 database query. An empty address is now tested by patterns that are
1685 lookups. However, all the other forms of pattern expect the subject to
1686 contain a local part and a domain, and therefore, for them, an empty
1687 address still always fails if the pattern is not itself empty.
1689 30. Exim went into a mad DNS loop when attempting to do a callout where the
1690 host was specified on an smtp transport, and looking it up yielded more
1691 than one IP address.
1693 31. Re-factored the code for checking spool and log partition space into a
1694 function that finds that data and another that does the check. The former
1695 is then used to implement four new variables: $spool_space, $log_space,
1696 $spool_inodes, and $log_inodes.
1698 32. The RFC2047 encoding function was originally intended for short strings
1699 such as real names; it was not keeping to the 75-character limit for
1700 encoded words that the RFC imposes. It now respects the limit, and
1701 generates multiple encoded words if necessary. To be on the safe side, I
1702 have increased the buffer size for the ${rfc2047: expansion operator from
1705 33. It is now permitted to omit both strings after an "if" condition; if the
1706 condition is true, the result is "true". As before, when the second string
1707 is omitted, a false condition yields an empty string. This makes it less
1708 cumbersome to write custom ACL and router conditions.
1710 34. Failure to deliver a bounce message always caused it to be frozen, even if
1711 there was an errors_to setting on the router. The errors_to setting is now
1714 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1715 canonical form (fully expanded) before being placed in
1716 $sender_host_address.
1718 36. The table in the code that translates DNS record types into text (T_A to
1719 "A" for instance) was missing entries for NS and CNAME. It is just possible
1720 that this could have caused confusion if both these types were looked up
1721 for the same domain, because the text type is used as part of Exim's
1722 per-process caching. But the chance of anyone hitting this buglet seems
1725 37. The dnsdb lookup has been extended in a number of ways.
1727 (1) There is a new type, "zns", which walks up the domain tree until it
1728 finds some nameserver records. It should be used with care.
1730 (2) There is a new type, "mxh", which is like "mx" except that it returns
1731 just the host names, not the priorities.
1733 (3) It is now possible to give a list of domains (or IP addresses) to be
1734 looked up. The behaviour when one of the lookups defers can be
1735 controlled by a keyword.
1737 (4) It is now possible to specify the separator character for use when
1738 multiple records are returned.
1740 38. The dnslists ACL condition has been extended: it is now possible to supply
1741 a list of IP addresses and/or domains to be looked up in a particular DNS
1744 39. Added log_selector=+queue_time_overall.
1746 40. When running the queue in the test harness, wait just a tad after forking a
1747 delivery process, to get repeatability of debugging output.
1749 41. Include certificate and key file names in error message when GnuTLS fails
1750 to set them up, because the GnuTLS error message doesn't include the name
1751 of the failing file when there is a problem reading it.
1753 42. Allow both -bf and -bF in the same test run.
1755 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1757 44. The "Exiscan patch" is now merged into the mainline Exim source.
1759 45. Sometimes the final signoff response after QUIT could fail to get
1760 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1761 < 0 before doing a fflush(). This bug looks as though it goes back to the
1762 introduction of TLS in release 3.20, but "sometimes" must have been rare
1763 because the tests only now provoked it.
1765 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1766 (this can affect the format of dates).
1768 47. exim_tidydb, when checking for the continued existence of a message for
1769 which it has found a message-specific retry record, was not finding
1770 messages that were in split spool directories. Consequently, it was
1771 deleting retry records that should have stayed in existence.
1773 48. Steve fixed some bugs in eximstats.
1775 49. The SPA authentication driver was not abandoning authentication and moving
1776 on to the next authenticator when an expansion was forced to fail,
1777 contradicting the general specification for all authenticators. Instead it
1778 was generating a temporary error. It now behaves as specified.
1780 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1781 (the order specifies the preference for clients). The order is now AES256,
1782 AES128, 3DES, ARCFOUR128.
1784 51. Small patch to Sieve code - explicitly set From: when generating an
1787 52. Exim crashed if a remote delivery caused a very long error message to be
1788 recorded - for instance if somebody sent an entire SpamAssassin report back
1789 as a large number of 550 error lines. This bug was coincidentally fixed by
1790 increasing the size of one of Exim's internal buffers (big_buffer) that
1791 happened as part of the Exiscan merge. However, to be on the safe side, I
1792 have made the code more robust (and fixed the comments that describe what
1795 53. Now that there can be additional text after "Completed" in log lines (if
1796 the queue_time_overall log selector is set), a one-byte patch to exigrep
1797 was needed to allow it to recognize "Completed" as not the last thing in
1800 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1801 patch that reportedly fixes this has been added. I am not expert enough to
1802 create a test for it. This is what the patch creator wrote:
1804 "I found a little strange behaviour of ldap code when working with
1805 Windows 2003 AD Domain, where users was placed in more than one
1806 Organization Units. When I tried to give exim partial DN, the exit code
1807 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1808 But simultaneously result of request was absolutely normal ldap result,
1809 so I produce this patch..."
1811 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1812 REFERENCE, so I have modified the code to exclude the patch when that macro
1815 55. Some experimental protocols are using DNS PTR records for new purposes. The
1816 keys for these records are domain names, not reversed IP addresses. The
1817 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1818 leaves it alone. Component reversal etc. now happens only for IP addresses.
1821 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1823 57. Double the size of the debug message buffer (to 2048) so that more of very
1824 long debug lines gets shown.
1826 58. The exicyclog utility now does better if the number of log files to keep
1827 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1829 59. Two changes related to the smtp_active_hostname option:
1831 (1) $smtp_active_hostname is now available as a variable.
1832 (2) The default for smtp_banner uses $smtp_active_hostname instead
1833 of $primary_hostname.
1835 60. The host_aton() function is supposed to be passed a string that is known
1836 to be a valid IP address. However, in the case of IPv6 addresses, it was
1837 not checking this. This is a hostage to fortune. Exim now panics and dies
1838 if the condition is not met. A case was found where this could be provoked
1839 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1840 components; fortuitously, this particular loophole had already been fixed
1841 by change 4.50/55 above.
1843 If there are any other similar loopholes, the new check in host_aton()
1844 itself should stop them being exploited. The report I received stated that
1845 data on the command line could provoke the exploit when Exim was running as
1846 exim, but did not say which command line option was involved. All I could
1847 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1848 running as the user.
1851 61. There was a buffer overflow vulnerability in the SPA authentication code
1852 (which came originally from the Samba project). I have added a test to the
1853 spa_base64_to_bits() function which I hope fixes it.
1856 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1857 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1859 63. The daemon start-up calls getloadavg() while still root for those OS that
1860 need the first call to be done as root, but it missed one case: when
1861 deliver_queue_load_max is set with deliver_drop_privilege. This is
1862 necessary for the benefit of the queue runner, because there is no re-exec
1863 when deliver_drop_privilege is set.
1865 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1866 This has been fixed.
1868 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1869 was in use, was not putting the data itself into the right store pool;
1870 consequently, it could be overwritten for a subsequent message in the same
1871 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1874 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1876 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1877 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1878 the function were treating the return as a boolean value, which happened to
1879 work because 0=false and not-0=true, but is not correct code.
1881 68. The host_aton() function was not handling scoped IPv6 addresses (those
1882 with, for example, "%eth0" on the end) correctly.
1884 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1885 negated items (that is, ~something) in unsigned ints. Some compilers
1886 apparently mutter when there is no cast.
1888 70. If an address verification called from an ACL failed, and did not produce a
1889 user-specific message (i.e. there was only a "system" message), nothing was
1890 put in $acl_verify_message. In this situation, it now puts the system
1893 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1894 session; change 4.31/43 added the unwanted input to the log line - except
1895 that it did not do this in the start of session case. It now does.
1897 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1898 This is wrong and can cause the other end to generate a synchronization
1899 error if it is another Exim or anything else that does the synchronization
1900 check. A QUIT command is no longer sent after a timeout.
1902 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1903 during host lookups.
1905 74. The defer_ok option of callout verification was not working if it was used
1906 when verifying addresses in header lines, that is, for this case:
1908 verify = header_sender/callout=defer_ok
1910 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1911 those file descriptors could be used for SMTP connections. If anything
1912 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1913 could be sent to the SMTP client, causing chaos. The daemon now opens
1914 stdin, stdout, and stderr to /dev/null when it puts itself into the
1917 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1918 log by default. The user can override this with suitable Perl magic.
1920 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1921 the log message when discard triggers, was not working for the DATA ACL or
1922 for the non-SMTP ACL.
1924 78. Error message wording change in sieve.c.
1926 79. If smtp_accept_max_per_host was set, the number of connections could be
1927 restricted to fewer than expected, because the daemon was trying to set up
1928 a new connection before checking whether the processes handling previous
1929 connections had finished. The check for completed processes is now done
1930 earlier. On busy systems, this bug wouldn't be noticed because something
1931 else would have woken the daemon, and it would have reaped the completed
1934 80. If a message was submitted locally by a user whose login name contained one
1935 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1936 It caused a spool format error. I have fixed the spool reading code. A
1937 related problem was that the "from" clause in the Received: line became
1938 illegal because of the space(s). It is now covered by ${quote_local_part.
1940 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1943 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1945 83. Patch from Sieve maintainer to fix unterminated string problem in
1946 "vacation" handling.
1948 84. Some minor changes to the Linux configuration files to help with other
1949 OS variants using glibc.
1951 85. One more patch for Sieve to update vacation handling to latest spec.
1954 ----------------------------------------------------
1955 See the note above about the 4.44 and 4.50 releases.
1956 ----------------------------------------------------
1962 1. Change 4.43/35 introduced a bug that caused file counts to be
1963 incorrectly computed when quota_filecount was set in an appendfile
1966 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1967 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1970 3. Give more explanation in the error message when the command for a transport
1971 filter fails to execute.
1973 4. There are several places where Exim runs a non-Exim command in a
1974 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1975 was being done only for the command run by the queryprogram router. It is
1976 now done for all such subprocesses. The other cases are: ${run, transport
1977 filters, and the commands run by the lmtp and pipe transports.
1979 5. Some older OS have a limit of 256 on the maximum number of file
1980 descriptors. Exim was using setrlimit() to set 1000 as a large value
1981 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1982 systems. I've change it so that if it can't get 1000, it tries for 256.
1984 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1985 was an oversight, and furthermore, ever since the addition of extra
1986 controls (e.g. 4.43/32), the checks on when to allow different forms of
1987 "control" were broken. There should now be diagnostics for all cases when a
1988 control that does not make sense is encountered.
1990 7. $recipients is now available in the predata ACL (oversight).
1992 8. Tidy the search cache before the fork to do a delivery from a message
1993 received from the command line. Otherwise the child will trigger a lookup
1994 failure and thereby defer the delivery if it tries to use (for example) a
1995 cached ldap connection that the parent has called unbind on.
1997 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1998 of $address_data from the recipient verification was clobbered by the
1999 sender verification.
2001 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2002 was its contents. (It was OK if the option was not defined at all.)
2004 11. A "Completed" log line is now written for messages that are removed from
2005 the spool by the -Mrm option.
2007 12. $host_address is now set to the target address during the checking of
2008 ignore_target_hosts.
2010 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2011 being passed; this would have caused $sender_host_name to have been used if
2012 matching the list had actually called for a host name (not very likely,
2013 since this list is usually IP addresses). A host name is now passed as
2016 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2017 code that specifies a non-restarting handler (typically sigaction() in
2018 modern systems) in an attempt to fix a rare and obscure crash bug.
2020 15. Narrowed the window for a race in the daemon that could cause it to ignore
2021 SIGCHLD signals. This is not a major problem, because they are used only to
2022 wake it up if nothing else does.
2024 16. A malformed maildirsize file could cause Exim to calculate negative values
2025 for the mailbox size or file count. Odd effects could occur as a result.
2026 The maildirsize information is now recalculated if the size or filecount
2029 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2030 support for a long time. Removed HAVE_SYS_VFS_H.
2032 18. Updated exipick to current release from John Jetmore.
2034 19. Allow an empty sender to be matched against a lookup in an address list.
2035 Previously the only cases considered were a regular expression, or an
2038 20. Exim went into a mad DNS lookup loop when doing a callout where the
2039 host was specified on the transport, if the DNS lookup yielded more than
2042 21. The RFC2047 encoding function was originally intended for short strings
2043 such as real names; it was not keeping to the 75-character limit for
2044 encoded words that the RFC imposes. It now respects the limit, and
2045 generates multiple encoded words if necessary. To be on the safe side, I
2046 have increased the buffer size for the ${rfc2047: expansion operator from
2049 22. Failure to deliver a bounce message always caused it to be frozen, even if
2050 there was an errors_to setting on the router. The errors_to setting is now
2053 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2054 canonical form (fully expanded) before being placed in
2055 $sender_host_address.
2057 24. Updated eximstats to version 1.33
2059 25. Include certificate and key file names in error message when GnuTLS fails
2060 to set them up, because the GnuTLS error message doesn't include the name
2061 of the failing file when there is a problem reading it.
2063 26. Expand error message when OpenSSL has problems setting up cert/key files.
2066 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2067 (this can affect the format of dates).
2069 28. exim_tidydb, when checking for the continued existence of a message for
2070 which it has found a message-specific retry record, was not finding
2071 messages that were in split spool directories. Consequently, it was
2072 deleting retry records that should have stayed in existence.
2074 29. eximstats updated to version 1.35
2075 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2076 1.35 - bugfix such that pie charts by volume are generated correctly
2078 30. The SPA authentication driver was not abandoning authentication and moving
2079 on to the next authenticator when an expansion was forced to fail,
2080 contradicting the general specification for all authenticators. Instead it
2081 was generating a temporary error. It now behaves as specified.
2083 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2084 (the order specifies the preference for clients). The order is now AES256,
2085 AES128, 3DES, ARCFOUR128.
2087 31. Small patch to Sieve code - explicitly set From: when generating an
2090 32. Exim crashed if a remote delivery caused a very long error message to be
2091 recorded - for instance if somebody sent an entire SpamAssassin report back
2092 as a large number of 550 error lines. This bug was coincidentally fixed by
2093 increasing the size of one of Exim's internal buffers (big_buffer) that
2094 happened as part of the Exiscan merge. However, to be on the safe side, I
2095 have made the code more robust (and fixed the comments that describe what
2098 33. Some experimental protocols are using DNS PTR records for new purposes. The
2099 keys for these records are domain names, not reversed IP addresses. The
2100 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2101 leaves it alone. Component reversal etc. now happens only for IP addresses.
2104 34. The host_aton() function is supposed to be passed a string that is known
2105 to be a valid IP address. However, in the case of IPv6 addresses, it was
2106 not checking this. This is a hostage to fortune. Exim now panics and dies
2107 if the condition is not met. A case was found where this could be provoked
2108 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2109 components; fortuitously, this particular loophole had already been fixed
2110 by change 4.50/55 or 4.44/33 above.
2112 If there are any other similar loopholes, the new check in host_aton()
2113 itself should stop them being exploited. The report I received stated that
2114 data on the command line could provoke the exploit when Exim was running as
2115 exim, but did not say which command line option was involved. All I could
2116 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2117 running as the user.
2120 35. There was a buffer overflow vulnerability in the SPA authentication code
2121 (which came originally from the Samba project). I have added a test to the
2122 spa_base64_to_bits() function which I hope fixes it.
2125 36. The daemon start-up calls getloadavg() while still root for those OS that
2126 need the first call to be done as root, but it missed one case: when
2127 deliver_queue_load_max is set with deliver_drop_privilege. This is
2128 necessary for the benefit of the queue runner, because there is no re-exec
2129 when deliver_drop_privilege is set.
2131 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2132 was in use, was not putting the data itself into the right store pool;
2133 consequently, it could be overwritten for a subsequent message in the same
2134 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2137 38. Sometimes the final signoff response after QUIT could fail to get
2138 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2139 < 0 before doing a fflush(). This bug looks as though it goes back to the
2140 introduction of TLS in release 3.20, but "sometimes" must have been rare
2141 because the tests only now provoked it.
2147 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2148 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2149 Now it returns an integer. A number of calls were still expecting a T/F
2150 return. Fortuitously, in all cases, the tests worked in OK situations,
2151 which is the norm. However, things would have gone wrong on any write
2152 failures on the smtp file descriptor. This function is used when sending
2153 messages over SMTP and also when doing verify callouts.
2155 2. When Exim is called to do synchronous delivery of a locally submitted
2156 message (the -odf or -odi options), it no longer closes stderr before doing
2159 3. Implemented the mua_wrapper option.
2161 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2163 5. Implemented the functions header_remove(), header_testname(),
2164 header_add_at_position(), and receive_remove_recipient(), and exported them
2167 6. If an ACL "warn" statement specified the addition of headers, Exim already
2168 inserted X-ACL-Warn: at the start if there was no header name. However, it
2169 was not making this test for the second and subsequent header lines if
2170 there were newlines in the string. This meant that an invalid header could
2171 be inserted if Exim was badly configured.
2173 7. Allow an ACL "warn" statement to add header lines at the start or after all
2174 the Received: headers, as well as at the end.
2176 8. Added the rcpt_4xx retry error code.
2178 9. Added postmaster_mailfrom=xxx to callout verification option.
2180 10. Added mailfrom=xxxx to the callout verification option, for verify=
2183 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2184 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2186 12. Inserted some casts to stop certain compilers warning when using pointer
2187 differences as field lengths or precisions in printf-type calls (mostly
2188 affecting debugging statements).
2190 13. Added optional readline() support for -be (dynamically loaded).
2192 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2193 same clock tick as a message's arrival, so that its received time was the
2194 same as the "first fail" time on the retry record, and that message
2195 remained on the queue past the ultimate address timeout, every queue runner
2196 would try a delivery (because it was past the ultimate address timeout) but
2197 after another failure, the ultimate address timeout, which should have then
2198 bounced the address, did not kick in. This was a "< instead of <=" error;
2199 in most cases the first failure would have been in the next clock tick
2200 after the received time, and all would be well.
2202 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2203 being recognized when the domain list was tested by the match_domain
2204 condition in an expansion string.
2206 16. Added the ${str2b64: operator.
2208 17. Exim was always calling setrlimit() to set a large limit for the number of
2209 processes, without checking whether the existing limit was already
2210 adequate. (It did check for the limit on file descriptors.) Furthermore,
2211 errors from getrlimit() and setrlimit() were being ignored. Now they are
2212 logged to the main and panic logs, but Exim does carry on, to try to do its
2213 job under whatever limits there are.
2215 18. Imported PCRE 5.0.
2217 19. Trivial typo in log message " temporarily refused connection" (the leading
2220 20. If the log selector return_path_on_delivery was set and an address was
2221 redirected to /dev/null, the delivery process crashed because it assumed
2222 that a return path would always be set for a "successful" delivery. In this
2223 case, the whole delivery is bypassed as an optimization, and therefore no
2226 21. Internal re-arrangement: the function for sending a challenge and reading
2227 a response while authentication was assuming a zero-terminated challenge
2228 string. It's now changed to take a pointer and a length, to allow for
2229 binary data in such strings.
2231 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2233 23. Exim was not respecting finduser_retries when seeking the login of the
2234 uid under which it was called; it was always trying 10 times. (The default
2235 setting of finduser_retries is zero.) Also, it was sleeping after the final
2236 failure, which is pointless.
2238 24. Implemented tls_on_connect_ports.
2240 25. Implemented acl_smtp_predata.
2242 26. If the domain in control=submission is set empty, Exim assumes that the
2243 authenticated id is a complete email address when it generates From: or
2244 Sender: header lines.
2246 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2247 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2248 chown and chgrp in /bin and hostname in /usr/bin.
2250 28. Exim was keeping the "process log" file open after each use, just as it
2251 does for the main log. This opens the possibility of it remaining open for
2252 long periods when the USR1 signal hits a daemon. Occasional processlog
2253 errors were reported, that could have been caused by this. Anyway, it seems
2254 much more sensible not to leave this file open at all, so that is what now
2257 29. The long-running daemon process does not normally write to the log once it
2258 has entered its main loop, and it closes the log before doing so. This is
2259 so that log files can straightforwardly be renamed and moved. However,
2260 there are a couple of unusual error situations where the daemon does write
2261 log entries, and I had neglected to close the log afterwards.
2263 30. The text of an SMTP error response that was received during a remote
2264 delivery was being truncated at 512 bytes. This is too short for some of
2265 the long messages that one sometimes sees. I've increased the limit to
2268 31. It is now possible to make retry rules that apply only when a message has a
2269 specific sender, in particular, an empty sender.
2271 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2272 it possible to be selective about when SMTP synchronization is enforced.
2274 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2276 32. Implemented hosts_connection_nolog.
2278 33. Added an ACL for QUIT.
2280 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2283 35. Added mailbox_size and mailbox_filecount to appendfile.
2285 36. Added control = no_multiline_responses to ACLs.
2287 37. There was a bug in the logic of the code that waits for the clock to tick
2288 in the case where the clock went backwards by a substantial amount such
2289 that the microsecond fraction of "now" was more than the microsecond
2290 fraction of "then" (but the whole seconds number was less).
2292 38. Added support for the libradius Radius client library this is found on
2293 FreeBSD (previously only the radiusclient library was supported).
2299 1. When certain lookups returned multiple values in the form name=value, the
2300 quoting of the values was not always being done properly. Specifically:
2301 (a) If the value started with a double quote, but contained no whitespace,
2303 (b) If the value contained whitespace other than a space character (i.e.
2304 tabs or newlines or carriage returns) it was not quoted.
2305 This fix has been applied to the mysql and pgsql lookups by writing a
2306 separate quoting function and calling it from the lookup code. The fix
2307 should probably also be applied to nisplus, ibase and oracle lookups, but
2308 since I cannot test any of those, I have not disturbed their existing code.
2310 2. A hit in the callout cache for a specific address caused a log line with no
2311 reason for rejecting RCPT. Now it says "Previous (cached) callout
2312 verification failure".
2314 3. There was an off-by-one bug in the queryprogram router. An over-long
2315 return line was truncated at 256 instead of 255 characters, thereby
2316 overflowing its buffer with the terminating zero. As well as fixing this, I
2317 have increased the buffer size to 1024 (and made a note to document this).
2319 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2320 when Exim is waiting for an SMTP response from a remote server, Exim
2321 restarts its select() call on the socket, thereby resetting its timeout.
2322 This is not a problem when such interrupts are rare. Somebody set up a cron
2323 job to run exiwhat every 2 minutes, which is less than the normal select()
2324 timeout (5 or 10 minutes). This meant that the select() timeout never
2325 kicked in because it was always reset. I have fixed this by comparing the
2326 time when an interrupt arrives with the time at the start of the first call
2327 to select(). If more time than the timeout has elapsed, the interrupt is
2328 treated as a timeout.
2330 5. Some internal re-factoring in preparation for the addition of Sieve
2331 extensions (by MH). In particular, the "personal" test is moved to a
2332 separate function, and given an option for scanning Cc: and Bcc: (which is
2333 not set for Exim filters).
2335 6. When Exim created an email address using the login of the caller as the
2336 local part (e.g. when creating a From: or Sender: header line), it was not
2337 quoting the local part when it contained special characters such as @.
2339 7. Installed new OpenBSD configuration files.
2341 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2342 try to make them clearer.
2344 9. Callout options, other than the timeout value, were being ignored when
2345 verifying sender addresses in header lines. For example, when using
2347 verify = header_sender/callout=no_cache
2349 the cache was (incorrectly) being used.
2351 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2352 only the Cygwin environment.
2354 11. When return_path_on_delivery was set as a log selector, if different remote
2355 addresses in the same message used different return paths and parallel
2356 remote delivery occurred, the wrong values would sometimes be logged.
2357 (Whenever a remote delivery process finished, the return path value from
2358 the most recently started remote delivery process was logged.)
2360 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2361 lines when AUTH and/or TLS are in use. This is the "received protocol"
2362 field. Exim used to use "asmtp" for authenticated SMTP, without any
2363 indication (in the protocol name) for TLS use. Now it follows the RFC and
2364 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2365 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2366 names appear in log lines as well as in Received: header lines.
2368 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2369 extensions, and comparison tests, and to fix some bugs.
2371 14. Changes to the "personal" filter test:
2373 (1) The test was buggy in that it was just doing the equivalent of
2374 "contains" tests on header lines. For example, if a user's address was
2375 anne@some.where, the "personal" test would incorrectly be true for
2377 To: susanne@some.where
2379 This test is now done by extracting each address from the header in turn,
2380 and checking the entire address. Other tests that are part of "personal"
2381 are now done using regular expressions (for example, to check local parts
2382 of addresses in From: header lines).
2384 (2) The list of non-personal local parts in From: addresses has been
2385 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2386 taken from the Sieve specification recommendations.
2388 (3) If the message contains any header line starting with "List-" it is
2389 treated as non-personal.
2391 (4) The test for "circular" in the Subject: header line has been removed
2392 because it now seems ill-conceived.
2394 15. Minor typos in src/EDITME comments corrected.
2396 16. Installed latest exipick from John Jetmore.
2398 17. If headers_add on a router specified a text string that was too long for
2399 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2400 of string_sprintf() is now avoided.
2402 18. $message_body_size was not set (it was always zero) when running the DATA
2403 ACL and the local_scan() function.
2405 19. For the "mail" command in an Exim filter, no default was being set for
2406 the once_repeat time, causing a random time value to be used if "once" was
2407 specified. (If the value happened to be <= 0, no repeat happened.) The
2408 default is now 0s, meaning "never repeat". The "vacation" command was OK
2409 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2410 (I found it when inspecting the code).
2412 20. There is now an overall timeout for performing a callout verification. It
2413 defaults to 4 times the callout timeout, which applies to individual SMTP
2414 commands during the callout. The overall timeout applies when there is more
2415 than one host that can be tried. The timeout is checked before trying the
2416 next host. This prevents very long delays if there are a large number of
2417 hosts and all are timing out (e.g. when the network connections are timing
2418 out). The value of the overall timeout can be changed by specifying an
2419 additional sub-option for "callout", called "maxwait". For example:
2421 verify = sender/callout=5s,maxwait=20s
2423 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2424 to the end before writing, but this should make it even safer).
2426 22. Exim was forgetting that it had advertised PIPELINING for the second and
2427 subsequent messages on an SMTP connection. It was also not resetting its
2428 memory on STARTTLS and an internal HELO.
2430 23. When Exim logs an SMTP synchronization error within a session, it now
2431 records whether PIPELINING has been advertised or not.
2433 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2434 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2435 rather than long int.
2437 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2439 26. Added the never_mail option to autoreply.
2445 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2446 crash if the getsockname() call failed; this can happen if a connection is
2447 closed very soon after it is established. The problem was simply in the
2448 order in which certain operations were done, causing Exim to try to write
2449 to the SMTP stream before it had set up the file descriptor. The bug has
2450 been fixed by making things happen in the correct order.
2456 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2457 before the connection was closed, thus losing the rejection response.
2459 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2460 some early Solaris releases, but causes trouble in current releases where
2461 socklen_t is defined.
2463 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2466 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2469 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2470 that releasing the top of it at the end releases what was used for sub-
2471 expansions (unless the block got too big). However, discard this block if
2472 the first thing is a variable or header, so that we can use its block when
2473 it is dynamic (useful for very large $message_headers, for example).
2475 6. Lookups now cache *every* query, not just the most recent. A new, separate
2476 store pool is used for this. It can be recovered when all lookup caches are
2477 flushed. Lookups now release memory at the end of their result strings.
2478 This has involved some general refactoring of the lookup sources.
2480 7. Some code has been added to the store_xxx() functions to reduce the amount
2481 of flapping under certain conditions.
2483 8. log_incoming_interface used to affect only the <= reception log lines. Now
2484 it causes the local interface and port to be added to several more SMTP log
2485 lines, for example "SMTP connection from", and rejection lines.
2487 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2489 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2491 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2492 could be overwritten at the end of the current message (or the start of a
2493 new message if it was set in a HELO ACL). The value is now preserved for
2494 the duration of the SMTP connection.
2496 12. If a transport had a headers_rewrite setting, and a matching header line
2497 contained an unqualified address, that address was qualified, even if it
2498 did not match any rewriting rules. The underlying bug was that the values
2499 of the flags that permit the existence of unqualified sender and recipient
2500 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2501 non-local messages, and by -bnq for local messages) were not being
2502 preserved with the message after it was received.
2504 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2505 "next input=" as part of the text comprising the host identity instead of
2506 the correct text. The code was using the same buffer for two different
2507 strings. However, depending on which order the printing function evaluated
2508 its arguments, the bug did not always show up. Under Linux, for example, my
2509 test suite worked just fine.
2511 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2512 doesn't work with some older versions of Perl. It has been changed to "my",
2513 which in any case is probably the better facility to use.
2515 15. A really picky compiler found some instances of statements for creating
2516 error messages that either had too many or two few arguments for the format
2519 16. The size of the buffer for calls to the DNS resolver has been increased
2520 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2521 for addresses that have a lot of PTR records. This alleviates a problem; it
2522 does not fully solve it.
2524 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2525 buffer now truncates the list and logs the incident, which is the same
2526 action as happens when Exim is looking up a host name and its aliases.
2527 Previously in this situation something unpredictable would happen;
2528 sometimes it was "internal error: store_reset failed".
2530 18. If a server dropped the connection unexpectedly when an Exim client was
2531 using GnuTLS and trying to read a response, the client delivery process
2532 crashed while trying to generate an error log message.
2534 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2535 string, for example:
2537 warn message = H1: something\nH2: something
2539 the text was added as a single header line from Exim's point of view
2540 though it ended up OK in the delivered message. However, searching for the
2541 second and subsequent header lines using $h_h2: did not work. This has been
2542 fixed. Similarly, if a system filter added multiple headers in this way,
2543 the routers could not see them.
2545 20. Expanded the error message when iplsearch is called with an invalid key to
2546 suggest using net-iplsearch in a host list.
2548 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2549 ACLs are no longer actually imposed (and a message to that effect is
2552 22. If a "gecos" field in a passwd entry contained escaped characters, in
2553 particular, if it contained a \" sequence, Exim got it wrong when building
2554 a From: or a Sender: header from that name. A second bug also caused
2555 incorrect handling when an unquoted " was present following a character
2556 that needed quoting.
2558 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2559 was not being matched caselessly.
2561 24. Arranged for all hyphens in the exim.8 source to be escaped with
2564 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2565 properly. Recipient callout cache records were still being keyed to include
2566 the sender, even when use_sender was set false. This led to far more
2567 callouts that were necessary. The sender is no longer included in the key
2568 when use_sender is false.
2570 26. Added "control = submission" modifier to ACLs.
2572 27. Added the ${base62d: operator to decode base 62 numbers.
2574 28. dnsdb lookups can now access SRV records.
2576 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2577 the configuration file.
2579 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2580 (-v) mode. This makes the output for a verbose queue run more intelligible.
2582 31. Added a use_postmaster feature to recipient callouts.
2584 32. Added the $body_zerocount variable, containing the number of binary zero
2585 bytes in the message body.
2587 33. The time of last modification of the "new" subdirectory is now used as the
2588 "mailbox time last read" when there is a quota error for a maildir
2591 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2593 35. Added +ignore_unknown as a special item in host lists.
2595 36. Code for decoding IPv6 addresses in host lists is now included, even if
2596 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2597 address was recognized as an IP address, but was then not correctly decoded
2598 into binary, causing unexpected and incorrect effects when compared with
2605 1. Very minor rewording of debugging text in manualroute to say "list of
2606 hosts" instead of "hostlist".
2608 2. If verify=header_syntax was set, and a header line with an unqualified
2609 address (no domain) and a large number of spaces between the end of the
2610 name and the colon was received, the reception process suffered a buffer
2611 overflow, and (when I tested it) crashed. This was caused by some obsolete
2612 code that should have been removed. The fix is to remove it!
2614 3. When running in the test harness, delay a bit after writing a bounce
2615 message to get a bit more predictability in the log output.
2617 4. Added a call to search_tidyup() just before forking a reception process. In
2618 theory, someone could use a lookup in the expansion of smtp_accept_max_
2619 per_host which, without the tidyup, could leave open a database connection.
2621 5. Added the variables $recipient_data and $sender_data which get set from a
2622 lookup success in an ACL "recipients" or "senders" condition, or a router
2623 "senders" option, similar to $domain_data and $local_part_data.
2625 6. Moved the writing of debug_print from before to after the "senders" test
2628 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2629 problems for message scanning, either using a data ACL, or using
2630 local_scan() because the Received: header was not generated till after they
2631 were called (in order to set the time as the time of reception completion).
2632 I have revised the way this works. The header is now generated after the
2633 body is received, but before the ACL or local_scan() are called. After they
2634 are run, the timestamp in the header is updated.
2640 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2641 before starting a queue runner without re-exec. This happened only when
2642 deliver_drop_privilege was set or when the Exim user was set to root. The
2643 effect of the bug was that timeouts during subsequent deliveries caused
2644 crashes instead of being properly handled. The handler is now left at its
2645 default (and expected) setting.
2647 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2648 message, again when deliver_drop_privilege is set or Exim is run as root.
2649 The bug described in (1) was not present in this case, but the tidying up
2650 of the other signals was missing. I have made the two cases consistent.
2652 3. The ignore_target_hosts setting on a manualroute router was being ignored
2653 for hosts that were looked up using the /MX notation.
2655 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2658 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2659 operated on the sender address. After changing the $sender_address to <>
2660 for the sender address verify, Exim was re-instated it as the original
2661 (before rewriting) address, but remembering that it had rewritten it, so it
2662 wasn't rewriting it again. This bug also had the effect of breaking the
2663 sender address verification caching when the sender address was rewritten.
2665 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2666 This has been changed so that if the ip literal address matches
2667 ignore_target_hosts, the router declines.
2669 7. Added expansion conditions match_domain, match_address, and match_local_
2670 part (NOT match_host).
2672 8. The placeholder for the Received: header didn't have a length field set.
2674 9. Added code to Exim itself and to exim_lock to test for a specific race
2675 condition that could lead to file corruption when using MBX delivery. The
2676 issue is with the lockfile that is created in /tmp. If this file is removed
2677 after a process has opened it but before that process has acquired a lock,
2678 there is the potential for a second process to recreate the file and also
2679 acquire a lock. This could lead to two Exim processes writing to the file
2680 at the same time. The added code performs the same test as UW imapd; it
2681 checks after acquiring the lock that its file descriptor still refers to
2682 the same named file.
2684 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2685 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2688 11. Added the smtp_active_hostname option. If used, this will typically be made
2689 to depend on the incoming interface address. Because $interface_address is
2690 not set up until the daemon has forked a reception process, error responses
2691 that can happen earlier (such as "too many connections") no longer contain
2694 12. If an expansion in a condition on a "warn" statement fails because a lookup
2695 defers, the "warn" statement is abandoned, and the next ACL statement is
2696 processed. Previously this caused the whole ACL to be aborted.
2698 13. Added the iplsearch lookup type.
2700 14. Added ident_timeout as a log selector.
2702 15. Added tls_certificate_verified as a log selector.
2704 16. Added a global option tls_require_ciphers (compare the smtp transport
2705 option of the same name). This controls incoming TLS connections.
2707 17. I finally figured out how to make tls_require_ciphers do a similar thing
2708 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2709 before starting the TLS session.
2711 18. Tabs are now shown as \t in -bP output.
2713 19. If the log selector return_path_on_delivery was set, Exim crashed when
2714 bouncing a message because it had too many Received: header lines.
2716 20. If two routers both had headers_remove settings, and the first one included
2717 a superfluous trailing colon, the final name in the first list and the
2718 first name in the second list were incorrectly joined into one item (with a
2719 colon in the middle).
2725 1. Added -C and -D options to the exinext utility, mainly to make it easier
2726 to include in the automated testing, but these could be helpful when
2727 multiple configurations are in use.
2729 2. The exinext utility was not formatting the output nicely when there was
2730 an alternate port involved in the retry record key, nor when there was a
2731 message id as well (for retries that were specific to a specific message
2732 and a specific host). It was also confused by IPv6 addresses, because of
2733 the additional colons they contain. I have fixed the IPv4 problem, and
2734 patched it up to do a reasonable job for IPv6.
2736 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2737 delivery, the log line now contains "pipelined" if PIPELINING was used.
2739 4. An SMTP transport process used to panic and die if the bind() call to set
2740 an explicit outgoing interface failed. This has been changed; it is now
2741 treated in the same way as a connect() failure.
2743 5. A reference to $sender_host_name in the part of a conditional expansion
2744 that was being skipped was still causing a DNS lookup. This no longer
2747 6. The def: expansion condition was not recognizing references to header lines
2748 that used bh_ and bheader_.
2750 7. Added the _cache feature to named lists.
2752 8. The code for checking quota_filecount in the appendfile transport was
2753 allowing one more file than it should have been.
2755 9. For compatibility with Sendmail, the command line option
2763 and sets the incoming protocol and host name (for trusted callers). The
2764 host name and its colon can be omitted when only the protocol is to be set.
2765 Note the Exim already has two private options, -pd and -ps, that refer to
2766 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2767 "s", but I don't think that's a major issue.
2769 10. A number of refactoring changes to the code, none of which should affect
2772 (a) The number of logging options was getting close to filling up the
2773 32-bit word that was used as a bit map. I have split them into two classes:
2774 those that are passed in the argument to log_write(), and those that are
2775 only ever tested independently outside of that function. These are now in
2776 separate 32-bit words, so there is plenty of room for expansion again.
2777 There is no change in the user interface or the logging behaviour.
2779 (b) When building, for example, log lines, the code previously used a
2780 macro that called string_cat() twice, in order to add two strings. This is
2781 not really sufficiently general. Furthermore, there was one instance where
2782 it was actually wrong because one of the argument was used twice, and in
2783 one call a function was used. (As it happened, calling the function twice
2784 did not affect the overall behaviour.) The macro has been replaced by a
2785 function that can join an arbitrary number of extra strings onto a growing
2788 (c) The code for expansion conditions now uses a table and a binary chop
2789 instead of a serial search (which was left over from when there were very
2790 few conditions). Also, it now recognizes conditions like "pam" even when
2791 the relevant support is not compiled in: a suitably worded error message is
2792 given if an attempt is made to use such a condition.
2794 11. Added ${time_interval:xxxxx}.
2796 12. A bug was causing one of the ddress fields not to be passed back correctly
2797 from remote delivery subprocesses. The field in question was not being
2798 subsequently used, so this caused to problems in practice.
2800 13. Added new log selectors queue_time and deliver_time.
2802 14. Might have fixed a bug in maildirsizefile handling that threw up
2803 "unexpected character" debug warnings, and recalculated the data
2804 unnecessarily. In any case, I expanded the warning message to give more
2807 15. Added the message "Restricted characters in address" to the statements in
2808 the default ACL that block characters like @ and % in local parts.
2810 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2811 Three changes have been made:
2813 (a) There was a serious bug; a negative response to MAIL caused the whole
2814 recipient domain to be cached as invalid, thereby blocking all messages
2815 to all local parts at the same domain, from all senders. This bug has
2816 been fixed. The domain is no longer cached after a negative response to
2817 MAIL if the sender used is not empty.
2819 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2822 (c) A new callout option, "use_sender" has been added for people who want
2823 the modified behaviour.
2829 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2832 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2833 indeed breaks things for older releases.
2835 3. Added additional logging to the case where there is a problem reading data
2836 from a filter that is running in a subprocess using a pipe, in order to
2837 try to track down a specific problem.
2839 4. Testing facility fudge: when running in the test harness and attempting
2840 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2841 getting "No route to host". Convert this to a timeout.
2843 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2846 6. Some OS don't have socklen_t but use size_t instead. This affects the
2847 fifth argument of getsockopt() amongst other things. This is now
2848 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2849 can be set for individual OS. I have set it for SunOS5, OSF1, and
2850 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2851 some earlier ones do not.
2853 7. Change 4.30/15 was not doing the test caselessly.
2855 8. The standard form for an IPv6 address literal was being rejected by address
2856 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2857 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2858 this, as well as the form without the "IPv6" on the front (but only when
2859 address literals are enabled, of course).
2861 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2863 10. Exim crashed if a message with an empty sender address specified by -f
2864 encountered a router with an errors_to setting. This could be provoked only
2865 by a command such as
2869 where an empty string was supplied; "<>" did not hit this bug.
2871 11. Installed PCRE release 4.5.
2873 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2874 remained set. It is now erased.
2876 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2877 times from message ids (which are base 36 rather than the normal 62).
2879 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2880 were being counted as actual protocol errors, and logged if the log
2881 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2882 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2883 and DATA following a set of rejected RCPTs do not count as protocol errors.
2884 In other words, Exim assumes they were pipelined, though this may not
2885 actually be the case. Of course, in all cases the client gets an
2886 appropriate error code.
2888 15. If a lookup fails in an ACL condition, a message about the failure may
2889 be available; it is used if testing the ACL cannot continue, because most
2890 such messages specify what the cause of the deferral is. However, some
2891 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2892 that caused an old message to be retained and used if a later statement
2893 caused a defer, replacing the real cause of the deferral.
2895 16. If an IP address had so many PTR records that the DNS lookup buffer
2896 was not large enough to hold them, Exim could crash while trying to process
2897 the truncated data. It now detects and logs this case.
2899 17. Further to 4.21/58, another change has been made: if (and only if) the
2900 first line of a message (the first header line) ends with CRLF, a bare LF
2901 in a subsequent header line has a space inserted after it, so as not to
2902 terminate the header.
2904 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2905 unnecessarily, used atoi() instead of strtol(), and didn't check the
2906 termination when getting file sizes from file names by regex.
2908 19. Completely re-implemented the support for maildirsize files, in the light
2909 of a number of problems with the previous contributed implementation
2910 (4.30/29). In particular:
2912 . If the quota is zero, the maildirsize file is maintained, but no quota is
2915 . If the maildir directory does not exist, it is created before any attempt
2916 to write a maildirsize file.
2918 . The quota value in the file is just a cache; if the quota is changed in
2919 the transport, the new value overrides.
2921 . A regular expression is available for excluding directories from the
2924 20. The autoreply transport checks the characters in options that define the
2925 message's headers; it allows continued headers, but it was checking with
2926 isspace() after an embedded newline instead of explicitly looking for a
2929 21. If all the "regular" hosts to which an address was routed had passed their
2930 expiry times, and had not reached their retry times, the address was
2931 bounced, even if fallback hosts were defined. Now Exim should go on to try
2934 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2935 equivalent code in the SMTP transport. Some hosts send humungous responses
2936 to HELO/EHLO, more than 1024 it seems.
2938 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2939 gives compiler warnings in some environments. I've now done it "properly",
2942 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2943 (because of problems with the built-in one) was declared to return uschar *
2944 instead of char *, causing compiler failure.
2946 25. Fixed a file descriptor leak when processing alias/forward files.
2948 26. Fixed a minor format string issue in dbfn.c.
2950 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2952 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2953 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2956 29. When a sender address is verified, it is cached, to save repeating the test
2957 when there is more than one recipient in a message. However, when the
2958 verification involves a callout, it is possible for different callout
2959 options to be set for different recipients. It is too complicated to keep
2960 track of this in the cache, so now Exim always runs a verification when a
2961 callout is required, relying on the callout cache for the optimization.
2962 The overhead is duplication of the address routing, but this should not be
2965 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2966 to be verified with callout=postmaster, and the main callout worked but the
2967 postmaster check failed, the verification correctly failed. However, if a
2968 subsequent RCPT command asked for sender verification *without* the
2969 postmaster check, incorrect caching caused this verification also to fail,
2972 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2973 it was not caching the DNS options (qualify_single, search_parents) that
2974 were used when the lookup failed. A subsequent lookup with different
2975 options therefore always gave the same answer, though there were cases
2976 where it should not have. (Example: a "domains = !$mx_any" option on a
2977 dnslookup router: the "domains" option is always processed without any
2978 widening, but the router might have qualify_single set.) Now Exim uses the
2979 cached value only when the same options are set.
2981 32. Added John Jetmore's "exipick" utility to the distribution.
2983 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2984 than a timeout (e.g. a certificate is required, and is not provided), an
2985 Exim server now closes the connection immediately. Previously it waited for
2986 the client to close - but if the client is SSL, it seems that they each
2987 wait for each other, leading to a delay before one of them times out.
2989 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2990 maintained 0.8.x compatibility because I don't think many are using it, and
2991 it is clearly obsolete.
2993 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2996 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2997 client certificate was expired. A simple patch fixes this, though I don't
2998 understand the full logic of why the verify callback is called multiple
3001 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3002 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3003 which causes problems with some clients (such as the Certicom SSL Plus
3004 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3005 disables the coutermeasure allowing Eudora to connect."
3007 38. Exim was not checking that a write() to a log file succeeded. This could
3008 lead to Bad Things if a log got too big, in particular if it hit a file
3009 size limit. Exim now panics and dies if it cannot write to a log file, just
3010 as it does if it cannot open a log file.
3012 39. Modified OS/Makefile-Linux so that it now contains
3014 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3016 The two -D definitions ensure that Exim is compiled with large file
3017 support, which makes it possible to handle log files that are bigger than
3020 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3021 instance) a domain was checked against a named list that involved a lookup,
3022 causing $domain_data to be set, then another domain was checked against the
3023 same list, then the first domain was re-checked, the value of $domain_data
3024 after the final check could be wrong. In particular, if the second check
3025 failed, it could be set empty. This bug probably also applied to
3028 41. The strip_trailing_dot option was not being applied to the address given
3029 with the -f command-line option.
3031 42. The code for reading a message's header from the spool was incrementing
3032 $received_count, but never initializing it. This meant that the value was
3033 incorrect (doubled) while delivering a message in the same process in which
3034 it was received. In the most common configuration of Exim, this never
3035 happens - a fresh exec is done - but it can happen when
3036 deliver_drop_privilege is set.
3038 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3039 it now includes up to 150 characters of the unexpected data in the log
3042 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3043 and building data strings. The size of both of these buffers was 10 000
3044 bytes - far larger than anybody would *ever* want, thought I. Needless to
3045 say, somebody hit the limit. I have increased the maximum line length to
3046 20 000 and the maximum data length of concatenated lines to 100 000. I have
3047 also fixed two bugs, because there was no checking on these buffers. Tsk,
3048 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3049 buffer is too small.
3051 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3052 lsearch lookups. Now it does.
3054 46. When parsing a route_list item in a manualroute router, a fixed-length
3055 buffer was used for the list of hosts. I made this 1024 bytes long,
3056 thinking that nobody would ever have a list of hosts that long. Wrong.
3057 Somebody had a whole pile of complicated expansion conditions, and the
3058 string was silently truncated, leading to an expansion error. It turns out
3059 that it is easier to change to an unlimited length (owing to other changes
3060 that have happened since this code was originally written) than to build
3061 structure for giving a limitation error. The length of the item that
3062 expands into the list of hosts is now unlimited.
3064 47. The lsearch lookup could not handle data where the length of text line was
3065 more than 4095 characters. Such lines were truncated, leading to shortened
3066 data being returned. It should now handle lines of any length.
3068 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3069 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3072 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3073 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3074 no longer happens in comment lines. A list of replacements is now placed
3075 at the head of all of the source files, except those whose only change is
3076 to replace PERL_COMMAND in the very first #! line.
3078 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3079 messages on the queue, with a bottom-up merge sort, using code contributed
3080 by Michael Haardt. This should make operations like -bp somewhat faster on
3081 large queues. It won't affect queue runners, except when queue_run_in_order
3084 51. Installed eximstats 1.31 in the distribution.
3086 52. Added support for SRV lookups to the dnslookup router.
3088 53. If an ACL referred to $message_body or $message_body_end, the value was not
3089 reset for any messages that followed in the same SMTP session.
3091 54. The store-handling optimization for building very long strings was not
3092 differentiating between the different store pools. I don't think this
3093 actually made any difference in practice, but I've tidied it.
3095 55. While running the routers to verify a sender address, $sender_address
3096 was still set to the sender address. This is wrong, because when routing to
3097 send a bounce to the sender, it would be empty. Therefore, I have changed
3098 it so that, while verifying a sender address, $sender_address is set to <>.
3099 (There is no change to what happens when verifying a recipient address.)
3101 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3102 target A or AAAA records (if not already returned) without resetting the
3103 qualify_single or search_parents options of the DNS resolver. These are
3104 inappropriate in this case because the targets of MX and SRV records must
3105 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3106 target that, when qualified, matched something in the local domain. These
3107 two options are now turned off when doing these lookups.
3109 57. It seems that at least some releases of Reiserfs (which does not have the
3110 concept of a fixed number of inodes) returns zero and not -1 for the
3111 number of available inodes. This interacted badly with check_spool_inodes,
3112 which assumed that -1 was the "no such thing" setting. What I have done is
3113 to check that the total number of inodes is greater than zero before doing
3114 the test of how many are available.
3116 58. When a "warn" ACL statement has a log_message modifier, the message is
3117 remembered, and not repeated. This is to avoid a lot of repetition when a
3118 message has many recipients that cause the same warning to be written.
3119 Howewer, Exim was preserving the list of already written lines for an
3120 entire SMTP session, which doesn't seem right. The memory is now reset if a
3121 new message is started.
3123 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3124 debugging output unless log_rewrite was also set.
3126 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3127 of (int)(handle) when we know that handle contains (void *)(-1).
3129 61. The Exim daemon panic-logs an error return when it closes the incoming
3130 connection. However "connection reset by peer" seems to be common, and
3131 isn't really an error worthy of noting specially, so that particular error
3134 62. When Exim is trying to find all the local interfaces, it used to panic and
3135 die if the ioctl to get the interface flags failed. However, it seems that
3136 on at least one OS (Solaris 9) it is possible to have an interface that is
3137 included in the list of interfaces, but for which you get a failure error
3138 for this call. This happens when the interface is not "plumbed" into a
3139 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3140 failure of the "get flags" call assumes that the interface is down.
3142 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3143 makes life easier for people who are doing arithmetic on fields extracted
3144 from dates, where you often get leading zeros that should not be
3145 interpreted as octal.
3147 64. Added qualify_domain to the redirect router, to override the global
3150 65. If a pathologically long header line contained very many addresses (the
3151 report of this problem mentioned 10 000) and each of them was rewritten,
3152 Exim could use up a very large amount of memory. (It kept on making new
3153 copies of the header line as it rewrote, and never released the old ones.)
3154 At the expense of a bit more processing, the header rewriting function has
3155 been changed so that it no longer eats memory in this way.
3157 66. The generation of the Received: header has been moved from the time that a
3158 message starts to be received, to the time that it finishes. The timestamp
3159 in the Received: header should now be very close to that of the <= log
3160 line. There are two side-effects of this change:
3162 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3163 logged header lines no longer include the local Received: line, because
3164 it has not yet been created. The same applies to a copy of the message
3165 that is returned to a non-SMTP sender when a message is rejected.
3167 (b) When a filter file is tested using -bf, no additional Received: header
3168 is added to the test message. After some thought, I decided that this
3171 This change does not affect the value of $received_for. It is still set
3172 after address rewriting, but before local_scan() is called.
3174 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3176 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3177 gave an unhelpful panic error message, and a defer error. I have managed to
3178 change this behaviour so that it now rejects any supplied certificate,
3179 which seems right, as the list of acceptable certificates is empty.
3181 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3182 gave an unhelpful defer error. I have not managed to make this reject any
3183 supplied certificates, but the error message it gives is "no certificate
3184 supplied", which is not helpful.
3186 70. exigrep's output now also includes lines that are not associated with any
3187 message, but which match the given pattern. Implemented by a patch from
3188 Martin Sluka, which also tidied up the Perl a bit.
3190 71. Recipient callout verification, like sender verification, was using <> in
3191 the MAIL FROM command. This isn't really the right thing, since the actual
3192 sender may affect whether the remote host accepts the recipient or not. I
3193 have changed it to use the actual sender in the callout; this means that
3194 the cache record is now keyed on a recipient/sender pair, not just the
3195 recipient address. There doesn't seem to be a real danger of callout loops,
3196 since a callout by the remote host to check the sender would use <>.
3197 [SEE ABOVE: changed after hitting problems.]
3199 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3200 temporary errors. However, in the case of such a code being given after
3201 the end of a data transmission (i.e. after ".") Exim was failing to write
3202 a retry record for the message. (Yes, there was some broken host that was
3203 actually sending 8xx at this point.)
3205 73. An unknown lookup type in a host list could cause Exim to panic-die when
3206 the list was checked. (An example that provoked this was putting <; in the
3207 middle of a list instead of at the start.) If this happened during a DATA
3208 ACL check, a -D file could be left lying around. This kind of configuration
3209 error no longer causes Exim to die; instead it causes a defer errror. The
3210 incident is still logged to the main and panic logs.
3212 74. Buglet left over from Exim 3 conversion. The message "too many messages
3213 in one connection" was written to the rejectlog but not the mainlog, except
3214 when address rewriting (yes!) was being logged.
3216 75. Added write_rejectlog option.
3218 76. When a system filter was run not as root (that is, when system_filter_user
3219 was set), the values of the $n variables were not being returned to the
3220 main process; thus, they were not subsequently available in the $sn
3223 77. Added +return_path_on_delivery log selector.
3225 78. A connection timeout was being treated differently from recipients deferred
3226 when testing hosts_max_try with a message that was older than the host's
3227 retry timeout. (The host should not be counted, thus allowing all hosts to
3228 be tried at least once before bouncing.) This may have been the cause of an
3229 occasionally reported bug whereby a message would remain on the queue
3230 longer than the retry timeout, but would be bounced if a delivery was
3231 forced. I say "may" because I never totally pinned down the problem;
3232 setting up timeout/retry tests is difficult. See also the next item.
3234 79. The ultimate address timeout was not being applied to errors that involved
3235 a combination of host plus message (for example, a timeout on a MAIL
3236 command). When an address resolved to a number of possible hosts, and they
3237 were not all tried for each delivery (e.g. because of hosts_max_try), a
3238 message could remain on the queue longer than the retry timeout.
3240 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3243 81. Fixed an obscure SMTP outgoing bug which required at least the following
3244 conditions: (a) there was another message waiting for the same server;
3245 (b) the server returned 5xx to all RCPT commands in the first message so
3246 that the message was not completed; (c) the server dropped the connection
3247 or gave a negative response to the RSET that Exim sends to abort the
3248 transaction. The observed case was a dropped connection after DATA that had
3249 been sent in pipelining mode. That is, the server had advertised PIPELINING
3250 but was not implementing it correctly. The effect of the bug was incorrect
3251 behaviour, such as trying another host, and this could lead to a crash.
3257 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3258 and daemon.c were passed as pointers to ints; they should have been
3259 pointers to socklen_t variables (which are typically unsigned ints).
3261 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3264 3. Fixed a really odd bug that affected only the testing scheme; patching a
3265 certain fixed string in the binary changed the value of another string that
3266 happened to be identical to the end of the original first string.
3268 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3269 name", it returns that address as the IP address. On some operating
3270 systems (e.g. Solaris), it also passes back the IP address string as the
3271 "host name". However, on others (e.g. Linux), it passes back an empty
3272 string. Exim wasn't checking for this, and was changing the host name to an
3273 empty string, assuming it had been canonicized.
3275 5. Although rare, it is permitted to have more than one PTR record for a given
3276 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3277 all the names associated with an address, because they do in Solaris.
3278 However, it seems that they do not in Linux for data that comes from the
3279 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3280 I found this out when I moved to a new Linux workstation and tried to run
3281 the Exim test suite.
3283 To get round this problem I have changed the code so that it now does its
3284 own call to the DNS to look up PTR records when searching for a host name.
3285 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3286 addresses that are only in /etc/hosts are still found.
3288 This behaviour is, however, controlled by an option called host_lookup_
3289 order, which defaults to "bydns:byaddr". If people want to use the other
3290 order, or indeed, just use one or the other means of lookup, they can
3291 specify it in this variable.
3293 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3294 some operating systems, this comes back from gethostbyaddr() as an empty
3295 string, and this is what Exim used to test for. However, it seems that in
3296 other systems, "." is yielded. Exim now tests for this case too.
3298 7. The values of check_spool_space and check_log_space are now held internally
3299 as a number of kilobytes instead of an absolute number of bytes. If a
3300 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3301 kilobyte. This means that much larger values can be stored.
3303 8. Exim monitor: an attempt to get the action menu when not actually pointing
3304 at a message produces an empty menu entitled "No message selected". This
3305 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3306 no entries in it ("Shell widget menu has zero width and/or height"). So I
3307 have added a single, blank menu entry in this case.
3309 9. Added ${quote_local_part.
3311 10. MIME decoding is now applied to the contents of Subject: header lines when
3314 11. Now that a reference to $sender_host_address automatically causes a reverse
3315 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3316 host lookup before query-style lookups in lists that might use this
3317 variable. This has therefore been abolished, and the "net-" prefix is no
3318 longer necessary for query-style lookups.
3320 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3321 have been a typo for LFLAGS, so it has been changed.
3323 13. The install script calls Exim with "-C /dev/null" in order to find the
3324 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3325 to be output. Howeve, since Exim outputs its version number before the
3326 error, it didn't break the script. It just looked ugly. I fixed this by
3327 always allowing "-C /dev/null" if the caller is root.
3329 14. Ignore overlarge ACL variable number when reading spool file - insurance
3330 against a later release with more variables having written the file.
3332 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3333 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3334 this, as well as the form without the "IPv6" on the front.
3336 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3337 OS/Makefile-Darwin file.
3339 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3340 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3342 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3343 variable. However, some OS use ints for inodes. I've added cast to long int
3344 to get rid of the compiler warning.
3346 19. I had forgotten to lock out "/../" in configuration file names when
3347 ALT_CONFIG_PREFIX was set.
3349 20. Routers used for verification do not need to specify transports. However,
3350 if such a router generated a host list, and callout was configured, Exim
3351 crashed, because it could not find a port number from the (non-existent)
3352 transport. It now assumes port 25 in this circumstance.
3354 21. Added the -t option to exigrep.
3356 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3357 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3358 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3359 them is set, LOOKUP_LSEARCH is forced.
3361 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3363 24. Added sender and host information to the "rejected by local_scan()" log
3364 line; previously there was no indication of these.
3366 25. Added .include_if_exists.
3368 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3369 Linux. It turns out that not all file systems support this. Apparently some
3370 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3371 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3372 not supported on the file descriptor, is now ignored when Exim is trying to
3373 sync a directory. This applies only to Linux.
3375 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3377 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3378 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3381 29. Added support for maildirsize files from supplied patch (modified a bit).
3383 30. The use of :fail: followed by an empty string could lead Exim to respond to
3384 sender verification failures with (e.g.):
3386 550 Verification failed for <xxx>
3387 550 Sender verify failed
3389 where the first response line was missing the '-' that indicates it is not
3390 the final line of the response.
3392 31. The loop for finding the name of the user that called Exim had a hardwired
3393 limit of 10; it now uses the value of finduser_retries, which is used for
3394 all other user lookups.
3396 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3399 33. Exim was neglecting to zero errno before one call of strtol() when
3400 expanding a string and expecting an integer value. On some systems this
3401 resulted in spurious "integer overflow" errors. Also, it was casting the
3402 result into an int without checking.
3404 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3405 did not work. The code looks as if it has *never* worked, though it appears
3406 to have been documented since at least releast 1.62. I have made it work.
3408 35. The "timeout_DNS" error in retry rules, also documented since at least
3409 1.62, also never worked. As it isn't clear exactly what this means, and
3410 clearly it isn't a major issue, I have abolished the feature by treating it
3411 as "timeout", and writing a warning to the main and panic logs.
3413 36. The display of retry rules for -brt wasn't always showing the error code
3416 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3417 timeout_connect_A, timeout_connect_MX.
3419 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3420 to the empty sender.
3422 39. The daemon was not analysing the content of -oX till after it had closed
3423 stderr and disconnected from the controlling terminal. This meant that any
3424 syntax errors were only noted on the panic log, and the return code from
3425 the command was 0. By re-arranging the code a little, I've made the
3426 decoding happen first, so such errors now appear on stderr, and the return
3427 code is 1. However, the actual setting up of the sockets still happens in
3428 the disconnected process, so errors there are still only recorded on the
3431 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3432 connections (as happens on some IP stacks) was logged at start up time as
3433 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3434 it from "IPv6 and IPv4", which means that two separate sockets are being
3437 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3438 says whether AF_INET or AF_INET6 was passed as an argument.
3440 42. Exiwhat output was messed up when time zones were included in log
3443 43. Exiwhat now gives more information about the daemon's listening ports,
3444 and whether -tls-on-connect was used.
3446 44. The "port" option of the smtp transport is now expanded.
3448 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3449 silently ignored. Now an error message is written to the main and panic
3452 46. There's a new ACL modifier called "logwrite" which writes to a log file
3453 as soon as it is encountered.
3455 47. Added $local_user_uid and $local_user_gid at routing time.
3457 48. Exim crashed when trying to verify a sender address that was being
3460 49. Exim was recognizing only a space character after ".include". It now also
3461 recognizes a tab character.
3463 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3464 extracting the relevant information from the specification. The man page no
3465 longer contains scrambled data for the -d option, and I've added a section
3466 at the front about calling Exim under different names.
3468 51. Added "extra_headers" argument to the "mail" command in filter files.
3470 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3473 53. Installed eximstats 1.29.
3475 54. Added transport_filter_timeout as a generic transport option.
3477 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3478 Cc: header lines. This was required by RFC 822, but it not required by RFC
3481 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3482 incoming messages that did not have them. Now it does so only if the
3483 message originates locally, that is, if there is no associated remote host
3484 address. When Resent- header lines are present, this applies to the Resent-
3485 lines rather than the non-Resent- lines.
3487 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3488 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3490 58. Messages for configuration errors now include the name of the main
3491 configuration file - useful now that there may be more than one file in a
3492 list (.included file names were always shown).
3494 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3495 for those rare installations that do not start the daemon as root or run it
3496 setuid root. I've cut out the call to initgroups() if the daemon is not
3499 60. The Exim user and group can now be bound into the binary as text strings
3500 that are looked up at the start of Exim's processing.
3502 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3504 62. Added $mailstore_basename variable.
3506 63. Installed patch to sieve.c from Michael Haardt.
3508 64. When Exim failed to open the panic log after failing to open the main log,
3509 the original message it was trying to log was written to stderr and debug
3510 output, but if they were not available (the usual case in production), it
3511 was lost. Now it is written to syslog before the two lines that record the
3512 failures to open the logs.
3514 65. Users' Exim filters run in subprocesses under the user's uid. It is
3515 possible for a "deliver" command or an alias in a "personal" command to
3516 provoke an address rewrite. If logging of address rewriting is configured,
3517 this fails because the process is not running as root or exim. There may be
3518 a better way of dealing with this, but for the moment (because 4.30 needs
3519 to be released), I have disabled address rewrite logging when running a
3520 filter in a non-root, non-exim process.
3526 1. The buildconfig auxiliary program wasn't quoting the value set for
3527 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3528 not defined. This bug was masked in 4.22 by the effect that was fixed in
3531 2. Some messages that were rejected after a message id was allocated were
3532 shown as "incomplete" by exigrep. It no longer does this for messages that
3533 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3535 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3536 have allow_domain_literals set, the ID did not get logged in the <= line.
3537 Domain literals are now always recognized in Message-ID: header lines.
3539 4. The first argument for a ${extract expansion item is the key name or field
3540 number. Leading and trailing spaces in this item were not being ignored,
3541 causing some misleading effects.
3543 5. When deliver_drop_privilege was set, single queue runner processes started
3544 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3545 same command in the process it spins off) were not dropping privilege.
3547 6. When the daemon running as "exim" started a queue runner, it always
3548 re-executed Exim in the spun-off process. This is a waste of effort when
3549 deliver_drop_privilege is set. The new process now just calls the
3550 queue-runner function directly.
3556 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3559 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3560 ignored. Though the use of -oP was forcing the writing of a pid file, it
3561 was always written to the default place.
3563 3. If the message "no IP address found for host xxxx" is generated during
3564 incoming verification, it is now followed by identification of the incoming
3565 connection (so you can more easily find what provoked it).
3567 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3569 5. Added some features to "harden" Exim a bit more against certain attacks:
3571 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3572 be put in Local/Makefile. This is like the never_users runtime option,
3573 but it cannot be overridden. The default setting is "root".
3575 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3576 prefix string with which any file named in a -C command line option
3579 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3580 is retained for -C and -D only if the caller of Exim is root. Without
3581 it, the exim user may also use -C and -D and retain privilege.
3583 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3584 command line option is disabled.
3586 6. Macro names set by the -D option must start with an upper case letter, just
3587 like macro names defined in the configuration file.
3589 7. Added "dereference=" facility to LDAP.
3591 8. Two instances of the typo "uknown" in the source files are fixed.
3593 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3594 the Configure-Makefile script screwed up while processing it.
3596 10. Incorporated PCRE 4.4.
3598 11. The SMTP synchronization check was not operating right at the start of an
3599 SMTP session. For example, it could not catch a HELO sent before the client
3600 waited for the greeting. There is now a check for outstanding input at the
3601 point when the greeting is written. Because of the duplex, asynchronous
3602 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3603 way, but not yet received, when the check is performed.
3605 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3606 on TCP/IP sockets, because this apparently causes some broken clients to
3609 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3610 unchanged) from the Cygwin maintainer.
3612 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3613 was supported instead of testing for mbx. Effectively a typo.
3615 15. The spa authenticator server code was not checking that the input it
3616 received was valid base64.
3618 16. The debug output line for the "set" modifier in ACLs was not showing the
3619 name of the variable that was being set.
3621 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3623 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3624 lookup on its own. Something else was needed to trigger the lookup. For
3625 example, a match in host_lookup or the need for a host name in a host list.
3626 Now, if $sender_host_name is referenced and the host name has not yet been
3627 looked up, a lookup is performed. If the lookup fails, the variable remains
3628 empty, and $host_lookup_failed is set to "1".
3630 19. Added "eqi" as a case-independent comparison operator.
3632 20. The saslauthd authentication condition could segfault if neither service
3633 nor realm was specified.
3635 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3636 error message that was logged was misleading, and incoming SMTP
3637 connections were dropped. The message is now more accurate, and temporary
3638 errors are given to SMTP connections.
3640 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3641 (except RSET) with a 421 error, until QUIT is received. However, it was
3642 failing to send a response to QUIT.
3644 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3645 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3646 ACL is now run after the helo_try_verify_hosts code.
3648 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3649 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3650 case-independent, so other case variants are also recognized). Apparently
3651 some systems use these upper case variants.
3653 25. If more than two messages were waiting for the same host, and a transport
3654 filter was specified for the transport, Exim sent two messages over the
3655 same TCP/IP connection, and then failed with "socket operation on non-
3656 socket" when it tried to send the third.
3658 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3660 27. The extern definition of crypt16() in expand.c was not being excluded when
3661 the OS had its own crypt16() function.
3663 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3664 as a preferred synonym for return_size_limit, both as an option and as an
3667 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3669 30. Changed the default configuration ACL to relax the local part checking rule
3670 for addresses that are not in any local domains. For these addresses,
3671 slashes and pipe symbols are allowed within local parts, but the sequence
3672 /../ is explicitly forbidden.
3674 31. SPA server authentication was not clearing the challenge buffer before
3677 32. log_message in a "warn" ACL statement was writing to the reject log as
3678 well as to the main log, which contradicts the documentation and doesn't
3679 seem right (because no rejection is happening). So I have stopped it.
3681 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3682 However, I am unable to do any testing of this.
3684 34. Fixed an infelicity in the appendfile transport. When checking directories
3685 for a mailbox, to see if any needed to be created, it was accidentally
3686 using path names with one or more superfluous leading slashes; tracing
3687 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3689 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3690 contents are added to the log line that is written for every discarded
3691 recipient. (Previously a log_message setting was ignored.)
3693 36. The ${quote: operator now quotes the string if it is empty.
3695 37. The install script runs exim in order to find its version number. If for
3696 some reason other than non-existence or emptiness, which it checks, it
3697 could not run './exim', it was installing it with an empty version number,
3698 i.e. as "exim-". This error state is now caught, and the installation is
3701 38. An argument was missing from the function that creates an error message
3702 when Exim fails to connect to the socket for saslauthd authentication.
3703 This could cause Exim to crash, or give a corrupted message.
3705 39. Added isip, isip4, and isip6 to ${if conditions.
3707 40. The ACL variables $acl_xx are now saved with the message, and can be
3708 accessed later in routers, transports, and filters.
3710 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3711 strings in the file are not string-expanded.
3713 42. If a MAIL command specified a SIZE value that was too large to fit into an
3714 int variable, the check against message_size_limit failed. Such values are
3715 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3716 day this will have to be increased, but I don't think I want to be around
3717 when emails are that large.
3724 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3725 iconv() is not standard in FreeBSD.
3727 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3728 IPv6 enabled. The observed symptom was a segmentation fault on return from
3729 the function os_common_find_running_interfaces() in src/os.c.
3731 3. In the check_special_case() function in daemon.c I had used "errno" as an
3732 argument name, which causes warnings on some systems. This was basically a
3733 typo, since it was named "eno" in the comments!
3735 4. The code that waits for the clock to tick (at a resolution of some fraction
3736 of a second) so as to ensure message-id uniqueness was always waiting for
3737 at least one whole tick, when it could have waited for less. [This is
3738 almost certainly not relevant at current processor speeds, where it is
3739 unlikely to ever wait at all. But we try to future-proof.]
3741 5. The function that sleeps for a time interval that includes fractions of a
3742 second contained a race. It did not block SIGALRM between setting the
3743 timer, and suspending (a couple of lines later). If the interval was short
3744 and the sigsuspend() was delayed until after it had expired, the suspension
3745 never ended. On busy systems this could lead to processes getting stuck for
3748 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3749 process, before it forks any delivery processes. The open lookup caching
3750 mechanism meant that the open file or database connection was passed into
3751 the delivery process. The problem was that delivery processes always tidy
3752 up cached lookup data. This could cause a problem for the next delivery
3753 process started by the queue runner, because the external queue runner
3754 process does not know about the closure. So the next delivery process
3755 still has data in the lookup cache. In the case of a file lookup, there was
3756 no problem because closing a file descriptor in a subprocess doesn't affect
3757 the parent. However, if the lookup was caching a connection to a database,
3758 the connection was closed, and the second delivery process was likely to
3759 see errors such as "PGSQL: query failed: server closed the connection
3760 unexpectedly". The problem has been fixed by closing all cached lookups
3761 in a queue runner before running a delivery process.
3763 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3764 seem to have the "const" qualifier which it has on other OS. I've
3767 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3768 *of the same type* (client or server) with the same public name that an
3769 error should be diagnosed.
3771 9. When Exim looked up a host name for an IP address, but failed to find the
3772 original IP address when looking up the host name (a safety check), it
3773 output the message "<ip address> does not match any IP for NULL", which was
3774 confusing, to say the least. The bug was that the host name should have
3775 appeared instead of "NULL".
3777 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3778 user that is built into the binary, and the -C or -D options is used, root
3779 privilege is dropped before the configuration file is read. In addition,
3780 logging is switched to stderr instead of the normal log files. If the
3781 configuration then re-defines the Exim user, the unprivileged environment
3782 is probably not what is expected, so Exim logs a panic warning message (but
3785 However, if deliver_drop_privilege is set, the unprivileged state may well
3786 be exactly what is intended, so the warning has been cut out in that case,
3787 and Exim is allowed to try to write to its normal log files.
3793 1. smtp_return_error_details was not giving details for temporary sender
3794 or receiver verification errors.
3796 2. Diagnose a configuration error if two authenticators have the same public
3799 3. Exim used not to create the message log file for a message until the first
3800 delivery attempt. This could be confusing when incoming messages were held
3801 for policy or load reasons. The message log file is now created at the time
3802 the message is received, and an initial "Received" line is written to it.
3804 4. The automatically generated man page for command line options had a minor
3805 bug that caused no ill effects; however, a more serious problem was that
3806 the procedure for building the man page automatically didn't always
3807 operate. Consequently, release 4.20 contains an out-of-date version. This
3808 shouldn't happen again.
3810 5. When building Exim with embedded Perl support, the script that builds the
3811 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3812 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3814 6. The freeze_tell option was not being used for messages that were frozen on
3815 arrival, either by an ACL or by local_scan().
3817 7. Added the smtp_incomplete_transaction log selector.
3819 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3820 was accepting AUTH without a new EHLO.
3822 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3823 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3826 10. Logging of TCP/IP connections (when configured) now happens in the main
3827 daemon process instead of the child process, so that the TCP/IP connection
3828 count is more accurate (but it can never be perfect).
3830 11. The use of "drop" in a nested ACL was not being handled correctly in the
3831 outer ACL. Now, if condition failure induced by the nested "drop" causes
3832 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3833 or "require"), the connection is dropped.
3835 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3836 that yield "discard" can now be used with an "accept" or a "discard" verb,
3837 but an error is generated for any others (because I can't see a useful way
3838 to define what should happen).
3840 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3841 are now processed in the same way as lines in the Exim configuration file.
3842 In particular, continuation lines are supported.
3844 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3846 15. Added -ti meaning -t -i.
3848 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3849 domains, and warn by logging if others are found.
3851 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3852 returned from GIFCONF when Exim is trying to find the list of interfaces on
3853 a host. The code in os.c has been modified to copy each ifreq to an aligned
3854 structure in all cases.
3856 Also, in some cases, the returned ifreq's were being copied to a 'struct
3857 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3858 means the last couple of bytes of an IPv6 address could be chopped if the
3859 ifreq contained only a normal sockaddr (14 bytes storage).
3861 18. Named domain lists were not supported in the hosts_treat_as_local option.
3862 An entry such as +xxxx was not recognized, and was treated as a literal
3865 19. Ensure that header lines added by a DATA ACL are included in the reject log
3866 if the ACL subsequently rejects the message.
3868 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3869 MD5 (which is deprecated).
3871 21. When testing a filter file using -bf, Exim was writing a message when it
3872 took the sender from a "From " line in the message, but it was not doing so
3873 when it took $return_path from a Return-Path: header line. It now does.
3875 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3876 with a valid header line field name (a series of printing characters
3877 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3879 23. Changed "disc" in the source to "disk" to conform to the documentation and
3880 the book and for uniformity.
3882 24. Ignore Sendmail's -Ooption=value command line item.
3884 25. When execve() failed while trying to run a command in a pipe transport,
3885 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3886 could be confused with a return value of 69 from the command itself. This
3887 has been changed to 127, the value the shell returns if it is asked to run
3888 a non-existent command. The wording for the related log line suggests a
3889 non-existent command as the problem.
3891 26. If received_header_text expands to an empty string, do not add a Received:
3892 header line to the message. (Well, it adds a token one on the spool, but
3893 marks it "old" so that it doesn't get used or transmitted.)
3895 27. Installed eximstats 1.28 (addition of -nt option).
3897 28. There was no check for failure on the call to getsockname() in the daemon
3898 code. This can fail if there is a shortage of resources on the system, with
3899 ENOMEM, for example. A temporary error is now given on failure.
3901 29. Contrary to the C standard, it seems that in some environments, the
3902 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3903 program. Exim now does this explicitly; it affects the formatting of
3904 timestamps using strftime().
3906 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3907 complaints. I've now initialized all the variables, to avoid this.
3909 32. Header lines added by a system filter were not being "seen" during
3910 transport-time rewrites.
3912 33. The info_callback() function passed to OpenSSL is set up with type void
3913 (*)(SSL *, int, int), as described somewhere. However, when calling the
3914 function (actually a macro) that sets it up, the type void(*)() is
3915 expected. I've put in a cast to prevent warnings from picky compilers.
3917 34. If a DNS black list lookup found a CNAME record, but there were no A
3918 records associated with the domain it pointed at, Exim crashed.
3920 35. If a DNS black list lookup returned more than one A record, Exim ignored
3921 all but the first. It now scans all returned addresses if a particular IP
3922 value is being sought. In this situation, the contents of the
3923 $dnslist_value variable are a list of all the addresses, separated by a
3926 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3927 to accept a host name and all its aliases if the forward lookup for any of
3928 them yielded the IP address of the incoming connection. Now it accepts only
3929 those names whose forward lookup yields the correct IP address. Any other
3930 names are discarded. This closes a loophole whereby a rogue DNS
3931 administrator could create reverse DNS records to break through a
3932 wildcarded host restriction in an ACL.
3934 37. If a user filter or a system filter that ran in a subprocess used any of
3935 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3936 the wrong values were passed to the pipe command ($thisaddress had the
3937 value of $0, $0 had the value of $1, etc). This bug was introduced by
3938 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3940 38. Improved the line breaking for long SMTP error messages from ACLs.
3941 Previously, if there was no break point between 40 and 75 characters, Exim
3942 left the rest of the message alone. Two changes have been made: (a) I've
3943 reduced the minimum length to 35 characters; (b) if it can't find a break
3944 point between 35 and 75 characters, it looks ahead and uses the first one
3945 that it finds. This may give the occasional overlong line, but at least the
3946 remaining text gets split now.
3948 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3949 file descriptors might be low, and that setting 1000 would always raise it.
3950 It turns out that in some environments, the limit is already over 1000 and
3951 that lowering it causes trouble. So now Exim takes care not to decrease it.
3953 40. When delivering a message, the value of $return_path is set to $sender_
3954 address at the start of routing (routers may change the value). By an
3955 oversight, this default was not being set up when an address was tested by
3956 -bt or -bv, which affected the outcome if any router or filter referred to
3959 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3960 log only when "message" or "log_message" are set. However, if one of the
3961 conditions was an address verification, or a call to a nested ACL, the
3962 messages generated by the underlying test were being passed through. This
3963 no longer happens. The underlying message is available in $acl_verify_
3964 message for both "message" and "log_message" expansions, so it can be
3965 passed through if needed.
3967 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3968 new expansion $bh_ to give the encoded byte string without charset
3969 translation. Translation happens only if iconv() is available; HAVE_ICONV
3970 indicates this at build time. HEADERS_CHARSET gives the charset to
3971 translate to; headers_charset can change it in the configuration, and
3972 "headers charset" can change it in an individual filter file.
3974 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3975 that creates RFC 2047 encoded "words" labels them as that charset instead
3976 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3977 expansion operator; (ii) when Exim creates a From: line for a local
3978 message; (iii) when a header line is rewritten to include a "phrase" part.
3980 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3981 buggy, causing it to skip the first lines of messages whose message ID
3982 ended in 'D'. This would not have bitten before Exim release 4.14, because
3983 message IDs were unlikely to end in 'D' before then. The effect was to have
3984 incorrect size information for certain domains.
3986 45. #include "config.h" was missing at the start of the crypt16.c module. This
3987 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3990 46. If there was a timeout during a "random" callout check, Exim treated it as
3991 a failure of the random address, and carried on sending RSET and the real
3992 address. If the delay was just some slowness somewhere, the response to the
3993 original RCPT would be taken as a response to RSET and so on, causing
3994 mayhem of various kinds.
3996 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3997 when I implemented it. It didn't allow for the fact that some option values
3998 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3999 the right test for positive values.
4001 48. Domain names in DNS records are case-independent. Exim always looks them up
4002 in lower case. Some resolvers return domain names in exactly the case they
4003 appear in the zone file, that is, they may contain uppercase letters. Not
4004 all resolvers do this - some return always lower case. Exim was treating a
4005 change of case by a resolver as a change of domain, similar to a widening
4006 of a domain abbreviation. This triggered its re-routing code and so it was
4007 trying to route what was effectively the same domain again. This normally
4008 caused routing to fail (because the router wouldn't handle the domain
4009 twice). Now Exim checks for this case specially, and just changes the
4010 casing of the domain that it ultimately uses when it transmits the message
4013 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4016 50. If a filter generated a file delivery with a non-absolute name (possible if
4017 no home directory exists for the router), the forbid_file option was not
4020 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4021 the existing equality matching.
4023 52. Exim was using ints instead of ino_t variables in some places where it was
4024 dealing with inode numbers.
4026 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4027 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4028 called TMPDIR, and if it finds it is different, it changes its value.
4030 54. The smtp_printf() function is now made available to local_scan() so
4031 additional output lines can be written before returning. There is also an
4032 smtp_fflush() function to enable the detection of a dropped connection.
4033 The variables smtp_input and smtp_batched_input are exported to
4036 55. Changed the default runtime configuration: the message "Unknown user"
4037 has been removed from the ACL, and instead placed on the localuser router,
4038 using the cannot_route_message feature. This means that any verification
4039 failures that generate their own messages won't get overridden. Similarly,
4040 the "Unrouteable address" message that was in the ACL for unverifiable
4041 relay addresses has also been removed.
4043 56. Added hosts_avoid_esmtp to the smtp transport.
4045 57. The exicyclog script was not checking for the esoteric option
4046 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4047 will work only if exicyclog is run under the appropriate euid.
4049 58. Following a discussion on the list, the rules by which Exim recognises line
4050 endings on incoming messages have been changed. The -dropcr and drop_cr
4051 options are now no-ops, retained only for backwards compatibility. The
4052 following line terminators are recognized: LF CRLF CR. However, special
4053 processing applies to CR:
4055 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4056 nor a local message in the state where . is a terminator.
4058 (ii) If a bare CR is encountered in a header line, an extra space is added
4059 after the line terminator so as not to end the header. The reasoning
4060 behind this is that bare CRs in header lines are most likely either
4061 to be mistakes, or people trying to play silly games.
4063 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4064 was being incorrectly given as 18 bytes larger than it should have been.
4065 This is a VOB (very old bug).
4067 60. This may never have affected anything current, but just in case it has:
4068 When the local host is found other than at the start of a list of hosts,
4069 the local host, those with the same MX, and any that follow, are discarded.
4070 When the list in question was part of a longer list of hosts, the following
4071 hosts (not currently being processed) were also being discarded. This no
4072 longer happens. I'm not sure if this situation could ever has previously
4075 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4078 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4079 header. This is something that is recommended in a new Internet Draft, and
4080 is something that is documented as being done by Sendmail. There are two
4081 possible values. For messages generated by the autoreply transport, Exim
4084 Auto-Submitted: auto-replied
4086 whereas for all other generated messages (e.g. bounces) it adds
4088 Auto-Submitted: auto-generated
4090 63. The "personal" condition in filters now includes a test for the
4091 Auto-Submitted: header. If it contains the string "auto-" the message it
4092 not considered personal.
4094 64. Added rcpt_include_affixes as a generic transport option.
4096 65. Added queue_only_override (default true).
4098 66. Added the syslog_duplication option.
4100 67. If what should have been the first header line of a message consisted of
4101 a space followed by a colon, Exim was mis-interpreting it as a header line.
4102 It isn't of course - it is syntactically invalid and should therefore be
4103 treated as the start of the message body. The misbehaviour could have
4104 caused a number of strange effects, including loss of data in subsequent
4105 header lines, and spool format errors.
4107 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4108 client host had authenticated. This control can now be exercised by an ACL
4109 for more flexibility.
4111 69. By default, callouts do not happen when testing with -bh. There is now a
4112 variant, -bhc, which does actually run the callout code, including
4113 consulting and updating the callout cache.
4115 70. Added support for saslauthd authentication, courtesy of Alexander
4118 71. If statvfs() failed on the spool or log directories while checking their
4119 size for availability, Exim confusingly gave the error "space shortage".
4120 Furthermore, in debugging mode it crashed with a floating point exception.
4121 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4122 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4123 serious problem, Exim now writes to the main and panic logs when this
4124 happens, with details of the failure. It then refuses to accept the
4125 incoming message, giving the message "spool directory problem" or "log
4126 directory problem" with a 421 code for SMTP messages.
4128 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4129 0, 1, and 2 exist, because some OS complain for execs without them (see
4130 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4131 descriptors. However, the code omitted to check that the open succeeded,
4132 causing mysterious errors if for some reason the permissions on /dev/null
4133 got screwed. Now Exim writes a message to the main and panic logs, and
4134 bombs out if it can't open /dev/null.
4136 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4137 interact so that it is all more flexible. It is supposed to remain
4138 backwards compatible. Also added extra_local_interfaces.
4140 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4141 to bomb out with an assertion failure - to the client this appears as a
4142 connection drop. This problem occurs in the part of the code that was taken
4143 from the Samba project. Fortunately, the assertion is in a very simple
4144 function, so I have fixed this by reproducing the function inline in the
4145 one place where it is called, and arranging for authentication to fail
4146 instead of killing the process with assert().
4148 75. The SPA client code was not working when the server requested OEM rather
4149 than Unicode encoding.
4151 76. Added code to make require_files with a specific uid setting more usable in
4152 the case where statting the file as root fails - usually a non-root-mounted
4153 NFS file system. When this happens and the failure is EACCES, Exim now
4154 forks a subprocess and does the per-uid checking as the relevant uid.
4156 77. Added process_log_path.
4158 78. If log_file_path was not explicitly set, a setting of check_log_space or
4159 check_log_inodes was ignored.
4161 79. If a space check for the spool or log partitions fails, the incident is now
4162 logged. Of course, in the latter case the data may get lost...
4164 80. Added the %p formatting code to string_format() so that it can be used to
4165 print addresses in debug_print(). Adjusted all the address printing in the
4166 debugging in store.c to use %p rather than %d.
4168 81. There was a concern that a line of code in smtp_in.c could overflow a
4169 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4170 initially expressed, the concern was not well-founded, because trailing
4171 spaces are removed early. However, if the trailing spaces were followed by
4172 a NULL, they did not get removed, so the overflow was possible. Two fixes
4175 (a) I re-wrote the offending code in a cleaner fashion.
4176 (b) If an incoming SMTP command contains a NULL character, it is rejected
4179 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4180 runs initgroups(), so that if the Exim user is in any additional groups,
4181 they will be used during message reception.
4187 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob