1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.395 2006/09/19 20:01:13 jetmore Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Efrurth's refactoring patch to tidy up spool_mbox.c.
39 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
41 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
43 PH/08 An error is now given if message_size_limit is specified negative.
45 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
46 to be given (somewhat) arbitrary names.
48 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
51 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
52 miscellaneous code fixes
58 SC/01 Use a glob alias rather than an array ref in eximstats generated
59 parser. This improves both readability and performance.
61 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
62 Don't display local sender or destination tables in eximstats unless
63 there is data to show.
64 Added average volumes into the eximstats top table text output.
66 SC/03 Collect data on the number of addresses (recipients) as well
67 as the number of messages in eximstats.
69 TF/01 Correct an error in the documentation for the redirect router. Exim
70 does (usually) call initgroups() when daemonizing.
72 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
73 with consistent privilege compared to when running as a daemon.
75 TF/03 Note in the spec that $authenticated_id is not set for local
76 submissions from trusted users.
78 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
79 Thanks to Dean Brooks <dean@iglou.com> for the patch.
81 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
82 by adding some example configuration directives to the default
83 configuration file. A little bit of work is required to uncomment the
84 directives and define how usernames and passwords are checked, but
85 there is now a framework to start from.
87 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
88 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
89 without this. I don't know how relevant this is to other LDAP libraries.
91 PH/02 Add the verb name to the "unknown ACL verb" error.
93 PH/03 Magnus Holmgren's patch for filter_prepend_home.
95 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
97 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
98 directory not expanded when it should be if an expanded home directory
99 was set for the address (which is overridden by the transport).
101 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
104 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
105 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
106 because it is too late at that time, and has no effect.
108 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
109 security issue with \' (bugzilla #107). I could not use the
110 PQescapeStringConn() function, because it needs a PGconn value as one of
113 PH/08 When testing addresses using -bt, indicate those final addresses that
114 are duplicates that would not cause an additional delivery. At least one
115 person was confused, thinking that -bt output corresponded to deliveries.
116 (Suppressing duplicates isn't a good idea as you lose the information
117 about possibly different redirections that led to the duplicates.)
119 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
120 systems where poll() doesn't work, in particular OS X.
122 PH/10 Added more information to debugging output for retry time not reached.
124 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
125 operations in malware.c.
127 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
130 PH/13 If write_rejectlog was set false when logging was sent to syslog with
131 syslog_duplication set false, log lines that would normally be written
132 both the the main log and to the reject log were not written to syslog at
135 PH/14 In the default configuration, change the use of "message" in ACL warn
136 statements to "add_header".
138 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
139 not followed by a command (e.g. "seen endif").
141 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
142 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
145 PH/17 Added extra conditions to the default value of delay_warning_condition
149 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
150 { match{$h_precedence:}{(?i)bulk|list|junk} } \
151 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
154 The Auto-Submitted: and various List- headers are standardised, whereas I
155 don't think Precedence: ever was.
157 PH/18 Refactored debugging code in route_finduser() to show more information,
158 in particular, the error code if getpwnam() issues one.
160 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
161 This is apparently needed in addition to the PH/07 change above to avoid
162 any possible encoding problems.
164 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
165 but not after initializing Perl.
167 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
168 output them only if debugging. By default they are written stderr,
169 apparently, which is not desirable.
171 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
174 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
177 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
179 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
180 authenticated or an ident call has been made. Suppress the default
181 values for $authenticated_id and $authenticated_sender (but permit -oMai
182 and -oMas) when testing with -bh.
184 PH/24 Re-jigged the order of the tests in the default configuration so that the
185 tests for valid domains and recipients precede the DNS black list and CSA
186 tests, on the grounds that those ones are more expensive.
188 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
189 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
190 This bug exists in every version of Exim that I still have, right back to
193 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
194 However, an attempt to turn on case-sensitivity in a regex key by
195 including (?-i) didn't work because the subject string was already
196 lowercased, and the effects were non-intuitive. It turns out that a
197 one-line patch can be used to allow (?-i) to work as expected.
203 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
204 other effects) broke the use of negated acl sub-conditions.
206 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
209 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
210 "Deny" causes Exim to reject the incoming connection with a 554 error.
211 Unfortunately, if there is a major crisis, such as a disk failure,
212 tcp-wrappers gives "deny", whereas what one would like would be some
213 kind of temporary error. A kludge has been added to help with this.
214 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
215 554 error is used if errno is still zero or contains ENOENT (which occurs
216 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
219 PH/03 Add -lutil to the default FreeBSD LIBS setting.
221 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
222 errors. Otherwise a message that provokes a temporary error (when other
223 messages do not) can cause a whole host to time out.
225 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
226 addresses were routed directly to files or pipes from a redirect router.
227 File deliveries just didn't batch; pipe deliveries might have suffered
230 PH/06 A failure to get a lock for a hints database would erroneously always say
231 "Failed to get write lock", even when it was really a read lock.
233 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
234 of 0600. This has been changed to use the value of the lockfile_mode
235 option (which defaults to 0600).
237 PH/08 Applied small patch from the Sieve maintainer.
239 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
240 folder from quota calculations, a direct delivery into this folder messed
241 up the contents of the maildirsize file. This was because the regex was
242 used only to exclude .Trash (or whatever) when the size of the mailbox
243 was calculated. There was no check that a delivery was happening into an
244 excluded directory. This bug has been fixed by ignoring all quota
245 processing for deliveries into excluded directories.
247 PH/10 Added the maildirfolder_create_regex option to appendfile.
253 PH/01 The code for finding all the local interface addresses on a FreeBSD
254 system running IPv6 was broken. This may well have applied to all BSD
255 systems, as well as to others that have similar system calls. The broken
256 code found IPv4 interfaces correctly, but gave incorrect values for the
257 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
258 that it would not match correctly against @[] and not recognize the IPv6
261 PH/02 The ipliteral router was not recognizing addresses of the form user@
262 [ipv6:....] because it didn't know about the "ipv6:" prefix.
264 PH/03 Added disable_ipv6.
266 PH/04 Changed $reply_address to use the raw form of the headers instead of the
267 decoded form, because it is most often used to construct To: headers
268 lines in autoreplies, and the decoded form may well be syntactically
269 invalid. However, $reply_address has leading white space removed, and all
270 newlines turned into spaces so that the autoreply transport does not
273 PH/05 If group was specified without a user on a router, and no group or user
274 was specified on a transport, the group from the router was ignored.
276 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
277 for visible compile-time settings that can be used to change these
278 numbers, for those that want even more. Backwards compatibility with old
279 spool files has been maintained. However, going back to a previous Exim
280 release will lost any variables that are in spool files.
282 PH/07 Two small changes when running in the test harness: increase delay when
283 passing a TCP/IP connection to a new process, in case the original
284 process has to generate a bounce, and remove special handling of
285 127.0.0.2 (sic), which is no longer necessary.
287 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
288 be the same on different OS.
290 PH/09 Moved a debug statement in filter processing to avoid a race problem when
293 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
294 whether --show-vars was specified or not
296 JJ/02 exipick: Added support for new ACL variable spool format introduced
299 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
300 syntactically invalid From: or Reply-to: line, and a filter used this to
301 generate an autoreply, and therefore failed to obtain an address for the
302 autoreply, Exim could try to deliver to a non-existent relative file
303 name, causing unrelated and misleading errors. What now happens is that
304 it logs this as a hard delivery error, but does not attempt to create a
307 PH/11 The exinext utility has a -C option for testing purposes, but although
308 the given file was scanned by exinext itself; it wasn't being passed on
309 when Exim was called.
311 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
312 an end-of-file indication when reading a command response.
314 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
315 compiled. In many other places in Exim, IPv6 addresses are always
316 recognized, so I have changed this. It also means that IPv4 domain
317 literals of the form [IPV4:n.n.n.n] are now always recognized.
319 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
320 used if the router is not running as root, for example, when verifying at
321 ACL time, or when using -bh. The debugging output from this situation was
322 non-existent - all you got was a failure to exec. I have made two
325 (a) Failures to set uid/gid, the current directory, or a process leader
326 in a subprocess such as that created by queryprogram now generate
327 suitable debugging ouput when -d is set.
329 (b) The queryprogram router detects when it is not running as root,
330 outputs suitable debugging information if -d is set, and then runs
331 the subprocess without attempting to change uid/gid.
333 PH/15 Minor change to Makefile for building test_host (undocumented testing
336 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
337 additional section of a DNS packet that returns MX or SRV records.
338 Instead, it always explicitly searches for A/AAAA records. This avoids
339 major problems that occur when a DNS server includes only records of one
340 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
341 fixed another bug: if SRV records were looked up and the corresponding
342 address records were *not* found in the additional section, the port
343 values from the SRV records were lost.
345 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
346 using the correct key (the original address) when searching the retry
347 rules in order to find which one to use for generating the retry hint.
349 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
350 adding the default one. Similarly, if it contains a Reply-To: header, the
351 errors_reply_to option, if set, is not used.
353 PH/19 When calculating a retry time, Exim used to measure the "time since
354 failure" by looking at the "first failed" field in the retry record. Now
355 it does not use this if it is later than than the arrival time of the
356 message. Instead it uses the arrival time. This makes for better
357 behaviour in cases where some deliveries succeed, thus re-setting the
358 "first failed" field. An example is a quota failure for a huge message
359 when small messages continue to be delivered. Without this change, the
360 "time since failure" will always be short, possible causing more frequent
361 delivery attempts for the huge message than are intended.
362 [Note: This change was subsequently modified - see PH/04 for 4.62.]
364 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
365 $1, $2, $3) because the numerical variables can be reset during some
366 expansion items (e.g. "match"), thereby losing the authentication data.
368 PH/21 Make -bV show the size of off_t variables so that the test suite can
369 decide whether to run tests for quotas > 2G.
371 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
372 mailbox_size, and mailbox_filecount in the appendfile transport. If a
373 filecount value is greater than 2G or if a quota value is greater than 2G
374 on a system where the size of off_t is not greater than 4, a panic error
377 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
378 never match. The debug and -bh output now contains an explicit error
379 message indicating a malformed IPv4 address or mask.
381 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
382 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
385 PH/25 Do not write to syslog when running in the test harness. The only
386 occasion when this arises is a failure to open the main or panic logs
387 (for which there is an explicit test).
389 PH/26 Added the /no_tell option to "control=freeze".
391 PH/27 If a host name lookup failed very early in a connection, for example, if
392 the IP address matched host_lookup and the reverse lookup yielded a name
393 that did not have a forward lookup, an error message of the form "no IP
394 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
395 could be logged. Now it outputs the IP address instead of "NULL".
397 PH/28 An enabling patch from MH: add new function child_open_exim2() which
398 allows the sender and the authenticated sender to be set when
399 submitting a message from within Exim. Since child_open_exim() is
400 documented for local_scan(), the new function should be too.
402 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
403 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
404 results in an empty string is now treated as unset.
406 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
408 PH/31 Added sender_verify_fail logging option.
410 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
411 needed by RFC 822 but not by RFC 2822 was commented out. I have now
412 tidied the source and removed it altogether.
414 PH/33 When a queue run was abandoned because the load average was too high, a
415 log line was always written; now it is written only if the queue_run log
416 selector is set. In addition, the log line for abandonment now contains
417 information about the queue run such as the pid. This is always present
418 in "start" and "stop" lines but was omitted from the "abandon" line.
420 PH/34 Omit spaces between a header name and the colon in the error message that
421 is given when verify = headers_syntax fails (if there are lots of them,
422 the message gets confusing).
424 PH/35 Change the default for dns_check_names_pattern to allow slashes within
425 names, as there are now some PTR records that contain slashes. This check
426 is only to protect against broken name servers that fall over on strange
427 characters, so the fact that it applies to all lookups doesn't matter.
429 PH/36 Now that the new test suite is complete, we can remove some of the
430 special code in Exim that was needed for the old test suite. For example,
431 sorting DNS records because real resolvers return them in an arbitrary
432 order. The new test suite's fake resolver always returns records in the
435 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
436 bounces) except when queue_only is set, to avoid logging races between
437 the different processes.
439 PH/38 Panic-die if .include specifies a non-absolute path.
441 PH/39 A tweak to the "H" retry rule from its user.
443 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
444 a label. They prevented compilation on older perls.
446 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
447 a warning to be raised on newish perls.
449 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
450 on queue. Changes to match documented behaviour of showing count of
451 messages matching specified criteria.
453 PH/40 Changed the default ident timeout from 30s to 5s.
455 PH/41 Added support for the use of login_cap features, on those BSD systems
456 that have them, for controlling the resources used by pipe deliveries.
458 PH/42 The content-scanning code uses fopen() to create files in which to put
459 message data. Previously it was not paying any attention to the mode of
460 the files. Exim runs with umask(0) because the rest of the code creates
461 files with open(), and sets the required mode explicitly. Thus, these
462 files were ending up world-writeable. This was not a big issue, because,
463 being within the spool directory, they were not world-accessible. I have
464 created a function called modefopen, which takes an additional mode
465 argument. It sets umask(777), creates the file, chmods it to the required
466 mode, then resets the umask. All the relevant calls to fopen() in the
467 content scanning code have been changed to use this function.
469 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
470 to 24 hours. This avoids potential overflow problems when processing G
471 and H retry rules. I suspect nobody ever tinkers with this value.
473 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
475 PH/45 When the plaintext authenticator is running as a client, the server's
476 challenges are checked to ensure they are valid base64 strings. By
477 default, the authentication attempt is cancelled if an invalid string is
478 received. Setting client_ignore_invalid_base64 true ignores these errors.
479 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
480 they are received. Thus, the responses can be made to depend on the
481 challenges. If an invalid string is ignored, an empty string is placed in
484 PH/46 Messages that are created by the autoreply transport now contains a
485 References: header, in accordance with RFCs 2822 and 3834.
487 PH/47 Added authenticated_sender_force to the smtp transport.
489 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
491 PH/49 Installed latest patch from the Sieve maintainer.
493 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
494 was also set, the appendfile transport was unnecessarily scanning a
495 directory of message files (e.g. for maildir delivery) to find the count
496 of files (along with the size), even though it did not need this
497 information. It now does the scan only if it needs to find either the
498 size of the count of files.
500 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
502 PH/52 Two bugs concerned with error handling when the smtp transport is
505 (i) Exim was not creating retry information for temporary errors given
506 for individual recipients after the DATA command when the smtp transport
507 was used in LMTP mode. This meant that they could be retried too
508 frequently, and not timed out correctly.
510 (ii) Exim was setting the flag that allows error details to be returned
511 for LMTP errors on RCPT commands, but not for LMTP errors for individual
512 recipients that were returned after the DATA command.
514 PH/53 This is related to PH/52, but is more general: for any failing address,
515 when detailed error information was permitted to be returned to the
516 sender, but the error was temporary, then after the final timeout, only
517 "retry timeout exceeded" was returned. Now it returns the full error as
518 well as "retry timeout exceeded".
520 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
521 do this, and (what is worse) MTAs that accept it.
523 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
524 will now be deprecated.
526 PH/56 New os.c-cygwin from the Cygwin maintainer.
528 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
529 formats (previously only available in exim formats via -bpr, -bpru,
530 and -bpra. Now also available in native and exiqgrep formats)
532 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
533 with very large, slow to parse queues
535 JJ/08 exipick: added ! as generic prefix to negate any criteria format
537 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
539 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
540 responses to authentication challenges, though it was showing the
541 challenges; (ii) I've removed the CR characters from the debug output for
544 PH/58 Allow for the insertion of a newline as well as a space when a string
545 is turned into more than one encoded-word during RFC 2047 encoding. The
546 Sieve code now uses this.
548 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
549 data_4xx, lost_connection, tls_required.
551 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
552 message was being sent as an SMTP response.
554 PH/61 Add -l and -k options to exicyclog.
556 PH/62 When verifying, if an address was redirected to one new address, so that
557 verification continued, and the new address failed or deferred after
558 having set something in $address_data, the value of $address_data was not
559 passed back to the ACL. This was different to the case when no
560 redirection occurred. The value is now passed back in both cases.
562 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
563 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
564 use login_cap.h, so on its own it isn't the distinguishing feature. The
565 new name refers directly to the setclassresources() function.
567 PH/65 Added configuration files for NetBSD3.
569 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
571 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
572 is preferred over IPv4.
574 PH/68 The bounce_return_message and bounce_return_body options were not being
575 honoured for bounces generated during the reception of non-SMTP messages.
576 In particular, this applied to messages rejected by the ACL. This bug has
577 been fixed. However, if bounce_return_message is true and bounce_return_
578 body is false, the headers that are returned for a non-SMTP message
579 include only those that have been read before the error was detected.
580 (In the case of an ACL rejection, they have all been read.)
582 PH/69 The HTML version of the specification is now built in a directory called
583 spec_html instead of spec.html, because the latter looks like a path with
584 a MIME-type, and this confuses some software.
586 PH/70 Catch two compiler warnings in sieve.c.
588 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
589 function verify_get_ident() calls ip_connect() to connect a socket, but
590 if the "connect()" function timed out, ip_connect() used to close the
591 socket. However, verify_get_ident() also closes the socket later, and in
592 between Exim writes to the log, which may get opened at this point. When
593 the socket was closed in ip_connect(), the log could get the same file
594 descriptor number as the socket. This naturally causes chaos. The fix is
595 not to close the socket in ip_connect(); the socket should be closed by
596 the function that creates it. There was only one place in the code where
597 this was missing, in the iplookup router, which I don't think anybody now
598 uses, but I've fixed it anyway.
600 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
601 well as to direct DNS lookups. Otherwise the handling of names in host
602 lists is inconsistent and therefore confusing.
608 PH/01 Two changes to the default runtime configuration:
610 (1) Move the checks for relay_from_hosts and authenticated clients from
611 after to before the (commented out) DNS black list checks.
613 (2) Add control=submission to the relay_from_hosts and authenticated
614 clients checks, on the grounds that messages accepted by these
615 statements are most likely to be submissions.
617 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
619 (1) Generate an error if the third argument for the ${prvs expansion is
622 (2) Treat a missing third argument of ${prvscheck as if it were an empty
625 (3) Reset the variables that are obtained from the first argument of
626 ${prvscheck and used in the second argument before leaving the code,
627 because their memory is reclaimed, so using them afterwards may do
630 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
631 one (it's much easier than Tom thought :-).
633 (5) Because of (4), we can now allow for the use of $prvscheck_result
634 inside the third argument.
636 PH/03 For some reason, the default setting of PATH when running a command from
637 a pipe transport was just "/usr/bin". I have changed it to
640 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
641 anything to be listed in the output from -bV.
643 PH/05 When a filter generated an autoreply, the entire To: header line was
644 quoted in the delivery log line, like this:
646 => >A.N.Other <ano@some.domain> <original@ddress> ...
648 This has been changed so that it extracts the operative address. There
649 may be more than one such address. If so, they are comma-separated, like
652 => >ano@some.domain,ona@other.domain <original@ddress> ...
654 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
655 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
656 not being looked up in the rDNS to get a host name, Exim was showing the
657 IP address twice in Received: lines, even though the IP addresses were
658 identical. For example:
660 Received: from [1.2.3.4] (helo=[1.2.3.4])
662 However, if the real host name was known, it was omitting the HELO data
663 if it matched the actual IP address. This has been tidied up so that it
664 doesn't show the same IP address twice.
666 PH/07 When both +timestamp and +memory debugging was on, the value given by
667 $tod_xxx expansions could be wrong, because the tod_stamp() function was
668 called by the debug printing, thereby overwriting the timestamp buffer.
669 Debugging no longer uses the tod_stamp() function when +timestamp is set.
671 PH/08 When the original message was included in an autoreply transport, it
672 always said "this is a copy of the message, including all the headers",
673 even if body_only or headers_only was set. It now gives an appropriate
676 PH/09 Applied a patch from the Sieve maintainer which:
678 o fixes some comments
679 o adds the (disabled) notify extension core
680 o adds some debug output for the result of if/elsif tests
681 o points to the current vacation draft in the documentation
682 and documents the missing references header update
686 o fixes a bug in processing the envelope test (when testing
687 multiple envelope elements, the last element determinted the
690 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
691 Electronic Mail") by including:
693 Auto-submitted: auto-generated
695 in the messages that it generates (bounce messages and others, such as
696 warnings). In the case of bounce messages for non-SMTP mesages, there was
697 also a typo: it was using "Auto_submitted" (underscore instead of
698 hyphen). Since every message generated by Exim is necessarily in response
699 to another message, thes have all been changed to:
701 Auto-Submitted: auto-replied
703 in accordance with these statements in the RFC:
705 The auto-replied keyword:
707 - SHOULD be used on messages sent in direct response to another
708 message by an automatic process,
710 - MUST NOT be used on manually-generated messages,
712 - MAY be used on Delivery Status Notifications (DSNs) and Message
713 Disposition Notifications (MDNs),
715 - MUST NOT be used on messages generated by automatic or periodic
716 processes, except for messages which are automatic responses to
719 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
720 to the default Received: header definition.
722 PH/12 Added log selector acl_warn_skipped (default on).
724 PH/13 After a successful wildlsearch lookup, discard the values of numeric
725 variables because (a) they are in the wrong storage pool and (b) even if
726 they were copied, it wouldn't work properly because of the caching.
728 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
729 checking when decoding. Apparently there are clients that generate
730 overlong encoded strings. Why am I not surprised?
732 PH/15 If the first argument of "${if match_address" was not empty, but did not
733 contain an "@" character, Exim crashed. Now it writes a panic log message
734 and treats the condition as false.
736 PH/16 In autoreply, treat an empty string for "once" the same as unset.
738 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
739 extension "envelope-auth". The code is finished and in agreement with
740 other implementations, but there is no documentation so far and in fact,
741 nobody wrote the draft yet. This extension is currently #undef'ed, thus
742 not changing the active code.
744 Print executed "if" and "elsif" statements when debugging is used. This
745 helps a great deal to understand what a filter does.
747 Document more things not specified clearly in RFC3028. I had all this
748 sorted out, when out of a sudden new issues came to my mind. Oops."
750 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
753 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
754 canonical form (as documented). However, after a host name lookup from
755 the IP address, check_host() was doing a simple string comparison with
756 addresses acquired from the DNS when checking that the found name did
757 have the original IP as one of its addresses. Since any found IPv6
758 addresses are likely to be in abbreviated form, the comparison could
759 fail. Luckily, there already exists a function for doing the comparison
760 by converting both addresses to binary, so now that is used instead of
763 PH/20 There was another similar case to PH/19, when a complete host name was
764 given in a host list; looking up its IP address could give an abbreviated
765 form, whereas the current host's name might or might not be abbreviated.
766 The same fix has been applied.
772 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
773 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
776 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
777 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
779 PH/03 Typo: missing ".o" in src/pcre/Makefile.
781 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
782 header line, restrict the check to what is listed in RFCs 2369 and 2929.
783 Also, for "Auto-Submitted", treat anything other than "no" as
784 non-personal, in accordance with RFC 3834. (Previously it treated
785 anything starting "auto-" as non-personal.)
787 TF/01 The control=submission/name=... option had a problem with syntax
788 errors if the name included a slash character. The /name= option
789 now slurps the rest of the string, so it can include any characters
790 but it must come last in the list of options (after /sender_retain
793 PH/05 Some modifications to the interface to the fake nameserver for the new
801 TK/01 Added the "success_on_redirect" address verification option. See
802 NewStuff for rationale and an example.
804 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
806 PH/02 Patch to exigrep to allow it to work on syslog lines.
808 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
809 fread() to skip over the body file's header line, because in Cygwin the
810 header line is locked and is inaccessible.
812 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
813 co-exist for some time) to make it clear that it is the Exim ID that is
814 referenced, not the Message-ID: header line.
816 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
817 string_format() function, because snprintf() does not exist on all
820 PH/06 The use of forbid_filter_existstest now also locks out the use of the
821 ${stat: expansion item.
823 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
824 protocol synchronization error", to keep the pedants happy.
826 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
827 well as for IRIX systems, when gcc is being used. See the host.c source
830 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
832 PH/10 Named domain lists were not working if used in a queue_smtp_domains
835 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
836 transport and to the smtp transport in LMTP mode.
838 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
840 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
841 run a filter in a subprocess. This could lead to confusion in subsequent
842 lookups in the parent process. There should also be a search_tidyup() at
843 the end of the subprocess.
845 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
846 only if the host matched helo_try_verify_hosts, which caused the
847 verification to occur when the EHLO/HELO command was issued. The ACL just
848 tested the remembered result. Now, if a previous verification attempt has
849 not happened, "verify = helo" does it there and then.
851 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
853 TK/03 Fix log output including CR from clamd.
855 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
856 exist provoked a memory error which could cause a segfault.
858 PH/15 Installed PCRE 6.2
860 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
862 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
863 of the problem. Specifically, suggested +O2 rather than +O1 for the
866 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
868 PH/20 If a delivery was routed to a non-standard port by means of an SRV
869 record, the port was not correctly logged when the outgoing_port log
870 selector was set (it logged the transort's default port).
872 PH/21 Added support for host-specific ports to manualroute, queryprogram,
873 fallback_hosts, and "hosts" in the smtp transport.
875 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
876 host errors such as "Connection refused".
878 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
879 authentication with radiusclient 0.4.9:
881 - Error returned from rc_read_config was caught wrongly
882 - Username/password not passed on to radius server due to wrong length.
884 The presumption is that some radiusclient API changes for 4.51/PH/17
885 were not taken care of correctly. The code is still untested by me (my
886 Linux distribution still has 0.3.2 of radiusclient), but it was
887 contributed by a Radius user.
889 PH/24 When doing a callout, the value of $domain wasn't set correctly when
890 expanding the "port" option of the smtp transport.
892 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
893 while reading a MIME header. Thanks to Tom Hughes for a patch.
895 PH/24 Include config.h inside local_scan.h so that configuration settings are
898 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
899 that in an ACL for RCPT (for example), you can examine exactly what was
902 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
903 commands, but it was not correctly comparing the address with the actual
904 client host address. Thus, it would show the EHLO address in Received:
905 header lines when this was not necessary.
907 PH/27 Added the % operator to ${eval:}.
909 PH/28 Exim tries to create and chdir to its spool directory when it starts;
910 it should be ignoring failures (because with -C, for example, it has lost
911 privilege). It wasn't ignoring creation failures other than "already
914 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
917 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
918 because an input line was too long, either on its own, or by virtue of
919 too many continuations, the temporary file was not being removed, and the
920 return code was incorrect.
922 PH/31 Missing "BOOL" in function definition in filtertest.c.
924 PH/32 Applied Sieve patches from the maintainer.
926 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
928 PH/33 Added "verify = not_blind".
930 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
931 Local/Makefile (with some defaults set). These are used in built scripts
932 such as exicyclog, but they have never been used in the exim_install
933 script (though there are many overriding facilities there). I have
934 arranged that the exim_install script now takes note of these two
937 PH/35 Installed configuration files for Dragonfly.
939 PH/36 When a locally submitted message by a trusted user did not contain a
940 From: header, and the sender address was obtained from -f or from an SMTP
941 MAIL command, and the trusted user did not use -F to supply a sender
942 name, $originator_name was incorrectly used when constructing a From:
943 header. Furthermore, $originator_name was used for submission mode
944 messages from external hosts without From: headers in a similar way,
945 which is clearly wrong.
947 PH/37 Added control=suppress_local_fixups.
949 PH/38 When log_selector = +received_sender was set, and the addition of the
950 sender made the log line's construction buffer exactly full, or one byte
951 less than full, an overflow happened when the terminating "\n" was
954 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
955 when the result of a list match is failure because a DNS lookup failed.
957 PH/40 RM_COMMAND is now used in the building process.
959 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
960 the "build-* directories that it finds.
962 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
963 address in a domain literal was a prefix of an interface address.
965 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
966 when verifying a sender address, unless rewrite_headers is false.
968 PH/44 Wrote a long comment about why errors_to addresses are verified as
969 recipients, not senders.
971 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
972 the ratelimit ACL was added.
974 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
976 PH/46 Added extra information about PostgreSQL errors to the error string.
978 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
979 avoiding the need to install special zones in a real server. This is
980 backwards compatible; if it can't find the fake resolver, it drops back.
981 Thus, both old and new test suites can be run.
983 TF/02 Added util/ratelimit.pl
985 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
986 clock is set back in time.
988 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
989 Candler <B.Candler@pobox.com>.
991 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
992 OK for addresses that are the result of redirections.
994 PH/48 A number of further additions for the benefit of the new test suite,
995 including a fake gethostbyname() that interfaces to the fake DNS resolver
998 TF/06 The fix for widen_domains has also been applied to qualify_single and
999 search_parents which are the other dnslookup options that can cause
1002 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1005 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1007 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1008 service" if it encountered an MX record with an empty target hostname.
1009 The message is now "an MX or SRV record indicated no SMTP service".
1011 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1012 if the DNS of the sending site is misconfigured. This is quite a
1013 common situation. This change restores the behaviour of treating a
1014 helo verification defer as a failure.
1016 PH/51 If self=fail was set on a router, the bounce message did not include the
1017 actual error message.
1023 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1025 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1026 command itself ended in error, the underlying message about the transport
1027 filter timeout was being overwritten with the pipe command error. Now the
1028 underlying error message should be appended to the second error message.
1030 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1032 PH/02 Reduce the amount of output that "make" produces by default. Full output
1033 can still be requested.
1035 PH/03 The warning log line about a condition test deferring for a "warn" verb
1036 was being output only once per connection, rather than after each
1037 occurrence (because it was using the same function as for successful
1038 "warn" verbs). This seems wrong, so I have changed it.
1040 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1041 it should not have, which might have caused a crash in the right
1042 circumstances, but probably never did.
1044 PH/04 Installed a modified version of Tony Finch's patch to make submission
1045 mode fix the return path as well as the Sender: header line, and to
1046 add a /name= option so that you can make the user's friendly name appear
1049 TF/03 Added the control = fakedefer ACL modifier.
1051 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1052 Mark Lowes for thorough testing.
1054 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1056 TK/03 Merged latest SRS patch from Miles Wilton.
1058 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1059 with the definition in sysexits.h (which is #included earlier).
1060 Fortunately, Exim does not actually use EX_OK. The code used to try to
1061 preserve the sysexits.h value, by assumimg that macro definitions were
1062 scanned for macro replacements. I have been disabused of this notion,
1063 so now the code just undefines EX_OK before #including unistd.h.
1065 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1066 in the smtp transport. When a block could not be written in a single
1067 write() function, the timeout was being re-applied to each part-write.
1068 This seems wrong - if the receiver was accepting one byte at a time it
1069 would take for ever. The timeout is now adjusted when this happens. It
1070 doesn't have to be particularly precise.
1072 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1073 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1075 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1076 without a domain if the check to <postmaster@domain> fails.
1078 SC/01 Eximstats: added -xls and the ability to specify output files
1079 (patch written by Frank Heydlauf).
1081 SC/02 Eximstats: use FileHandles for outputing results.
1083 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1085 SC/04 Eximstats: fixed display of large numbers with -nvr option
1087 SC/05 Eximstats: fixed merging of reports with empty tables.
1089 SC/06 Eximstats: added the -include_original_destination flag
1091 SC/07 Eximstats: removed tabs and trailing whitespace.
1093 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1095 TK/06 MBOX spool code: Add real "From " MBOX separator line
1096 so the .eml file is really in mbox format (even though
1097 most programs do not really care). Patch from Alex Miller.
1099 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1100 The latter is generated from $received_to and is only set if the
1101 message has one envelope recipient. SA can use these headers,
1102 obviously out-of-the-box. Patch from Alex Miller.
1104 PH/08 The ${def test on a variable was returning false if the variable's
1105 value was "0", contrary to what the specification has always said!
1106 The result should be true unless the variable is empty.
1108 PH/09 The syntax error of a character other than { following "${if
1109 def:variable_name" (after optional whitespace) was not being diagnosed.
1110 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1111 accidental colon was present, for example, could give incorrect results.
1113 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1114 result is used (not including appendfile, where other changes are about
1117 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1118 This involved changing a lot of size variables from int to off_t. It
1119 should work with maildirs and everything.
1121 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1122 spamd dying while we are connected to it.
1124 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1127 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1128 function for BDB 4.3.
1130 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1132 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1133 the sources for PCRE have been re-arranged and more files are now
1136 PH/15 The code I had for printing potentially long long variables in PH/11
1137 above was not the best (it lost precision). The length of off_t variables
1138 is now inspected at build time, and an appropriate printing format (%ld
1139 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1140 to be "long long int" or "long int". This is needed for the internal
1141 formatting function string_vformat().
1143 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1144 the configuration file to be ":syslog", then the script "guesses" where
1145 the logs files are, rather than using the compiled in default. In our
1146 case the guess is not the same as the compiled default, so the script
1147 suddenly stopped working when I started to use syslog. The patch checks
1148 to see if log_file_path is "". If so, it attempts to read it from exim
1149 with no configuration file to get the compiled in version, before it
1150 falls back to the previous guessing code."
1152 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1153 implementing BATV in an Exim configuration. See NewStuff for the gory
1156 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1157 Makefile that are specific to HP-UX.
1159 PH/18 If the "use_postmaster" option was set for a recipient callout together
1160 with the "random" option, the postmaster address was used as the MAIL
1161 FROM address for the random test, but not for the subsequent recipient
1162 test. It is now used for both.
1164 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1165 patch removes a few documentation additions to RFC 3028, because the
1166 latest draft now contains them. It adds the new en;ascii-case comparator
1167 and a new error check for 8bit text in MIME parts. Comparator and
1168 require names are now matched exactly. I enabled the subaddress
1169 extension, but it is not well tested yet (read: it works for me)."
1171 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1172 rework some of the code of TK/09 above to avoid the hardwired use of
1173 "%lld" and "long long". Replaced the call to snprintf() with a call to
1176 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1177 records point to non-existent hosts", "retry timeout exceeded", and
1178 "retry time not reached for any host after a long failure period".
1180 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1181 experimental DomainKeys support:
1183 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1184 (2) On an error such as an illegally used "control", the wrong name for
1185 the control was given.
1187 These problems did NOT occur unless DomainKeys support was compiled.
1189 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1191 PH/24 Added ${if match_ip condition.
1193 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1194 databases so that it will be absolutely obvious if a crash occurs in the
1195 DB library. This is a regular occurrence (often caused by mis-matched
1198 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1199 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1200 picked up on a user's system that detects such things. There doesn't seem
1201 to be a gcc warning option for this - only an attribute that has to be
1202 put on the function's prototype. It seems that in Fedora Core 4 they have
1203 set this on a number of new functions. No doubt there will be more in due
1206 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1207 specify a transport. However, if an address that was verified by such a
1208 router was the subject of a callout, Exim crashed because it tried to
1209 read the rcpt_include_affixes from the non-existent transport. Now it
1210 just assumes that the setting of that option is false. This bug was
1211 introduced by 4.51/PH/31.
1213 PH/28 Changed -d+all to exclude +memory, because that information is very
1214 rarely of interest, but it makes the output a lot bigger. People tend to
1215 do -d+all out of habit.
1217 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1218 code in os-type was giving problems when libc.so lives in lib64, like on
1221 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1222 aren't the modern standard, and it seems that some systems' include files
1223 don't always have them. Exim was already checking for some of the newer
1224 ones like T_AAAA, and defining it itself. I've added checks for all the
1225 record types that Exim uses.
1227 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1228 not automatically generating a new one, as it is supposed to. This
1229 prevented TLS from working. If the file did exist, but contained invalid
1230 data, a new version was generated, as expected. It was only the case of a
1231 non-existent file that was broken.
1233 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1234 with a change in libdomainkeys > 0.64.
1236 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1237 from DNS. If the selector record carries the flag, it now has
1238 precedence over the domain-wide flag.
1240 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1242 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1243 the use of a port name that isn't defined in /etc/services) occurred, the
1244 message was deferred as in a normal delivery, and thus remained on the
1245 spool, instead of being failed because of the mua_wrapper setting. This
1246 is now fixed, and I tidied up some of the mua_wrapper messages at the
1249 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1250 the messages in a hash of arrays rather than using individual hashes.
1251 This is a bit cleaner and results in dramatic memory savings, albeit
1252 at a slight CPU cost.
1254 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1255 as requested by Marc Sherman.
1257 SC/10 Eximstats: added histograms for user specified patterns as requested
1260 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1262 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1263 fopen() in the content-scanning modules that did not already have it.
1269 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1270 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1272 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1274 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1276 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1277 to test Sieve filters that use "vacation".
1279 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1280 that changes the way the GnuTLS parameters are stored in the cache file.
1281 The new format can be generated externally. For backward compatibility,
1282 if the data in the cache doesn't make sense, Exim assumes it has read an
1283 old-format file, and it generates new data and writes a new file. This
1284 means that you can't go back to an older release without removing the
1287 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1288 work if there are any delivery delays because "one_time" forces the
1289 parent to be marked "delivered", so its unseen clone is never tried
1290 again. For this reason, Exim now forbids the simultaneous setting of
1293 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1294 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1295 entry if you want to know the details. The fix, however, neglected to
1296 consider the case when local delivery batching is involved. The test for
1297 "previously delivered" was not happening when checking to see if an
1298 address could be batched with a previous (undelivered) one; under
1299 certain circumstances this could lead to multiple deliveries to the same
1302 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1303 in its include files, and this causes problems building Exim.
1305 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1306 header_syntax) but Exim was just ignoring anything given after a slash.
1307 In particular, this caused confusion with an attempt to use "verify =
1308 reverse_host_lookup/defer_ok". An error is now given when options are
1309 supplied for verify items that do not have them. (Maybe reverse_host_
1310 lookup should have a defer_ok option, but that's a different point.)
1312 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1313 defined by RFC 821) to 2048, because there were problems with some AUTH
1314 commands, and RFC 1869 says the size should be increased for extended
1315 SMTP commands that take arguments.
1317 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1320 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1321 "unknown" error; now it says that the functionality isn't in the binary.
1323 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1324 an address' error message when a string expansion fails (syntax or
1325 whatever). Otherwise the password may appear in the log. Following change
1326 PH/42 below, there is no longer a chance of it appearing in a bounce
1329 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1331 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1332 of its addresses was ever tried. (Bugzilla bug #2.)
1334 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1335 the result incorrectly in the debug output. (It correctly added a newline
1336 to what was transported.)
1338 TF/01 Added $received_time.
1340 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1341 commented out examples of how to interface to a virus scanner and to
1342 SpamAssassin. Also added commented examples of av_scanner and
1343 spamd_address settings.
1345 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1346 and controls are allowed in which ACLs. There were a couple of minor
1347 errors. Some of the entries in the conditions table (which is a table of
1348 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1349 negation of where the condition IS allowed.
1351 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1353 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1354 header file does not have a version number, so I've had to invent a new
1355 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1356 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1357 radiusclient), but it was contributed by a Radius user.
1359 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1360 files or directories, for OpenSSL.
1362 PH/19 When an Exim process that is running as root has to create an Exim log
1363 file, it does so in a subprocess that runs as exim:exim so as to get the
1364 ownership right at creation (otherwise, other Exim processes might see
1365 the file with the wrong ownership). There was no test for failure of this
1366 fork() call, which would lead to the process getting stuck as it waited
1367 for a non-existent subprocess. Forks do occasionally fail when resources
1368 run out. I reviewed all the other calls to fork(); they all seem to check
1371 PH/20 When checking for unexpected SMTP input at connect time (before writing
1372 the banner), Exim was not dealing correctly with a non-positive return
1373 from the read() function. If the client had disconnected by this time,
1374 the result was a log entry for a synchronization error with an empty
1375 string after "input=" when read() returned zero. If read() returned -1
1376 (an event I could not check), uninitialized data bytes were printed.
1377 There were reports of junk text (parts of files, etc) appearing after
1380 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1382 PH/22 Added support for macro redefinition, and (re)definition in between
1383 driver and ACL definitions.
1385 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1386 forgetting to use the resulting value; it was using the unexpanded value.
1388 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1389 hadn't been configured. The fix is from Juergen Kreileder, who
1390 understands it better than I do:
1392 "Here's what I see happening with three configured cyrus_sasl
1393 authenticators configured (plain, login, cram-md5):
1395 On startup auth_cyrus_sasl_init() gets called for each of these.
1396 This means three calls to sasl_listmech() without a specified mech_list.
1397 => SASL tests which mechs of all available mechs actually work
1398 => three warnings about OTP not working
1399 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1401 With the patch, sasl_listmech() also gets called three times. But now
1402 SASL's mech_list option is set to the server_mech specified in the the
1403 authenticator. Or in other words, the answer from sasl_listmech()
1404 gets limited to just the mech you're testing for (which is different
1406 => the return list contains just 'plain' or 'login', 'cram-md5' or
1407 nothing depending on the value of ob->server_mech.
1409 I've just tested the patch: Authentication still works fine,
1410 unavailable mechs specified in the exim configuration are still
1411 caught, and the auth.log warnings about OTP are gone."
1413 PH/25 When debugging is enabled, the contents of the command line are added
1414 to the debugging output, even when log_selector=+arguments is not
1417 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1418 answer is "GNU", and only if the return is "GNU/something" is the answer
1421 PH/27 $acl_verify_message is now set immediately after the failure of a
1422 verification in an ACL, and so is available in subsequent modifiers. In
1423 particular, the message can be preserved by coding like this:
1425 warn !verify = sender
1426 set acl_m0 = $acl_verify_message
1428 Previously, $acl_verify_message was set only while expanding "message"
1429 and "log_message" when a very denied access.
1431 PH/28 Modified OS/os.c-Linux with
1433 -#ifndef OS_LOAD_AVERAGE
1434 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1436 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1437 nomenclature these days.)
1439 PH/29 Installed patch from the Sieve maintainer that adds the options
1440 sieve_useraddress and sieve_subaddress to the redirect router.
1442 PH/30 In these circumstances:
1443 . Two addresses routed to the same list of hosts;
1444 . First host does not offer TLS;
1445 . First host accepts first address;
1446 . First host gives temporary error to second address;
1447 . Second host offers TLS and a TLS session is established;
1448 . Second host accepts second address.
1449 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1450 and peerdn, if requested) that were in fact used only for the second
1453 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1454 attention to any local part prefix or suffix that was matched by the
1455 router that accepted the address. It now behaves in the same way as it
1456 does for delivery: the affixes are removed from the local part unless
1457 rcpt_include_affixes is set on the transport.
1459 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1460 timeout during the DATA phase of an incoming message.
1462 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1463 applied a patch sanctioned by the Sieve maintainer.
1465 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1466 the uid or gid is negative. A case of a negative gid caused this to be
1467 noticed. The fix allows for either to be negative.
1469 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1470 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1473 PH/36 The change PH/12 above was broken. Fixed it.
1475 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1476 the grounds that routing the same address twice would always produce the
1477 same answer. This might have been true once, but it is certainly no
1478 longer true now. Routing a child address may depend on the previous
1479 routing that produced that child. Some complicated redirection strategies
1480 went wrong when messages had multiple recipients, and made Exim's
1481 behaviour dependent on the order in which the addresses were given.
1483 I have moved the duplicate checking until after the routing is complete.
1484 Exim scans the addresses that are assigned to local and remote
1485 transports, and removes any duplicates. This means that more work will be
1486 done, as duplicates will always all be routed, but duplicates are
1487 presumably rare, so I don't expect this is of any significance.
1489 For deliveries to pipes, files, and autoreplies, the duplicate checking
1490 still happens during the routing process, since they are not going to be
1493 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1494 It corrects a timeout issue with spamd. This is Ian's comment: "The
1495 background is that sometimes spamd either never reads data from a
1496 connection it has accepted, or it never writes response data. The exiscan
1497 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1498 blindly assumes that writes won't block so it may never time out."
1500 PH/39 Allow G after quota size as well as K and M.
1502 PH/40 The value set for $authenticated_id in an authenticator may not contain
1503 binary zeroes or newlines because the value is written to log lines and
1504 to spool files. There was no check on this. Now the value is run through
1505 the string_printing() function so that such characters are converted to
1506 printable escape sequences.
1508 PH/41 $message_linecount is a new variable that contains the total number of
1509 lines in the message. Compare $body_linecount, which is the count for the
1512 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1513 bounce and delay warning messages, except in certain special cases, which
1516 (a) An SMTP error message from a remote host;
1517 (b) A message specified in a :fail: redirection;
1518 (c) A message specified in a "fail" command in a system filter;
1519 (d) A message specified in a FAIL return from the queryprogram router;
1520 (e) A message specified by the cannot_route_message router option.
1522 In these cases only, Exim does include the error details in bounce and
1523 warning messages. There are also a few cases where bland messages such
1524 as "unrouteable address" or "local delivery error" are given.
1526 PH/43 $value is now also set for the "else" part of a ${run expansion.
1528 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1529 being worked on, but at least Exim now implements the latest version to
1532 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1533 process to complete was treated as a delivery failure, a timeout while
1534 writing the message to the pipe was logged, but erroneously treated as a
1535 successful delivery. Such timeouts include transport filter timeouts. For
1536 consistency with the overall process timeout, these timeouts are now
1537 treated as errors, giving rise to delivery failures by default. However,
1538 there is now a new Boolean option for the pipe transport called
1539 timeout_defer, which, if set TRUE, converts the failures into defers for
1540 both kinds of timeout. A transport filter timeout is now identified in
1543 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1544 systems where "make" and "gmake" are different, calling "gmake" at top
1545 level broke things. I've arranged for the value of $(MAKE) to be passed
1546 from the Makefile to this script so that it can call the same version of
1550 A note about Exim versions 4.44 and 4.50
1551 ----------------------------------------
1553 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1554 changes of various kinds. As a consequence, a big documentation update was
1555 needed. This delayed the release for rather longer than seemed good, especially
1556 in the light of a couple of (minor) security issues. Therefore, the changes
1557 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1558 release. So 4.44 and 4.50 are in effect two different branches that both start
1561 I have left the 4.50 change log unchanged; it contains all the changes since
1562 4.43. The change log for 4.44 is below; many of its items are identical to
1563 those for 4.50. This seems to be the most sensible way to preserve the
1564 historical information.
1570 1. Minor wording change to the doc/README.SIEVE file.
1572 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1573 computation of the current number of files was incorrect.
1575 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1576 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1579 4. Give more explanation in the error message when the command for a transport
1580 filter fails to execute.
1582 5. There are several places where Exim runs a non-Exim command in a
1583 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1584 was being done only for the command run by the queryprogram router. It is
1585 now done for all such subprocesses. The other cases are: ${run, transport
1586 filters, and the commands run by the lmtp and pipe transports.
1588 6. Added CONFIGURE_GROUP build-time option.
1590 7. Some older OS have a limit of 256 on the maximum number of file
1591 descriptors. Exim was using setrlimit() to set 1000 as a large value
1592 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1593 systems. I've change it so that if it can't get 1000, it tries for 256.
1595 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1596 was an oversight, and furthermore, ever since the addition of extra
1597 controls (e.g. 4.43/32), the checks on when to allow different forms of
1598 "control" were broken. There should now be diagnostics for all cases when a
1599 control that does not make sense is encountered.
1601 9. Added the /retain_sender option to "control=submission".
1603 10. $recipients is now available in the predata ACL (oversight).
1605 11. Tidy the search cache before the fork to do a delivery from a message
1606 received from the command line. Otherwise the child will trigger a lookup
1607 failure and thereby defer the delivery if it tries to use (for example) a
1608 cached ldap connection that the parent has called unbind on.
1610 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1611 of $address_data from the recipient verification was clobbered by the
1612 sender verification.
1614 13. The value of address_data from a sender verification is now available in
1615 $sender_address_data in subsequent conditions in the ACL statement.
1617 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1619 15. Added a new option "connect=<time>" to callout options, to set a different
1622 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1623 was its contents. (It was OK if the option was not defined at all.)
1625 17. A "Completed" log line is now written for messages that are removed from
1626 the spool by the -Mrm option.
1628 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1629 information about exactly what failed.
1631 19. Added -dd to debug only the daemon process.
1633 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1634 handles timeouts, both on the server side and network timeouts. Renamed the
1635 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1637 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1638 It is now set to "smtps".
1640 22. $host_address is now set to the target address during the checking of
1641 ignore_target_hosts.
1643 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1644 being passed; this would have caused $sender_host_name to have been used if
1645 matching the list had actually called for a host name (not very likely,
1646 since this list is usually IP addresses). A host name is now passed as
1649 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1650 code that specifies a non-restarting handler (typically sigaction() in
1651 modern systems) in an attempt to fix a rare and obscure crash bug.
1653 25. Narrowed the window for a race in the daemon that could cause it to ignore
1654 SIGCHLD signals. This is not a major problem, because they are used only to
1655 wake it up if nothing else does.
1657 26. A malformed maildirsize file could cause Exim to calculate negative values
1658 for the mailbox size or file count. Odd effects could occur as a result.
1659 The maildirsize information is now recalculated if the size or filecount
1662 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1663 support for a long time. Removed HAVE_SYS_VFS_H.
1665 28. Installed the latest version of exipick from John Jetmore.
1667 29. In an address list, if the pattern was not a regular expression, an empty
1668 subject address (from a bounce message) matched only if the pattern was an
1669 empty string. Non-empty patterns were not even tested. This was the wrong
1670 because it is perfectly reasonable to use an empty address as part of a
1671 database query. An empty address is now tested by patterns that are
1672 lookups. However, all the other forms of pattern expect the subject to
1673 contain a local part and a domain, and therefore, for them, an empty
1674 address still always fails if the pattern is not itself empty.
1676 30. Exim went into a mad DNS loop when attempting to do a callout where the
1677 host was specified on an smtp transport, and looking it up yielded more
1678 than one IP address.
1680 31. Re-factored the code for checking spool and log partition space into a
1681 function that finds that data and another that does the check. The former
1682 is then used to implement four new variables: $spool_space, $log_space,
1683 $spool_inodes, and $log_inodes.
1685 32. The RFC2047 encoding function was originally intended for short strings
1686 such as real names; it was not keeping to the 75-character limit for
1687 encoded words that the RFC imposes. It now respects the limit, and
1688 generates multiple encoded words if necessary. To be on the safe side, I
1689 have increased the buffer size for the ${rfc2047: expansion operator from
1692 33. It is now permitted to omit both strings after an "if" condition; if the
1693 condition is true, the result is "true". As before, when the second string
1694 is omitted, a false condition yields an empty string. This makes it less
1695 cumbersome to write custom ACL and router conditions.
1697 34. Failure to deliver a bounce message always caused it to be frozen, even if
1698 there was an errors_to setting on the router. The errors_to setting is now
1701 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1702 canonical form (fully expanded) before being placed in
1703 $sender_host_address.
1705 36. The table in the code that translates DNS record types into text (T_A to
1706 "A" for instance) was missing entries for NS and CNAME. It is just possible
1707 that this could have caused confusion if both these types were looked up
1708 for the same domain, because the text type is used as part of Exim's
1709 per-process caching. But the chance of anyone hitting this buglet seems
1712 37. The dnsdb lookup has been extended in a number of ways.
1714 (1) There is a new type, "zns", which walks up the domain tree until it
1715 finds some nameserver records. It should be used with care.
1717 (2) There is a new type, "mxh", which is like "mx" except that it returns
1718 just the host names, not the priorities.
1720 (3) It is now possible to give a list of domains (or IP addresses) to be
1721 looked up. The behaviour when one of the lookups defers can be
1722 controlled by a keyword.
1724 (4) It is now possible to specify the separator character for use when
1725 multiple records are returned.
1727 38. The dnslists ACL condition has been extended: it is now possible to supply
1728 a list of IP addresses and/or domains to be looked up in a particular DNS
1731 39. Added log_selector=+queue_time_overall.
1733 40. When running the queue in the test harness, wait just a tad after forking a
1734 delivery process, to get repeatability of debugging output.
1736 41. Include certificate and key file names in error message when GnuTLS fails
1737 to set them up, because the GnuTLS error message doesn't include the name
1738 of the failing file when there is a problem reading it.
1740 42. Allow both -bf and -bF in the same test run.
1742 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1744 44. The "Exiscan patch" is now merged into the mainline Exim source.
1746 45. Sometimes the final signoff response after QUIT could fail to get
1747 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1748 < 0 before doing a fflush(). This bug looks as though it goes back to the
1749 introduction of TLS in release 3.20, but "sometimes" must have been rare
1750 because the tests only now provoked it.
1752 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1753 (this can affect the format of dates).
1755 47. exim_tidydb, when checking for the continued existence of a message for
1756 which it has found a message-specific retry record, was not finding
1757 messages that were in split spool directories. Consequently, it was
1758 deleting retry records that should have stayed in existence.
1760 48. Steve fixed some bugs in eximstats.
1762 49. The SPA authentication driver was not abandoning authentication and moving
1763 on to the next authenticator when an expansion was forced to fail,
1764 contradicting the general specification for all authenticators. Instead it
1765 was generating a temporary error. It now behaves as specified.
1767 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1768 (the order specifies the preference for clients). The order is now AES256,
1769 AES128, 3DES, ARCFOUR128.
1771 51. Small patch to Sieve code - explicitly set From: when generating an
1774 52. Exim crashed if a remote delivery caused a very long error message to be
1775 recorded - for instance if somebody sent an entire SpamAssassin report back
1776 as a large number of 550 error lines. This bug was coincidentally fixed by
1777 increasing the size of one of Exim's internal buffers (big_buffer) that
1778 happened as part of the Exiscan merge. However, to be on the safe side, I
1779 have made the code more robust (and fixed the comments that describe what
1782 53. Now that there can be additional text after "Completed" in log lines (if
1783 the queue_time_overall log selector is set), a one-byte patch to exigrep
1784 was needed to allow it to recognize "Completed" as not the last thing in
1787 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1788 patch that reportedly fixes this has been added. I am not expert enough to
1789 create a test for it. This is what the patch creator wrote:
1791 "I found a little strange behaviour of ldap code when working with
1792 Windows 2003 AD Domain, where users was placed in more than one
1793 Organization Units. When I tried to give exim partial DN, the exit code
1794 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1795 But simultaneously result of request was absolutely normal ldap result,
1796 so I produce this patch..."
1798 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1799 REFERENCE, so I have modified the code to exclude the patch when that macro
1802 55. Some experimental protocols are using DNS PTR records for new purposes. The
1803 keys for these records are domain names, not reversed IP addresses. The
1804 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1805 leaves it alone. Component reversal etc. now happens only for IP addresses.
1808 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1810 57. Double the size of the debug message buffer (to 2048) so that more of very
1811 long debug lines gets shown.
1813 58. The exicyclog utility now does better if the number of log files to keep
1814 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1816 59. Two changes related to the smtp_active_hostname option:
1818 (1) $smtp_active_hostname is now available as a variable.
1819 (2) The default for smtp_banner uses $smtp_active_hostname instead
1820 of $primary_hostname.
1822 60. The host_aton() function is supposed to be passed a string that is known
1823 to be a valid IP address. However, in the case of IPv6 addresses, it was
1824 not checking this. This is a hostage to fortune. Exim now panics and dies
1825 if the condition is not met. A case was found where this could be provoked
1826 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1827 components; fortuitously, this particular loophole had already been fixed
1828 by change 4.50/55 above.
1830 If there are any other similar loopholes, the new check in host_aton()
1831 itself should stop them being exploited. The report I received stated that
1832 data on the command line could provoke the exploit when Exim was running as
1833 exim, but did not say which command line option was involved. All I could
1834 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1835 running as the user.
1838 61. There was a buffer overflow vulnerability in the SPA authentication code
1839 (which came originally from the Samba project). I have added a test to the
1840 spa_base64_to_bits() function which I hope fixes it.
1843 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1844 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1846 63. The daemon start-up calls getloadavg() while still root for those OS that
1847 need the first call to be done as root, but it missed one case: when
1848 deliver_queue_load_max is set with deliver_drop_privilege. This is
1849 necessary for the benefit of the queue runner, because there is no re-exec
1850 when deliver_drop_privilege is set.
1852 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1853 This has been fixed.
1855 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1856 was in use, was not putting the data itself into the right store pool;
1857 consequently, it could be overwritten for a subsequent message in the same
1858 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1861 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1863 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1864 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1865 the function were treating the return as a boolean value, which happened to
1866 work because 0=false and not-0=true, but is not correct code.
1868 68. The host_aton() function was not handling scoped IPv6 addresses (those
1869 with, for example, "%eth0" on the end) correctly.
1871 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1872 negated items (that is, ~something) in unsigned ints. Some compilers
1873 apparently mutter when there is no cast.
1875 70. If an address verification called from an ACL failed, and did not produce a
1876 user-specific message (i.e. there was only a "system" message), nothing was
1877 put in $acl_verify_message. In this situation, it now puts the system
1880 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1881 session; change 4.31/43 added the unwanted input to the log line - except
1882 that it did not do this in the start of session case. It now does.
1884 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1885 This is wrong and can cause the other end to generate a synchronization
1886 error if it is another Exim or anything else that does the synchronization
1887 check. A QUIT command is no longer sent after a timeout.
1889 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1890 during host lookups.
1892 74. The defer_ok option of callout verification was not working if it was used
1893 when verifying addresses in header lines, that is, for this case:
1895 verify = header_sender/callout=defer_ok
1897 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1898 those file descriptors could be used for SMTP connections. If anything
1899 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1900 could be sent to the SMTP client, causing chaos. The daemon now opens
1901 stdin, stdout, and stderr to /dev/null when it puts itself into the
1904 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1905 log by default. The user can override this with suitable Perl magic.
1907 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1908 the log message when discard triggers, was not working for the DATA ACL or
1909 for the non-SMTP ACL.
1911 78. Error message wording change in sieve.c.
1913 79. If smtp_accept_max_per_host was set, the number of connections could be
1914 restricted to fewer than expected, because the daemon was trying to set up
1915 a new connection before checking whether the processes handling previous
1916 connections had finished. The check for completed processes is now done
1917 earlier. On busy systems, this bug wouldn't be noticed because something
1918 else would have woken the daemon, and it would have reaped the completed
1921 80. If a message was submitted locally by a user whose login name contained one
1922 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1923 It caused a spool format error. I have fixed the spool reading code. A
1924 related problem was that the "from" clause in the Received: line became
1925 illegal because of the space(s). It is now covered by ${quote_local_part.
1927 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1930 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1932 83. Patch from Sieve maintainer to fix unterminated string problem in
1933 "vacation" handling.
1935 84. Some minor changes to the Linux configuration files to help with other
1936 OS variants using glibc.
1938 85. One more patch for Sieve to update vacation handling to latest spec.
1941 ----------------------------------------------------
1942 See the note above about the 4.44 and 4.50 releases.
1943 ----------------------------------------------------
1949 1. Change 4.43/35 introduced a bug that caused file counts to be
1950 incorrectly computed when quota_filecount was set in an appendfile
1953 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1954 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1957 3. Give more explanation in the error message when the command for a transport
1958 filter fails to execute.
1960 4. There are several places where Exim runs a non-Exim command in a
1961 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1962 was being done only for the command run by the queryprogram router. It is
1963 now done for all such subprocesses. The other cases are: ${run, transport
1964 filters, and the commands run by the lmtp and pipe transports.
1966 5. Some older OS have a limit of 256 on the maximum number of file
1967 descriptors. Exim was using setrlimit() to set 1000 as a large value
1968 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1969 systems. I've change it so that if it can't get 1000, it tries for 256.
1971 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1972 was an oversight, and furthermore, ever since the addition of extra
1973 controls (e.g. 4.43/32), the checks on when to allow different forms of
1974 "control" were broken. There should now be diagnostics for all cases when a
1975 control that does not make sense is encountered.
1977 7. $recipients is now available in the predata ACL (oversight).
1979 8. Tidy the search cache before the fork to do a delivery from a message
1980 received from the command line. Otherwise the child will trigger a lookup
1981 failure and thereby defer the delivery if it tries to use (for example) a
1982 cached ldap connection that the parent has called unbind on.
1984 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1985 of $address_data from the recipient verification was clobbered by the
1986 sender verification.
1988 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1989 was its contents. (It was OK if the option was not defined at all.)
1991 11. A "Completed" log line is now written for messages that are removed from
1992 the spool by the -Mrm option.
1994 12. $host_address is now set to the target address during the checking of
1995 ignore_target_hosts.
1997 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1998 being passed; this would have caused $sender_host_name to have been used if
1999 matching the list had actually called for a host name (not very likely,
2000 since this list is usually IP addresses). A host name is now passed as
2003 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2004 code that specifies a non-restarting handler (typically sigaction() in
2005 modern systems) in an attempt to fix a rare and obscure crash bug.
2007 15. Narrowed the window for a race in the daemon that could cause it to ignore
2008 SIGCHLD signals. This is not a major problem, because they are used only to
2009 wake it up if nothing else does.
2011 16. A malformed maildirsize file could cause Exim to calculate negative values
2012 for the mailbox size or file count. Odd effects could occur as a result.
2013 The maildirsize information is now recalculated if the size or filecount
2016 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2017 support for a long time. Removed HAVE_SYS_VFS_H.
2019 18. Updated exipick to current release from John Jetmore.
2021 19. Allow an empty sender to be matched against a lookup in an address list.
2022 Previously the only cases considered were a regular expression, or an
2025 20. Exim went into a mad DNS lookup loop when doing a callout where the
2026 host was specified on the transport, if the DNS lookup yielded more than
2029 21. The RFC2047 encoding function was originally intended for short strings
2030 such as real names; it was not keeping to the 75-character limit for
2031 encoded words that the RFC imposes. It now respects the limit, and
2032 generates multiple encoded words if necessary. To be on the safe side, I
2033 have increased the buffer size for the ${rfc2047: expansion operator from
2036 22. Failure to deliver a bounce message always caused it to be frozen, even if
2037 there was an errors_to setting on the router. The errors_to setting is now
2040 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2041 canonical form (fully expanded) before being placed in
2042 $sender_host_address.
2044 24. Updated eximstats to version 1.33
2046 25. Include certificate and key file names in error message when GnuTLS fails
2047 to set them up, because the GnuTLS error message doesn't include the name
2048 of the failing file when there is a problem reading it.
2050 26. Expand error message when OpenSSL has problems setting up cert/key files.
2053 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2054 (this can affect the format of dates).
2056 28. exim_tidydb, when checking for the continued existence of a message for
2057 which it has found a message-specific retry record, was not finding
2058 messages that were in split spool directories. Consequently, it was
2059 deleting retry records that should have stayed in existence.
2061 29. eximstats updated to version 1.35
2062 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2063 1.35 - bugfix such that pie charts by volume are generated correctly
2065 30. The SPA authentication driver was not abandoning authentication and moving
2066 on to the next authenticator when an expansion was forced to fail,
2067 contradicting the general specification for all authenticators. Instead it
2068 was generating a temporary error. It now behaves as specified.
2070 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2071 (the order specifies the preference for clients). The order is now AES256,
2072 AES128, 3DES, ARCFOUR128.
2074 31. Small patch to Sieve code - explicitly set From: when generating an
2077 32. Exim crashed if a remote delivery caused a very long error message to be
2078 recorded - for instance if somebody sent an entire SpamAssassin report back
2079 as a large number of 550 error lines. This bug was coincidentally fixed by
2080 increasing the size of one of Exim's internal buffers (big_buffer) that
2081 happened as part of the Exiscan merge. However, to be on the safe side, I
2082 have made the code more robust (and fixed the comments that describe what
2085 33. Some experimental protocols are using DNS PTR records for new purposes. The
2086 keys for these records are domain names, not reversed IP addresses. The
2087 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2088 leaves it alone. Component reversal etc. now happens only for IP addresses.
2091 34. The host_aton() function is supposed to be passed a string that is known
2092 to be a valid IP address. However, in the case of IPv6 addresses, it was
2093 not checking this. This is a hostage to fortune. Exim now panics and dies
2094 if the condition is not met. A case was found where this could be provoked
2095 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2096 components; fortuitously, this particular loophole had already been fixed
2097 by change 4.50/55 or 4.44/33 above.
2099 If there are any other similar loopholes, the new check in host_aton()
2100 itself should stop them being exploited. The report I received stated that
2101 data on the command line could provoke the exploit when Exim was running as
2102 exim, but did not say which command line option was involved. All I could
2103 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2104 running as the user.
2107 35. There was a buffer overflow vulnerability in the SPA authentication code
2108 (which came originally from the Samba project). I have added a test to the
2109 spa_base64_to_bits() function which I hope fixes it.
2112 36. The daemon start-up calls getloadavg() while still root for those OS that
2113 need the first call to be done as root, but it missed one case: when
2114 deliver_queue_load_max is set with deliver_drop_privilege. This is
2115 necessary for the benefit of the queue runner, because there is no re-exec
2116 when deliver_drop_privilege is set.
2118 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2119 was in use, was not putting the data itself into the right store pool;
2120 consequently, it could be overwritten for a subsequent message in the same
2121 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2124 38. Sometimes the final signoff response after QUIT could fail to get
2125 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2126 < 0 before doing a fflush(). This bug looks as though it goes back to the
2127 introduction of TLS in release 3.20, but "sometimes" must have been rare
2128 because the tests only now provoked it.
2134 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2135 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2136 Now it returns an integer. A number of calls were still expecting a T/F
2137 return. Fortuitously, in all cases, the tests worked in OK situations,
2138 which is the norm. However, things would have gone wrong on any write
2139 failures on the smtp file descriptor. This function is used when sending
2140 messages over SMTP and also when doing verify callouts.
2142 2. When Exim is called to do synchronous delivery of a locally submitted
2143 message (the -odf or -odi options), it no longer closes stderr before doing
2146 3. Implemented the mua_wrapper option.
2148 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2150 5. Implemented the functions header_remove(), header_testname(),
2151 header_add_at_position(), and receive_remove_recipient(), and exported them
2154 6. If an ACL "warn" statement specified the addition of headers, Exim already
2155 inserted X-ACL-Warn: at the start if there was no header name. However, it
2156 was not making this test for the second and subsequent header lines if
2157 there were newlines in the string. This meant that an invalid header could
2158 be inserted if Exim was badly configured.
2160 7. Allow an ACL "warn" statement to add header lines at the start or after all
2161 the Received: headers, as well as at the end.
2163 8. Added the rcpt_4xx retry error code.
2165 9. Added postmaster_mailfrom=xxx to callout verification option.
2167 10. Added mailfrom=xxxx to the callout verification option, for verify=
2170 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2171 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2173 12. Inserted some casts to stop certain compilers warning when using pointer
2174 differences as field lengths or precisions in printf-type calls (mostly
2175 affecting debugging statements).
2177 13. Added optional readline() support for -be (dynamically loaded).
2179 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2180 same clock tick as a message's arrival, so that its received time was the
2181 same as the "first fail" time on the retry record, and that message
2182 remained on the queue past the ultimate address timeout, every queue runner
2183 would try a delivery (because it was past the ultimate address timeout) but
2184 after another failure, the ultimate address timeout, which should have then
2185 bounced the address, did not kick in. This was a "< instead of <=" error;
2186 in most cases the first failure would have been in the next clock tick
2187 after the received time, and all would be well.
2189 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2190 being recognized when the domain list was tested by the match_domain
2191 condition in an expansion string.
2193 16. Added the ${str2b64: operator.
2195 17. Exim was always calling setrlimit() to set a large limit for the number of
2196 processes, without checking whether the existing limit was already
2197 adequate. (It did check for the limit on file descriptors.) Furthermore,
2198 errors from getrlimit() and setrlimit() were being ignored. Now they are
2199 logged to the main and panic logs, but Exim does carry on, to try to do its
2200 job under whatever limits there are.
2202 18. Imported PCRE 5.0.
2204 19. Trivial typo in log message " temporarily refused connection" (the leading
2207 20. If the log selector return_path_on_delivery was set and an address was
2208 redirected to /dev/null, the delivery process crashed because it assumed
2209 that a return path would always be set for a "successful" delivery. In this
2210 case, the whole delivery is bypassed as an optimization, and therefore no
2213 21. Internal re-arrangement: the function for sending a challenge and reading
2214 a response while authentication was assuming a zero-terminated challenge
2215 string. It's now changed to take a pointer and a length, to allow for
2216 binary data in such strings.
2218 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2220 23. Exim was not respecting finduser_retries when seeking the login of the
2221 uid under which it was called; it was always trying 10 times. (The default
2222 setting of finduser_retries is zero.) Also, it was sleeping after the final
2223 failure, which is pointless.
2225 24. Implemented tls_on_connect_ports.
2227 25. Implemented acl_smtp_predata.
2229 26. If the domain in control=submission is set empty, Exim assumes that the
2230 authenticated id is a complete email address when it generates From: or
2231 Sender: header lines.
2233 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2234 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2235 chown and chgrp in /bin and hostname in /usr/bin.
2237 28. Exim was keeping the "process log" file open after each use, just as it
2238 does for the main log. This opens the possibility of it remaining open for
2239 long periods when the USR1 signal hits a daemon. Occasional processlog
2240 errors were reported, that could have been caused by this. Anyway, it seems
2241 much more sensible not to leave this file open at all, so that is what now
2244 29. The long-running daemon process does not normally write to the log once it
2245 has entered its main loop, and it closes the log before doing so. This is
2246 so that log files can straightforwardly be renamed and moved. However,
2247 there are a couple of unusual error situations where the daemon does write
2248 log entries, and I had neglected to close the log afterwards.
2250 30. The text of an SMTP error response that was received during a remote
2251 delivery was being truncated at 512 bytes. This is too short for some of
2252 the long messages that one sometimes sees. I've increased the limit to
2255 31. It is now possible to make retry rules that apply only when a message has a
2256 specific sender, in particular, an empty sender.
2258 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2259 it possible to be selective about when SMTP synchronization is enforced.
2261 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2263 32. Implemented hosts_connection_nolog.
2265 33. Added an ACL for QUIT.
2267 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2270 35. Added mailbox_size and mailbox_filecount to appendfile.
2272 36. Added control = no_multiline_responses to ACLs.
2274 37. There was a bug in the logic of the code that waits for the clock to tick
2275 in the case where the clock went backwards by a substantial amount such
2276 that the microsecond fraction of "now" was more than the microsecond
2277 fraction of "then" (but the whole seconds number was less).
2279 38. Added support for the libradius Radius client library this is found on
2280 FreeBSD (previously only the radiusclient library was supported).
2286 1. When certain lookups returned multiple values in the form name=value, the
2287 quoting of the values was not always being done properly. Specifically:
2288 (a) If the value started with a double quote, but contained no whitespace,
2290 (b) If the value contained whitespace other than a space character (i.e.
2291 tabs or newlines or carriage returns) it was not quoted.
2292 This fix has been applied to the mysql and pgsql lookups by writing a
2293 separate quoting function and calling it from the lookup code. The fix
2294 should probably also be applied to nisplus, ibase and oracle lookups, but
2295 since I cannot test any of those, I have not disturbed their existing code.
2297 2. A hit in the callout cache for a specific address caused a log line with no
2298 reason for rejecting RCPT. Now it says "Previous (cached) callout
2299 verification failure".
2301 3. There was an off-by-one bug in the queryprogram router. An over-long
2302 return line was truncated at 256 instead of 255 characters, thereby
2303 overflowing its buffer with the terminating zero. As well as fixing this, I
2304 have increased the buffer size to 1024 (and made a note to document this).
2306 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2307 when Exim is waiting for an SMTP response from a remote server, Exim
2308 restarts its select() call on the socket, thereby resetting its timeout.
2309 This is not a problem when such interrupts are rare. Somebody set up a cron
2310 job to run exiwhat every 2 minutes, which is less than the normal select()
2311 timeout (5 or 10 minutes). This meant that the select() timeout never
2312 kicked in because it was always reset. I have fixed this by comparing the
2313 time when an interrupt arrives with the time at the start of the first call
2314 to select(). If more time than the timeout has elapsed, the interrupt is
2315 treated as a timeout.
2317 5. Some internal re-factoring in preparation for the addition of Sieve
2318 extensions (by MH). In particular, the "personal" test is moved to a
2319 separate function, and given an option for scanning Cc: and Bcc: (which is
2320 not set for Exim filters).
2322 6. When Exim created an email address using the login of the caller as the
2323 local part (e.g. when creating a From: or Sender: header line), it was not
2324 quoting the local part when it contained special characters such as @.
2326 7. Installed new OpenBSD configuration files.
2328 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2329 try to make them clearer.
2331 9. Callout options, other than the timeout value, were being ignored when
2332 verifying sender addresses in header lines. For example, when using
2334 verify = header_sender/callout=no_cache
2336 the cache was (incorrectly) being used.
2338 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2339 only the Cygwin environment.
2341 11. When return_path_on_delivery was set as a log selector, if different remote
2342 addresses in the same message used different return paths and parallel
2343 remote delivery occurred, the wrong values would sometimes be logged.
2344 (Whenever a remote delivery process finished, the return path value from
2345 the most recently started remote delivery process was logged.)
2347 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2348 lines when AUTH and/or TLS are in use. This is the "received protocol"
2349 field. Exim used to use "asmtp" for authenticated SMTP, without any
2350 indication (in the protocol name) for TLS use. Now it follows the RFC and
2351 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2352 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2353 names appear in log lines as well as in Received: header lines.
2355 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2356 extensions, and comparison tests, and to fix some bugs.
2358 14. Changes to the "personal" filter test:
2360 (1) The test was buggy in that it was just doing the equivalent of
2361 "contains" tests on header lines. For example, if a user's address was
2362 anne@some.where, the "personal" test would incorrectly be true for
2364 To: susanne@some.where
2366 This test is now done by extracting each address from the header in turn,
2367 and checking the entire address. Other tests that are part of "personal"
2368 are now done using regular expressions (for example, to check local parts
2369 of addresses in From: header lines).
2371 (2) The list of non-personal local parts in From: addresses has been
2372 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2373 taken from the Sieve specification recommendations.
2375 (3) If the message contains any header line starting with "List-" it is
2376 treated as non-personal.
2378 (4) The test for "circular" in the Subject: header line has been removed
2379 because it now seems ill-conceived.
2381 15. Minor typos in src/EDITME comments corrected.
2383 16. Installed latest exipick from John Jetmore.
2385 17. If headers_add on a router specified a text string that was too long for
2386 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2387 of string_sprintf() is now avoided.
2389 18. $message_body_size was not set (it was always zero) when running the DATA
2390 ACL and the local_scan() function.
2392 19. For the "mail" command in an Exim filter, no default was being set for
2393 the once_repeat time, causing a random time value to be used if "once" was
2394 specified. (If the value happened to be <= 0, no repeat happened.) The
2395 default is now 0s, meaning "never repeat". The "vacation" command was OK
2396 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2397 (I found it when inspecting the code).
2399 20. There is now an overall timeout for performing a callout verification. It
2400 defaults to 4 times the callout timeout, which applies to individual SMTP
2401 commands during the callout. The overall timeout applies when there is more
2402 than one host that can be tried. The timeout is checked before trying the
2403 next host. This prevents very long delays if there are a large number of
2404 hosts and all are timing out (e.g. when the network connections are timing
2405 out). The value of the overall timeout can be changed by specifying an
2406 additional sub-option for "callout", called "maxwait". For example:
2408 verify = sender/callout=5s,maxwait=20s
2410 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2411 to the end before writing, but this should make it even safer).
2413 22. Exim was forgetting that it had advertised PIPELINING for the second and
2414 subsequent messages on an SMTP connection. It was also not resetting its
2415 memory on STARTTLS and an internal HELO.
2417 23. When Exim logs an SMTP synchronization error within a session, it now
2418 records whether PIPELINING has been advertised or not.
2420 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2421 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2422 rather than long int.
2424 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2426 26. Added the never_mail option to autoreply.
2432 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2433 crash if the getsockname() call failed; this can happen if a connection is
2434 closed very soon after it is established. The problem was simply in the
2435 order in which certain operations were done, causing Exim to try to write
2436 to the SMTP stream before it had set up the file descriptor. The bug has
2437 been fixed by making things happen in the correct order.
2443 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2444 before the connection was closed, thus losing the rejection response.
2446 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2447 some early Solaris releases, but causes trouble in current releases where
2448 socklen_t is defined.
2450 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2453 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2456 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2457 that releasing the top of it at the end releases what was used for sub-
2458 expansions (unless the block got too big). However, discard this block if
2459 the first thing is a variable or header, so that we can use its block when
2460 it is dynamic (useful for very large $message_headers, for example).
2462 6. Lookups now cache *every* query, not just the most recent. A new, separate
2463 store pool is used for this. It can be recovered when all lookup caches are
2464 flushed. Lookups now release memory at the end of their result strings.
2465 This has involved some general refactoring of the lookup sources.
2467 7. Some code has been added to the store_xxx() functions to reduce the amount
2468 of flapping under certain conditions.
2470 8. log_incoming_interface used to affect only the <= reception log lines. Now
2471 it causes the local interface and port to be added to several more SMTP log
2472 lines, for example "SMTP connection from", and rejection lines.
2474 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2476 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2478 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2479 could be overwritten at the end of the current message (or the start of a
2480 new message if it was set in a HELO ACL). The value is now preserved for
2481 the duration of the SMTP connection.
2483 12. If a transport had a headers_rewrite setting, and a matching header line
2484 contained an unqualified address, that address was qualified, even if it
2485 did not match any rewriting rules. The underlying bug was that the values
2486 of the flags that permit the existence of unqualified sender and recipient
2487 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2488 non-local messages, and by -bnq for local messages) were not being
2489 preserved with the message after it was received.
2491 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2492 "next input=" as part of the text comprising the host identity instead of
2493 the correct text. The code was using the same buffer for two different
2494 strings. However, depending on which order the printing function evaluated
2495 its arguments, the bug did not always show up. Under Linux, for example, my
2496 test suite worked just fine.
2498 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2499 doesn't work with some older versions of Perl. It has been changed to "my",
2500 which in any case is probably the better facility to use.
2502 15. A really picky compiler found some instances of statements for creating
2503 error messages that either had too many or two few arguments for the format
2506 16. The size of the buffer for calls to the DNS resolver has been increased
2507 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2508 for addresses that have a lot of PTR records. This alleviates a problem; it
2509 does not fully solve it.
2511 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2512 buffer now truncates the list and logs the incident, which is the same
2513 action as happens when Exim is looking up a host name and its aliases.
2514 Previously in this situation something unpredictable would happen;
2515 sometimes it was "internal error: store_reset failed".
2517 18. If a server dropped the connection unexpectedly when an Exim client was
2518 using GnuTLS and trying to read a response, the client delivery process
2519 crashed while trying to generate an error log message.
2521 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2522 string, for example:
2524 warn message = H1: something\nH2: something
2526 the text was added as a single header line from Exim's point of view
2527 though it ended up OK in the delivered message. However, searching for the
2528 second and subsequent header lines using $h_h2: did not work. This has been
2529 fixed. Similarly, if a system filter added multiple headers in this way,
2530 the routers could not see them.
2532 20. Expanded the error message when iplsearch is called with an invalid key to
2533 suggest using net-iplsearch in a host list.
2535 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2536 ACLs are no longer actually imposed (and a message to that effect is
2539 22. If a "gecos" field in a passwd entry contained escaped characters, in
2540 particular, if it contained a \" sequence, Exim got it wrong when building
2541 a From: or a Sender: header from that name. A second bug also caused
2542 incorrect handling when an unquoted " was present following a character
2543 that needed quoting.
2545 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2546 was not being matched caselessly.
2548 24. Arranged for all hyphens in the exim.8 source to be escaped with
2551 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2552 properly. Recipient callout cache records were still being keyed to include
2553 the sender, even when use_sender was set false. This led to far more
2554 callouts that were necessary. The sender is no longer included in the key
2555 when use_sender is false.
2557 26. Added "control = submission" modifier to ACLs.
2559 27. Added the ${base62d: operator to decode base 62 numbers.
2561 28. dnsdb lookups can now access SRV records.
2563 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2564 the configuration file.
2566 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2567 (-v) mode. This makes the output for a verbose queue run more intelligible.
2569 31. Added a use_postmaster feature to recipient callouts.
2571 32. Added the $body_zerocount variable, containing the number of binary zero
2572 bytes in the message body.
2574 33. The time of last modification of the "new" subdirectory is now used as the
2575 "mailbox time last read" when there is a quota error for a maildir
2578 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2580 35. Added +ignore_unknown as a special item in host lists.
2582 36. Code for decoding IPv6 addresses in host lists is now included, even if
2583 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2584 address was recognized as an IP address, but was then not correctly decoded
2585 into binary, causing unexpected and incorrect effects when compared with
2592 1. Very minor rewording of debugging text in manualroute to say "list of
2593 hosts" instead of "hostlist".
2595 2. If verify=header_syntax was set, and a header line with an unqualified
2596 address (no domain) and a large number of spaces between the end of the
2597 name and the colon was received, the reception process suffered a buffer
2598 overflow, and (when I tested it) crashed. This was caused by some obsolete
2599 code that should have been removed. The fix is to remove it!
2601 3. When running in the test harness, delay a bit after writing a bounce
2602 message to get a bit more predictability in the log output.
2604 4. Added a call to search_tidyup() just before forking a reception process. In
2605 theory, someone could use a lookup in the expansion of smtp_accept_max_
2606 per_host which, without the tidyup, could leave open a database connection.
2608 5. Added the variables $recipient_data and $sender_data which get set from a
2609 lookup success in an ACL "recipients" or "senders" condition, or a router
2610 "senders" option, similar to $domain_data and $local_part_data.
2612 6. Moved the writing of debug_print from before to after the "senders" test
2615 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2616 problems for message scanning, either using a data ACL, or using
2617 local_scan() because the Received: header was not generated till after they
2618 were called (in order to set the time as the time of reception completion).
2619 I have revised the way this works. The header is now generated after the
2620 body is received, but before the ACL or local_scan() are called. After they
2621 are run, the timestamp in the header is updated.
2627 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2628 before starting a queue runner without re-exec. This happened only when
2629 deliver_drop_privilege was set or when the Exim user was set to root. The
2630 effect of the bug was that timeouts during subsequent deliveries caused
2631 crashes instead of being properly handled. The handler is now left at its
2632 default (and expected) setting.
2634 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2635 message, again when deliver_drop_privilege is set or Exim is run as root.
2636 The bug described in (1) was not present in this case, but the tidying up
2637 of the other signals was missing. I have made the two cases consistent.
2639 3. The ignore_target_hosts setting on a manualroute router was being ignored
2640 for hosts that were looked up using the /MX notation.
2642 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2645 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2646 operated on the sender address. After changing the $sender_address to <>
2647 for the sender address verify, Exim was re-instated it as the original
2648 (before rewriting) address, but remembering that it had rewritten it, so it
2649 wasn't rewriting it again. This bug also had the effect of breaking the
2650 sender address verification caching when the sender address was rewritten.
2652 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2653 This has been changed so that if the ip literal address matches
2654 ignore_target_hosts, the router declines.
2656 7. Added expansion conditions match_domain, match_address, and match_local_
2657 part (NOT match_host).
2659 8. The placeholder for the Received: header didn't have a length field set.
2661 9. Added code to Exim itself and to exim_lock to test for a specific race
2662 condition that could lead to file corruption when using MBX delivery. The
2663 issue is with the lockfile that is created in /tmp. If this file is removed
2664 after a process has opened it but before that process has acquired a lock,
2665 there is the potential for a second process to recreate the file and also
2666 acquire a lock. This could lead to two Exim processes writing to the file
2667 at the same time. The added code performs the same test as UW imapd; it
2668 checks after acquiring the lock that its file descriptor still refers to
2669 the same named file.
2671 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2672 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2675 11. Added the smtp_active_hostname option. If used, this will typically be made
2676 to depend on the incoming interface address. Because $interface_address is
2677 not set up until the daemon has forked a reception process, error responses
2678 that can happen earlier (such as "too many connections") no longer contain
2681 12. If an expansion in a condition on a "warn" statement fails because a lookup
2682 defers, the "warn" statement is abandoned, and the next ACL statement is
2683 processed. Previously this caused the whole ACL to be aborted.
2685 13. Added the iplsearch lookup type.
2687 14. Added ident_timeout as a log selector.
2689 15. Added tls_certificate_verified as a log selector.
2691 16. Added a global option tls_require_ciphers (compare the smtp transport
2692 option of the same name). This controls incoming TLS connections.
2694 17. I finally figured out how to make tls_require_ciphers do a similar thing
2695 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2696 before starting the TLS session.
2698 18. Tabs are now shown as \t in -bP output.
2700 19. If the log selector return_path_on_delivery was set, Exim crashed when
2701 bouncing a message because it had too many Received: header lines.
2703 20. If two routers both had headers_remove settings, and the first one included
2704 a superfluous trailing colon, the final name in the first list and the
2705 first name in the second list were incorrectly joined into one item (with a
2706 colon in the middle).
2712 1. Added -C and -D options to the exinext utility, mainly to make it easier
2713 to include in the automated testing, but these could be helpful when
2714 multiple configurations are in use.
2716 2. The exinext utility was not formatting the output nicely when there was
2717 an alternate port involved in the retry record key, nor when there was a
2718 message id as well (for retries that were specific to a specific message
2719 and a specific host). It was also confused by IPv6 addresses, because of
2720 the additional colons they contain. I have fixed the IPv4 problem, and
2721 patched it up to do a reasonable job for IPv6.
2723 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2724 delivery, the log line now contains "pipelined" if PIPELINING was used.
2726 4. An SMTP transport process used to panic and die if the bind() call to set
2727 an explicit outgoing interface failed. This has been changed; it is now
2728 treated in the same way as a connect() failure.
2730 5. A reference to $sender_host_name in the part of a conditional expansion
2731 that was being skipped was still causing a DNS lookup. This no longer
2734 6. The def: expansion condition was not recognizing references to header lines
2735 that used bh_ and bheader_.
2737 7. Added the _cache feature to named lists.
2739 8. The code for checking quota_filecount in the appendfile transport was
2740 allowing one more file than it should have been.
2742 9. For compatibility with Sendmail, the command line option
2750 and sets the incoming protocol and host name (for trusted callers). The
2751 host name and its colon can be omitted when only the protocol is to be set.
2752 Note the Exim already has two private options, -pd and -ps, that refer to
2753 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2754 "s", but I don't think that's a major issue.
2756 10. A number of refactoring changes to the code, none of which should affect
2759 (a) The number of logging options was getting close to filling up the
2760 32-bit word that was used as a bit map. I have split them into two classes:
2761 those that are passed in the argument to log_write(), and those that are
2762 only ever tested independently outside of that function. These are now in
2763 separate 32-bit words, so there is plenty of room for expansion again.
2764 There is no change in the user interface or the logging behaviour.
2766 (b) When building, for example, log lines, the code previously used a
2767 macro that called string_cat() twice, in order to add two strings. This is
2768 not really sufficiently general. Furthermore, there was one instance where
2769 it was actually wrong because one of the argument was used twice, and in
2770 one call a function was used. (As it happened, calling the function twice
2771 did not affect the overall behaviour.) The macro has been replaced by a
2772 function that can join an arbitrary number of extra strings onto a growing
2775 (c) The code for expansion conditions now uses a table and a binary chop
2776 instead of a serial search (which was left over from when there were very
2777 few conditions). Also, it now recognizes conditions like "pam" even when
2778 the relevant support is not compiled in: a suitably worded error message is
2779 given if an attempt is made to use such a condition.
2781 11. Added ${time_interval:xxxxx}.
2783 12. A bug was causing one of the ddress fields not to be passed back correctly
2784 from remote delivery subprocesses. The field in question was not being
2785 subsequently used, so this caused to problems in practice.
2787 13. Added new log selectors queue_time and deliver_time.
2789 14. Might have fixed a bug in maildirsizefile handling that threw up
2790 "unexpected character" debug warnings, and recalculated the data
2791 unnecessarily. In any case, I expanded the warning message to give more
2794 15. Added the message "Restricted characters in address" to the statements in
2795 the default ACL that block characters like @ and % in local parts.
2797 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2798 Three changes have been made:
2800 (a) There was a serious bug; a negative response to MAIL caused the whole
2801 recipient domain to be cached as invalid, thereby blocking all messages
2802 to all local parts at the same domain, from all senders. This bug has
2803 been fixed. The domain is no longer cached after a negative response to
2804 MAIL if the sender used is not empty.
2806 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2809 (c) A new callout option, "use_sender" has been added for people who want
2810 the modified behaviour.
2816 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2819 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2820 indeed breaks things for older releases.
2822 3. Added additional logging to the case where there is a problem reading data
2823 from a filter that is running in a subprocess using a pipe, in order to
2824 try to track down a specific problem.
2826 4. Testing facility fudge: when running in the test harness and attempting
2827 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2828 getting "No route to host". Convert this to a timeout.
2830 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2833 6. Some OS don't have socklen_t but use size_t instead. This affects the
2834 fifth argument of getsockopt() amongst other things. This is now
2835 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2836 can be set for individual OS. I have set it for SunOS5, OSF1, and
2837 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2838 some earlier ones do not.
2840 7. Change 4.30/15 was not doing the test caselessly.
2842 8. The standard form for an IPv6 address literal was being rejected by address
2843 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2844 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2845 this, as well as the form without the "IPv6" on the front (but only when
2846 address literals are enabled, of course).
2848 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2850 10. Exim crashed if a message with an empty sender address specified by -f
2851 encountered a router with an errors_to setting. This could be provoked only
2852 by a command such as
2856 where an empty string was supplied; "<>" did not hit this bug.
2858 11. Installed PCRE release 4.5.
2860 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2861 remained set. It is now erased.
2863 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2864 times from message ids (which are base 36 rather than the normal 62).
2866 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2867 were being counted as actual protocol errors, and logged if the log
2868 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2869 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2870 and DATA following a set of rejected RCPTs do not count as protocol errors.
2871 In other words, Exim assumes they were pipelined, though this may not
2872 actually be the case. Of course, in all cases the client gets an
2873 appropriate error code.
2875 15. If a lookup fails in an ACL condition, a message about the failure may
2876 be available; it is used if testing the ACL cannot continue, because most
2877 such messages specify what the cause of the deferral is. However, some
2878 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2879 that caused an old message to be retained and used if a later statement
2880 caused a defer, replacing the real cause of the deferral.
2882 16. If an IP address had so many PTR records that the DNS lookup buffer
2883 was not large enough to hold them, Exim could crash while trying to process
2884 the truncated data. It now detects and logs this case.
2886 17. Further to 4.21/58, another change has been made: if (and only if) the
2887 first line of a message (the first header line) ends with CRLF, a bare LF
2888 in a subsequent header line has a space inserted after it, so as not to
2889 terminate the header.
2891 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2892 unnecessarily, used atoi() instead of strtol(), and didn't check the
2893 termination when getting file sizes from file names by regex.
2895 19. Completely re-implemented the support for maildirsize files, in the light
2896 of a number of problems with the previous contributed implementation
2897 (4.30/29). In particular:
2899 . If the quota is zero, the maildirsize file is maintained, but no quota is
2902 . If the maildir directory does not exist, it is created before any attempt
2903 to write a maildirsize file.
2905 . The quota value in the file is just a cache; if the quota is changed in
2906 the transport, the new value overrides.
2908 . A regular expression is available for excluding directories from the
2911 20. The autoreply transport checks the characters in options that define the
2912 message's headers; it allows continued headers, but it was checking with
2913 isspace() after an embedded newline instead of explicitly looking for a
2916 21. If all the "regular" hosts to which an address was routed had passed their
2917 expiry times, and had not reached their retry times, the address was
2918 bounced, even if fallback hosts were defined. Now Exim should go on to try
2921 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2922 equivalent code in the SMTP transport. Some hosts send humungous responses
2923 to HELO/EHLO, more than 1024 it seems.
2925 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2926 gives compiler warnings in some environments. I've now done it "properly",
2929 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2930 (because of problems with the built-in one) was declared to return uschar *
2931 instead of char *, causing compiler failure.
2933 25. Fixed a file descriptor leak when processing alias/forward files.
2935 26. Fixed a minor format string issue in dbfn.c.
2937 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2939 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2940 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2943 29. When a sender address is verified, it is cached, to save repeating the test
2944 when there is more than one recipient in a message. However, when the
2945 verification involves a callout, it is possible for different callout
2946 options to be set for different recipients. It is too complicated to keep
2947 track of this in the cache, so now Exim always runs a verification when a
2948 callout is required, relying on the callout cache for the optimization.
2949 The overhead is duplication of the address routing, but this should not be
2952 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2953 to be verified with callout=postmaster, and the main callout worked but the
2954 postmaster check failed, the verification correctly failed. However, if a
2955 subsequent RCPT command asked for sender verification *without* the
2956 postmaster check, incorrect caching caused this verification also to fail,
2959 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2960 it was not caching the DNS options (qualify_single, search_parents) that
2961 were used when the lookup failed. A subsequent lookup with different
2962 options therefore always gave the same answer, though there were cases
2963 where it should not have. (Example: a "domains = !$mx_any" option on a
2964 dnslookup router: the "domains" option is always processed without any
2965 widening, but the router might have qualify_single set.) Now Exim uses the
2966 cached value only when the same options are set.
2968 32. Added John Jetmore's "exipick" utility to the distribution.
2970 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2971 than a timeout (e.g. a certificate is required, and is not provided), an
2972 Exim server now closes the connection immediately. Previously it waited for
2973 the client to close - but if the client is SSL, it seems that they each
2974 wait for each other, leading to a delay before one of them times out.
2976 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2977 maintained 0.8.x compatibility because I don't think many are using it, and
2978 it is clearly obsolete.
2980 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2983 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2984 client certificate was expired. A simple patch fixes this, though I don't
2985 understand the full logic of why the verify callback is called multiple
2988 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2989 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2990 which causes problems with some clients (such as the Certicom SSL Plus
2991 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2992 disables the coutermeasure allowing Eudora to connect."
2994 38. Exim was not checking that a write() to a log file succeeded. This could
2995 lead to Bad Things if a log got too big, in particular if it hit a file
2996 size limit. Exim now panics and dies if it cannot write to a log file, just
2997 as it does if it cannot open a log file.
2999 39. Modified OS/Makefile-Linux so that it now contains
3001 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3003 The two -D definitions ensure that Exim is compiled with large file
3004 support, which makes it possible to handle log files that are bigger than
3007 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3008 instance) a domain was checked against a named list that involved a lookup,
3009 causing $domain_data to be set, then another domain was checked against the
3010 same list, then the first domain was re-checked, the value of $domain_data
3011 after the final check could be wrong. In particular, if the second check
3012 failed, it could be set empty. This bug probably also applied to
3015 41. The strip_trailing_dot option was not being applied to the address given
3016 with the -f command-line option.
3018 42. The code for reading a message's header from the spool was incrementing
3019 $received_count, but never initializing it. This meant that the value was
3020 incorrect (doubled) while delivering a message in the same process in which
3021 it was received. In the most common configuration of Exim, this never
3022 happens - a fresh exec is done - but it can happen when
3023 deliver_drop_privilege is set.
3025 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3026 it now includes up to 150 characters of the unexpected data in the log
3029 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3030 and building data strings. The size of both of these buffers was 10 000
3031 bytes - far larger than anybody would *ever* want, thought I. Needless to
3032 say, somebody hit the limit. I have increased the maximum line length to
3033 20 000 and the maximum data length of concatenated lines to 100 000. I have
3034 also fixed two bugs, because there was no checking on these buffers. Tsk,
3035 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3036 buffer is too small.
3038 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3039 lsearch lookups. Now it does.
3041 46. When parsing a route_list item in a manualroute router, a fixed-length
3042 buffer was used for the list of hosts. I made this 1024 bytes long,
3043 thinking that nobody would ever have a list of hosts that long. Wrong.
3044 Somebody had a whole pile of complicated expansion conditions, and the
3045 string was silently truncated, leading to an expansion error. It turns out
3046 that it is easier to change to an unlimited length (owing to other changes
3047 that have happened since this code was originally written) than to build
3048 structure for giving a limitation error. The length of the item that
3049 expands into the list of hosts is now unlimited.
3051 47. The lsearch lookup could not handle data where the length of text line was
3052 more than 4095 characters. Such lines were truncated, leading to shortened
3053 data being returned. It should now handle lines of any length.
3055 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3056 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3059 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3060 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3061 no longer happens in comment lines. A list of replacements is now placed
3062 at the head of all of the source files, except those whose only change is
3063 to replace PERL_COMMAND in the very first #! line.
3065 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3066 messages on the queue, with a bottom-up merge sort, using code contributed
3067 by Michael Haardt. This should make operations like -bp somewhat faster on
3068 large queues. It won't affect queue runners, except when queue_run_in_order
3071 51. Installed eximstats 1.31 in the distribution.
3073 52. Added support for SRV lookups to the dnslookup router.
3075 53. If an ACL referred to $message_body or $message_body_end, the value was not
3076 reset for any messages that followed in the same SMTP session.
3078 54. The store-handling optimization for building very long strings was not
3079 differentiating between the different store pools. I don't think this
3080 actually made any difference in practice, but I've tidied it.
3082 55. While running the routers to verify a sender address, $sender_address
3083 was still set to the sender address. This is wrong, because when routing to
3084 send a bounce to the sender, it would be empty. Therefore, I have changed
3085 it so that, while verifying a sender address, $sender_address is set to <>.
3086 (There is no change to what happens when verifying a recipient address.)
3088 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3089 target A or AAAA records (if not already returned) without resetting the
3090 qualify_single or search_parents options of the DNS resolver. These are
3091 inappropriate in this case because the targets of MX and SRV records must
3092 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3093 target that, when qualified, matched something in the local domain. These
3094 two options are now turned off when doing these lookups.
3096 57. It seems that at least some releases of Reiserfs (which does not have the
3097 concept of a fixed number of inodes) returns zero and not -1 for the
3098 number of available inodes. This interacted badly with check_spool_inodes,
3099 which assumed that -1 was the "no such thing" setting. What I have done is
3100 to check that the total number of inodes is greater than zero before doing
3101 the test of how many are available.
3103 58. When a "warn" ACL statement has a log_message modifier, the message is
3104 remembered, and not repeated. This is to avoid a lot of repetition when a
3105 message has many recipients that cause the same warning to be written.
3106 Howewer, Exim was preserving the list of already written lines for an
3107 entire SMTP session, which doesn't seem right. The memory is now reset if a
3108 new message is started.
3110 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3111 debugging output unless log_rewrite was also set.
3113 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3114 of (int)(handle) when we know that handle contains (void *)(-1).
3116 61. The Exim daemon panic-logs an error return when it closes the incoming
3117 connection. However "connection reset by peer" seems to be common, and
3118 isn't really an error worthy of noting specially, so that particular error
3121 62. When Exim is trying to find all the local interfaces, it used to panic and
3122 die if the ioctl to get the interface flags failed. However, it seems that
3123 on at least one OS (Solaris 9) it is possible to have an interface that is
3124 included in the list of interfaces, but for which you get a failure error
3125 for this call. This happens when the interface is not "plumbed" into a
3126 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3127 failure of the "get flags" call assumes that the interface is down.
3129 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3130 makes life easier for people who are doing arithmetic on fields extracted
3131 from dates, where you often get leading zeros that should not be
3132 interpreted as octal.
3134 64. Added qualify_domain to the redirect router, to override the global
3137 65. If a pathologically long header line contained very many addresses (the
3138 report of this problem mentioned 10 000) and each of them was rewritten,
3139 Exim could use up a very large amount of memory. (It kept on making new
3140 copies of the header line as it rewrote, and never released the old ones.)
3141 At the expense of a bit more processing, the header rewriting function has
3142 been changed so that it no longer eats memory in this way.
3144 66. The generation of the Received: header has been moved from the time that a
3145 message starts to be received, to the time that it finishes. The timestamp
3146 in the Received: header should now be very close to that of the <= log
3147 line. There are two side-effects of this change:
3149 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3150 logged header lines no longer include the local Received: line, because
3151 it has not yet been created. The same applies to a copy of the message
3152 that is returned to a non-SMTP sender when a message is rejected.
3154 (b) When a filter file is tested using -bf, no additional Received: header
3155 is added to the test message. After some thought, I decided that this
3158 This change does not affect the value of $received_for. It is still set
3159 after address rewriting, but before local_scan() is called.
3161 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3163 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3164 gave an unhelpful panic error message, and a defer error. I have managed to
3165 change this behaviour so that it now rejects any supplied certificate,
3166 which seems right, as the list of acceptable certificates is empty.
3168 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3169 gave an unhelpful defer error. I have not managed to make this reject any
3170 supplied certificates, but the error message it gives is "no certificate
3171 supplied", which is not helpful.
3173 70. exigrep's output now also includes lines that are not associated with any
3174 message, but which match the given pattern. Implemented by a patch from
3175 Martin Sluka, which also tidied up the Perl a bit.
3177 71. Recipient callout verification, like sender verification, was using <> in
3178 the MAIL FROM command. This isn't really the right thing, since the actual
3179 sender may affect whether the remote host accepts the recipient or not. I
3180 have changed it to use the actual sender in the callout; this means that
3181 the cache record is now keyed on a recipient/sender pair, not just the
3182 recipient address. There doesn't seem to be a real danger of callout loops,
3183 since a callout by the remote host to check the sender would use <>.
3184 [SEE ABOVE: changed after hitting problems.]
3186 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3187 temporary errors. However, in the case of such a code being given after
3188 the end of a data transmission (i.e. after ".") Exim was failing to write
3189 a retry record for the message. (Yes, there was some broken host that was
3190 actually sending 8xx at this point.)
3192 73. An unknown lookup type in a host list could cause Exim to panic-die when
3193 the list was checked. (An example that provoked this was putting <; in the
3194 middle of a list instead of at the start.) If this happened during a DATA
3195 ACL check, a -D file could be left lying around. This kind of configuration
3196 error no longer causes Exim to die; instead it causes a defer errror. The
3197 incident is still logged to the main and panic logs.
3199 74. Buglet left over from Exim 3 conversion. The message "too many messages
3200 in one connection" was written to the rejectlog but not the mainlog, except
3201 when address rewriting (yes!) was being logged.
3203 75. Added write_rejectlog option.
3205 76. When a system filter was run not as root (that is, when system_filter_user
3206 was set), the values of the $n variables were not being returned to the
3207 main process; thus, they were not subsequently available in the $sn
3210 77. Added +return_path_on_delivery log selector.
3212 78. A connection timeout was being treated differently from recipients deferred
3213 when testing hosts_max_try with a message that was older than the host's
3214 retry timeout. (The host should not be counted, thus allowing all hosts to
3215 be tried at least once before bouncing.) This may have been the cause of an
3216 occasionally reported bug whereby a message would remain on the queue
3217 longer than the retry timeout, but would be bounced if a delivery was
3218 forced. I say "may" because I never totally pinned down the problem;
3219 setting up timeout/retry tests is difficult. See also the next item.
3221 79. The ultimate address timeout was not being applied to errors that involved
3222 a combination of host plus message (for example, a timeout on a MAIL
3223 command). When an address resolved to a number of possible hosts, and they
3224 were not all tried for each delivery (e.g. because of hosts_max_try), a
3225 message could remain on the queue longer than the retry timeout.
3227 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3230 81. Fixed an obscure SMTP outgoing bug which required at least the following
3231 conditions: (a) there was another message waiting for the same server;
3232 (b) the server returned 5xx to all RCPT commands in the first message so
3233 that the message was not completed; (c) the server dropped the connection
3234 or gave a negative response to the RSET that Exim sends to abort the
3235 transaction. The observed case was a dropped connection after DATA that had
3236 been sent in pipelining mode. That is, the server had advertised PIPELINING
3237 but was not implementing it correctly. The effect of the bug was incorrect
3238 behaviour, such as trying another host, and this could lead to a crash.
3244 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3245 and daemon.c were passed as pointers to ints; they should have been
3246 pointers to socklen_t variables (which are typically unsigned ints).
3248 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3251 3. Fixed a really odd bug that affected only the testing scheme; patching a
3252 certain fixed string in the binary changed the value of another string that
3253 happened to be identical to the end of the original first string.
3255 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3256 name", it returns that address as the IP address. On some operating
3257 systems (e.g. Solaris), it also passes back the IP address string as the
3258 "host name". However, on others (e.g. Linux), it passes back an empty
3259 string. Exim wasn't checking for this, and was changing the host name to an
3260 empty string, assuming it had been canonicized.
3262 5. Although rare, it is permitted to have more than one PTR record for a given
3263 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3264 all the names associated with an address, because they do in Solaris.
3265 However, it seems that they do not in Linux for data that comes from the
3266 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3267 I found this out when I moved to a new Linux workstation and tried to run
3268 the Exim test suite.
3270 To get round this problem I have changed the code so that it now does its
3271 own call to the DNS to look up PTR records when searching for a host name.
3272 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3273 addresses that are only in /etc/hosts are still found.
3275 This behaviour is, however, controlled by an option called host_lookup_
3276 order, which defaults to "bydns:byaddr". If people want to use the other
3277 order, or indeed, just use one or the other means of lookup, they can
3278 specify it in this variable.
3280 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3281 some operating systems, this comes back from gethostbyaddr() as an empty
3282 string, and this is what Exim used to test for. However, it seems that in
3283 other systems, "." is yielded. Exim now tests for this case too.
3285 7. The values of check_spool_space and check_log_space are now held internally
3286 as a number of kilobytes instead of an absolute number of bytes. If a
3287 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3288 kilobyte. This means that much larger values can be stored.
3290 8. Exim monitor: an attempt to get the action menu when not actually pointing
3291 at a message produces an empty menu entitled "No message selected". This
3292 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3293 no entries in it ("Shell widget menu has zero width and/or height"). So I
3294 have added a single, blank menu entry in this case.
3296 9. Added ${quote_local_part.
3298 10. MIME decoding is now applied to the contents of Subject: header lines when
3301 11. Now that a reference to $sender_host_address automatically causes a reverse
3302 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3303 host lookup before query-style lookups in lists that might use this
3304 variable. This has therefore been abolished, and the "net-" prefix is no
3305 longer necessary for query-style lookups.
3307 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3308 have been a typo for LFLAGS, so it has been changed.
3310 13. The install script calls Exim with "-C /dev/null" in order to find the
3311 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3312 to be output. Howeve, since Exim outputs its version number before the
3313 error, it didn't break the script. It just looked ugly. I fixed this by
3314 always allowing "-C /dev/null" if the caller is root.
3316 14. Ignore overlarge ACL variable number when reading spool file - insurance
3317 against a later release with more variables having written the file.
3319 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3320 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3321 this, as well as the form without the "IPv6" on the front.
3323 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3324 OS/Makefile-Darwin file.
3326 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3327 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3329 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3330 variable. However, some OS use ints for inodes. I've added cast to long int
3331 to get rid of the compiler warning.
3333 19. I had forgotten to lock out "/../" in configuration file names when
3334 ALT_CONFIG_PREFIX was set.
3336 20. Routers used for verification do not need to specify transports. However,
3337 if such a router generated a host list, and callout was configured, Exim
3338 crashed, because it could not find a port number from the (non-existent)
3339 transport. It now assumes port 25 in this circumstance.
3341 21. Added the -t option to exigrep.
3343 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3344 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3345 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3346 them is set, LOOKUP_LSEARCH is forced.
3348 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3350 24. Added sender and host information to the "rejected by local_scan()" log
3351 line; previously there was no indication of these.
3353 25. Added .include_if_exists.
3355 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3356 Linux. It turns out that not all file systems support this. Apparently some
3357 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3358 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3359 not supported on the file descriptor, is now ignored when Exim is trying to
3360 sync a directory. This applies only to Linux.
3362 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3364 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3365 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3368 29. Added support for maildirsize files from supplied patch (modified a bit).
3370 30. The use of :fail: followed by an empty string could lead Exim to respond to
3371 sender verification failures with (e.g.):
3373 550 Verification failed for <xxx>
3374 550 Sender verify failed
3376 where the first response line was missing the '-' that indicates it is not
3377 the final line of the response.
3379 31. The loop for finding the name of the user that called Exim had a hardwired
3380 limit of 10; it now uses the value of finduser_retries, which is used for
3381 all other user lookups.
3383 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3386 33. Exim was neglecting to zero errno before one call of strtol() when
3387 expanding a string and expecting an integer value. On some systems this
3388 resulted in spurious "integer overflow" errors. Also, it was casting the
3389 result into an int without checking.
3391 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3392 did not work. The code looks as if it has *never* worked, though it appears
3393 to have been documented since at least releast 1.62. I have made it work.
3395 35. The "timeout_DNS" error in retry rules, also documented since at least
3396 1.62, also never worked. As it isn't clear exactly what this means, and
3397 clearly it isn't a major issue, I have abolished the feature by treating it
3398 as "timeout", and writing a warning to the main and panic logs.
3400 36. The display of retry rules for -brt wasn't always showing the error code
3403 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3404 timeout_connect_A, timeout_connect_MX.
3406 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3407 to the empty sender.
3409 39. The daemon was not analysing the content of -oX till after it had closed
3410 stderr and disconnected from the controlling terminal. This meant that any
3411 syntax errors were only noted on the panic log, and the return code from
3412 the command was 0. By re-arranging the code a little, I've made the
3413 decoding happen first, so such errors now appear on stderr, and the return
3414 code is 1. However, the actual setting up of the sockets still happens in
3415 the disconnected process, so errors there are still only recorded on the
3418 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3419 connections (as happens on some IP stacks) was logged at start up time as
3420 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3421 it from "IPv6 and IPv4", which means that two separate sockets are being
3424 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3425 says whether AF_INET or AF_INET6 was passed as an argument.
3427 42. Exiwhat output was messed up when time zones were included in log
3430 43. Exiwhat now gives more information about the daemon's listening ports,
3431 and whether -tls-on-connect was used.
3433 44. The "port" option of the smtp transport is now expanded.
3435 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3436 silently ignored. Now an error message is written to the main and panic
3439 46. There's a new ACL modifier called "logwrite" which writes to a log file
3440 as soon as it is encountered.
3442 47. Added $local_user_uid and $local_user_gid at routing time.
3444 48. Exim crashed when trying to verify a sender address that was being
3447 49. Exim was recognizing only a space character after ".include". It now also
3448 recognizes a tab character.
3450 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3451 extracting the relevant information from the specification. The man page no
3452 longer contains scrambled data for the -d option, and I've added a section
3453 at the front about calling Exim under different names.
3455 51. Added "extra_headers" argument to the "mail" command in filter files.
3457 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3460 53. Installed eximstats 1.29.
3462 54. Added transport_filter_timeout as a generic transport option.
3464 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3465 Cc: header lines. This was required by RFC 822, but it not required by RFC
3468 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3469 incoming messages that did not have them. Now it does so only if the
3470 message originates locally, that is, if there is no associated remote host
3471 address. When Resent- header lines are present, this applies to the Resent-
3472 lines rather than the non-Resent- lines.
3474 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3475 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3477 58. Messages for configuration errors now include the name of the main
3478 configuration file - useful now that there may be more than one file in a
3479 list (.included file names were always shown).
3481 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3482 for those rare installations that do not start the daemon as root or run it
3483 setuid root. I've cut out the call to initgroups() if the daemon is not
3486 60. The Exim user and group can now be bound into the binary as text strings
3487 that are looked up at the start of Exim's processing.
3489 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3491 62. Added $mailstore_basename variable.
3493 63. Installed patch to sieve.c from Michael Haardt.
3495 64. When Exim failed to open the panic log after failing to open the main log,
3496 the original message it was trying to log was written to stderr and debug
3497 output, but if they were not available (the usual case in production), it
3498 was lost. Now it is written to syslog before the two lines that record the
3499 failures to open the logs.
3501 65. Users' Exim filters run in subprocesses under the user's uid. It is
3502 possible for a "deliver" command or an alias in a "personal" command to
3503 provoke an address rewrite. If logging of address rewriting is configured,
3504 this fails because the process is not running as root or exim. There may be
3505 a better way of dealing with this, but for the moment (because 4.30 needs
3506 to be released), I have disabled address rewrite logging when running a
3507 filter in a non-root, non-exim process.
3513 1. The buildconfig auxiliary program wasn't quoting the value set for
3514 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3515 not defined. This bug was masked in 4.22 by the effect that was fixed in
3518 2. Some messages that were rejected after a message id was allocated were
3519 shown as "incomplete" by exigrep. It no longer does this for messages that
3520 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3522 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3523 have allow_domain_literals set, the ID did not get logged in the <= line.
3524 Domain literals are now always recognized in Message-ID: header lines.
3526 4. The first argument for a ${extract expansion item is the key name or field
3527 number. Leading and trailing spaces in this item were not being ignored,
3528 causing some misleading effects.
3530 5. When deliver_drop_privilege was set, single queue runner processes started
3531 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3532 same command in the process it spins off) were not dropping privilege.
3534 6. When the daemon running as "exim" started a queue runner, it always
3535 re-executed Exim in the spun-off process. This is a waste of effort when
3536 deliver_drop_privilege is set. The new process now just calls the
3537 queue-runner function directly.
3543 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3546 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3547 ignored. Though the use of -oP was forcing the writing of a pid file, it
3548 was always written to the default place.
3550 3. If the message "no IP address found for host xxxx" is generated during
3551 incoming verification, it is now followed by identification of the incoming
3552 connection (so you can more easily find what provoked it).
3554 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3556 5. Added some features to "harden" Exim a bit more against certain attacks:
3558 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3559 be put in Local/Makefile. This is like the never_users runtime option,
3560 but it cannot be overridden. The default setting is "root".
3562 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3563 prefix string with which any file named in a -C command line option
3566 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3567 is retained for -C and -D only if the caller of Exim is root. Without
3568 it, the exim user may also use -C and -D and retain privilege.
3570 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3571 command line option is disabled.
3573 6. Macro names set by the -D option must start with an upper case letter, just
3574 like macro names defined in the configuration file.
3576 7. Added "dereference=" facility to LDAP.
3578 8. Two instances of the typo "uknown" in the source files are fixed.
3580 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3581 the Configure-Makefile script screwed up while processing it.
3583 10. Incorporated PCRE 4.4.
3585 11. The SMTP synchronization check was not operating right at the start of an
3586 SMTP session. For example, it could not catch a HELO sent before the client
3587 waited for the greeting. There is now a check for outstanding input at the
3588 point when the greeting is written. Because of the duplex, asynchronous
3589 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3590 way, but not yet received, when the check is performed.
3592 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3593 on TCP/IP sockets, because this apparently causes some broken clients to
3596 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3597 unchanged) from the Cygwin maintainer.
3599 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3600 was supported instead of testing for mbx. Effectively a typo.
3602 15. The spa authenticator server code was not checking that the input it
3603 received was valid base64.
3605 16. The debug output line for the "set" modifier in ACLs was not showing the
3606 name of the variable that was being set.
3608 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3610 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3611 lookup on its own. Something else was needed to trigger the lookup. For
3612 example, a match in host_lookup or the need for a host name in a host list.
3613 Now, if $sender_host_name is referenced and the host name has not yet been
3614 looked up, a lookup is performed. If the lookup fails, the variable remains
3615 empty, and $host_lookup_failed is set to "1".
3617 19. Added "eqi" as a case-independent comparison operator.
3619 20. The saslauthd authentication condition could segfault if neither service
3620 nor realm was specified.
3622 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3623 error message that was logged was misleading, and incoming SMTP
3624 connections were dropped. The message is now more accurate, and temporary
3625 errors are given to SMTP connections.
3627 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3628 (except RSET) with a 421 error, until QUIT is received. However, it was
3629 failing to send a response to QUIT.
3631 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3632 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3633 ACL is now run after the helo_try_verify_hosts code.
3635 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3636 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3637 case-independent, so other case variants are also recognized). Apparently
3638 some systems use these upper case variants.
3640 25. If more than two messages were waiting for the same host, and a transport
3641 filter was specified for the transport, Exim sent two messages over the
3642 same TCP/IP connection, and then failed with "socket operation on non-
3643 socket" when it tried to send the third.
3645 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3647 27. The extern definition of crypt16() in expand.c was not being excluded when
3648 the OS had its own crypt16() function.
3650 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3651 as a preferred synonym for return_size_limit, both as an option and as an
3654 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3656 30. Changed the default configuration ACL to relax the local part checking rule
3657 for addresses that are not in any local domains. For these addresses,
3658 slashes and pipe symbols are allowed within local parts, but the sequence
3659 /../ is explicitly forbidden.
3661 31. SPA server authentication was not clearing the challenge buffer before
3664 32. log_message in a "warn" ACL statement was writing to the reject log as
3665 well as to the main log, which contradicts the documentation and doesn't
3666 seem right (because no rejection is happening). So I have stopped it.
3668 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3669 However, I am unable to do any testing of this.
3671 34. Fixed an infelicity in the appendfile transport. When checking directories
3672 for a mailbox, to see if any needed to be created, it was accidentally
3673 using path names with one or more superfluous leading slashes; tracing
3674 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3676 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3677 contents are added to the log line that is written for every discarded
3678 recipient. (Previously a log_message setting was ignored.)
3680 36. The ${quote: operator now quotes the string if it is empty.
3682 37. The install script runs exim in order to find its version number. If for
3683 some reason other than non-existence or emptiness, which it checks, it
3684 could not run './exim', it was installing it with an empty version number,
3685 i.e. as "exim-". This error state is now caught, and the installation is
3688 38. An argument was missing from the function that creates an error message
3689 when Exim fails to connect to the socket for saslauthd authentication.
3690 This could cause Exim to crash, or give a corrupted message.
3692 39. Added isip, isip4, and isip6 to ${if conditions.
3694 40. The ACL variables $acl_xx are now saved with the message, and can be
3695 accessed later in routers, transports, and filters.
3697 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3698 strings in the file are not string-expanded.
3700 42. If a MAIL command specified a SIZE value that was too large to fit into an
3701 int variable, the check against message_size_limit failed. Such values are
3702 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3703 day this will have to be increased, but I don't think I want to be around
3704 when emails are that large.
3711 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3712 iconv() is not standard in FreeBSD.
3714 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3715 IPv6 enabled. The observed symptom was a segmentation fault on return from
3716 the function os_common_find_running_interfaces() in src/os.c.
3718 3. In the check_special_case() function in daemon.c I had used "errno" as an
3719 argument name, which causes warnings on some systems. This was basically a
3720 typo, since it was named "eno" in the comments!
3722 4. The code that waits for the clock to tick (at a resolution of some fraction
3723 of a second) so as to ensure message-id uniqueness was always waiting for
3724 at least one whole tick, when it could have waited for less. [This is
3725 almost certainly not relevant at current processor speeds, where it is
3726 unlikely to ever wait at all. But we try to future-proof.]
3728 5. The function that sleeps for a time interval that includes fractions of a
3729 second contained a race. It did not block SIGALRM between setting the
3730 timer, and suspending (a couple of lines later). If the interval was short
3731 and the sigsuspend() was delayed until after it had expired, the suspension
3732 never ended. On busy systems this could lead to processes getting stuck for
3735 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3736 process, before it forks any delivery processes. The open lookup caching
3737 mechanism meant that the open file or database connection was passed into
3738 the delivery process. The problem was that delivery processes always tidy
3739 up cached lookup data. This could cause a problem for the next delivery
3740 process started by the queue runner, because the external queue runner
3741 process does not know about the closure. So the next delivery process
3742 still has data in the lookup cache. In the case of a file lookup, there was
3743 no problem because closing a file descriptor in a subprocess doesn't affect
3744 the parent. However, if the lookup was caching a connection to a database,
3745 the connection was closed, and the second delivery process was likely to
3746 see errors such as "PGSQL: query failed: server closed the connection
3747 unexpectedly". The problem has been fixed by closing all cached lookups
3748 in a queue runner before running a delivery process.
3750 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3751 seem to have the "const" qualifier which it has on other OS. I've
3754 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3755 *of the same type* (client or server) with the same public name that an
3756 error should be diagnosed.
3758 9. When Exim looked up a host name for an IP address, but failed to find the
3759 original IP address when looking up the host name (a safety check), it
3760 output the message "<ip address> does not match any IP for NULL", which was
3761 confusing, to say the least. The bug was that the host name should have
3762 appeared instead of "NULL".
3764 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3765 user that is built into the binary, and the -C or -D options is used, root
3766 privilege is dropped before the configuration file is read. In addition,
3767 logging is switched to stderr instead of the normal log files. If the
3768 configuration then re-defines the Exim user, the unprivileged environment
3769 is probably not what is expected, so Exim logs a panic warning message (but
3772 However, if deliver_drop_privilege is set, the unprivileged state may well
3773 be exactly what is intended, so the warning has been cut out in that case,
3774 and Exim is allowed to try to write to its normal log files.
3780 1. smtp_return_error_details was not giving details for temporary sender
3781 or receiver verification errors.
3783 2. Diagnose a configuration error if two authenticators have the same public
3786 3. Exim used not to create the message log file for a message until the first
3787 delivery attempt. This could be confusing when incoming messages were held
3788 for policy or load reasons. The message log file is now created at the time
3789 the message is received, and an initial "Received" line is written to it.
3791 4. The automatically generated man page for command line options had a minor
3792 bug that caused no ill effects; however, a more serious problem was that
3793 the procedure for building the man page automatically didn't always
3794 operate. Consequently, release 4.20 contains an out-of-date version. This
3795 shouldn't happen again.
3797 5. When building Exim with embedded Perl support, the script that builds the
3798 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3799 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3801 6. The freeze_tell option was not being used for messages that were frozen on
3802 arrival, either by an ACL or by local_scan().
3804 7. Added the smtp_incomplete_transaction log selector.
3806 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3807 was accepting AUTH without a new EHLO.
3809 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3810 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3813 10. Logging of TCP/IP connections (when configured) now happens in the main
3814 daemon process instead of the child process, so that the TCP/IP connection
3815 count is more accurate (but it can never be perfect).
3817 11. The use of "drop" in a nested ACL was not being handled correctly in the
3818 outer ACL. Now, if condition failure induced by the nested "drop" causes
3819 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3820 or "require"), the connection is dropped.
3822 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3823 that yield "discard" can now be used with an "accept" or a "discard" verb,
3824 but an error is generated for any others (because I can't see a useful way
3825 to define what should happen).
3827 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3828 are now processed in the same way as lines in the Exim configuration file.
3829 In particular, continuation lines are supported.
3831 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3833 15. Added -ti meaning -t -i.
3835 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3836 domains, and warn by logging if others are found.
3838 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3839 returned from GIFCONF when Exim is trying to find the list of interfaces on
3840 a host. The code in os.c has been modified to copy each ifreq to an aligned
3841 structure in all cases.
3843 Also, in some cases, the returned ifreq's were being copied to a 'struct
3844 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3845 means the last couple of bytes of an IPv6 address could be chopped if the
3846 ifreq contained only a normal sockaddr (14 bytes storage).
3848 18. Named domain lists were not supported in the hosts_treat_as_local option.
3849 An entry such as +xxxx was not recognized, and was treated as a literal
3852 19. Ensure that header lines added by a DATA ACL are included in the reject log
3853 if the ACL subsequently rejects the message.
3855 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3856 MD5 (which is deprecated).
3858 21. When testing a filter file using -bf, Exim was writing a message when it
3859 took the sender from a "From " line in the message, but it was not doing so
3860 when it took $return_path from a Return-Path: header line. It now does.
3862 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3863 with a valid header line field name (a series of printing characters
3864 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3866 23. Changed "disc" in the source to "disk" to conform to the documentation and
3867 the book and for uniformity.
3869 24. Ignore Sendmail's -Ooption=value command line item.
3871 25. When execve() failed while trying to run a command in a pipe transport,
3872 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3873 could be confused with a return value of 69 from the command itself. This
3874 has been changed to 127, the value the shell returns if it is asked to run
3875 a non-existent command. The wording for the related log line suggests a
3876 non-existent command as the problem.
3878 26. If received_header_text expands to an empty string, do not add a Received:
3879 header line to the message. (Well, it adds a token one on the spool, but
3880 marks it "old" so that it doesn't get used or transmitted.)
3882 27. Installed eximstats 1.28 (addition of -nt option).
3884 28. There was no check for failure on the call to getsockname() in the daemon
3885 code. This can fail if there is a shortage of resources on the system, with
3886 ENOMEM, for example. A temporary error is now given on failure.
3888 29. Contrary to the C standard, it seems that in some environments, the
3889 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3890 program. Exim now does this explicitly; it affects the formatting of
3891 timestamps using strftime().
3893 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3894 complaints. I've now initialized all the variables, to avoid this.
3896 32. Header lines added by a system filter were not being "seen" during
3897 transport-time rewrites.
3899 33. The info_callback() function passed to OpenSSL is set up with type void
3900 (*)(SSL *, int, int), as described somewhere. However, when calling the
3901 function (actually a macro) that sets it up, the type void(*)() is
3902 expected. I've put in a cast to prevent warnings from picky compilers.
3904 34. If a DNS black list lookup found a CNAME record, but there were no A
3905 records associated with the domain it pointed at, Exim crashed.
3907 35. If a DNS black list lookup returned more than one A record, Exim ignored
3908 all but the first. It now scans all returned addresses if a particular IP
3909 value is being sought. In this situation, the contents of the
3910 $dnslist_value variable are a list of all the addresses, separated by a
3913 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3914 to accept a host name and all its aliases if the forward lookup for any of
3915 them yielded the IP address of the incoming connection. Now it accepts only
3916 those names whose forward lookup yields the correct IP address. Any other
3917 names are discarded. This closes a loophole whereby a rogue DNS
3918 administrator could create reverse DNS records to break through a
3919 wildcarded host restriction in an ACL.
3921 37. If a user filter or a system filter that ran in a subprocess used any of
3922 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3923 the wrong values were passed to the pipe command ($thisaddress had the
3924 value of $0, $0 had the value of $1, etc). This bug was introduced by
3925 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3927 38. Improved the line breaking for long SMTP error messages from ACLs.
3928 Previously, if there was no break point between 40 and 75 characters, Exim
3929 left the rest of the message alone. Two changes have been made: (a) I've
3930 reduced the minimum length to 35 characters; (b) if it can't find a break
3931 point between 35 and 75 characters, it looks ahead and uses the first one
3932 that it finds. This may give the occasional overlong line, but at least the
3933 remaining text gets split now.
3935 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3936 file descriptors might be low, and that setting 1000 would always raise it.
3937 It turns out that in some environments, the limit is already over 1000 and
3938 that lowering it causes trouble. So now Exim takes care not to decrease it.
3940 40. When delivering a message, the value of $return_path is set to $sender_
3941 address at the start of routing (routers may change the value). By an
3942 oversight, this default was not being set up when an address was tested by
3943 -bt or -bv, which affected the outcome if any router or filter referred to
3946 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3947 log only when "message" or "log_message" are set. However, if one of the
3948 conditions was an address verification, or a call to a nested ACL, the
3949 messages generated by the underlying test were being passed through. This
3950 no longer happens. The underlying message is available in $acl_verify_
3951 message for both "message" and "log_message" expansions, so it can be
3952 passed through if needed.
3954 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3955 new expansion $bh_ to give the encoded byte string without charset
3956 translation. Translation happens only if iconv() is available; HAVE_ICONV
3957 indicates this at build time. HEADERS_CHARSET gives the charset to
3958 translate to; headers_charset can change it in the configuration, and
3959 "headers charset" can change it in an individual filter file.
3961 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3962 that creates RFC 2047 encoded "words" labels them as that charset instead
3963 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3964 expansion operator; (ii) when Exim creates a From: line for a local
3965 message; (iii) when a header line is rewritten to include a "phrase" part.
3967 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3968 buggy, causing it to skip the first lines of messages whose message ID
3969 ended in 'D'. This would not have bitten before Exim release 4.14, because
3970 message IDs were unlikely to end in 'D' before then. The effect was to have
3971 incorrect size information for certain domains.
3973 45. #include "config.h" was missing at the start of the crypt16.c module. This
3974 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3977 46. If there was a timeout during a "random" callout check, Exim treated it as
3978 a failure of the random address, and carried on sending RSET and the real
3979 address. If the delay was just some slowness somewhere, the response to the
3980 original RCPT would be taken as a response to RSET and so on, causing
3981 mayhem of various kinds.
3983 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3984 when I implemented it. It didn't allow for the fact that some option values
3985 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3986 the right test for positive values.
3988 48. Domain names in DNS records are case-independent. Exim always looks them up
3989 in lower case. Some resolvers return domain names in exactly the case they
3990 appear in the zone file, that is, they may contain uppercase letters. Not
3991 all resolvers do this - some return always lower case. Exim was treating a
3992 change of case by a resolver as a change of domain, similar to a widening
3993 of a domain abbreviation. This triggered its re-routing code and so it was
3994 trying to route what was effectively the same domain again. This normally
3995 caused routing to fail (because the router wouldn't handle the domain
3996 twice). Now Exim checks for this case specially, and just changes the
3997 casing of the domain that it ultimately uses when it transmits the message
4000 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4003 50. If a filter generated a file delivery with a non-absolute name (possible if
4004 no home directory exists for the router), the forbid_file option was not
4007 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4008 the existing equality matching.
4010 52. Exim was using ints instead of ino_t variables in some places where it was
4011 dealing with inode numbers.
4013 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4014 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4015 called TMPDIR, and if it finds it is different, it changes its value.
4017 54. The smtp_printf() function is now made available to local_scan() so
4018 additional output lines can be written before returning. There is also an
4019 smtp_fflush() function to enable the detection of a dropped connection.
4020 The variables smtp_input and smtp_batched_input are exported to
4023 55. Changed the default runtime configuration: the message "Unknown user"
4024 has been removed from the ACL, and instead placed on the localuser router,
4025 using the cannot_route_message feature. This means that any verification
4026 failures that generate their own messages won't get overridden. Similarly,
4027 the "Unrouteable address" message that was in the ACL for unverifiable
4028 relay addresses has also been removed.
4030 56. Added hosts_avoid_esmtp to the smtp transport.
4032 57. The exicyclog script was not checking for the esoteric option
4033 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4034 will work only if exicyclog is run under the appropriate euid.
4036 58. Following a discussion on the list, the rules by which Exim recognises line
4037 endings on incoming messages have been changed. The -dropcr and drop_cr
4038 options are now no-ops, retained only for backwards compatibility. The
4039 following line terminators are recognized: LF CRLF CR. However, special
4040 processing applies to CR:
4042 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4043 nor a local message in the state where . is a terminator.
4045 (ii) If a bare CR is encountered in a header line, an extra space is added
4046 after the line terminator so as not to end the header. The reasoning
4047 behind this is that bare CRs in header lines are most likely either
4048 to be mistakes, or people trying to play silly games.
4050 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4051 was being incorrectly given as 18 bytes larger than it should have been.
4052 This is a VOB (very old bug).
4054 60. This may never have affected anything current, but just in case it has:
4055 When the local host is found other than at the start of a list of hosts,
4056 the local host, those with the same MX, and any that follow, are discarded.
4057 When the list in question was part of a longer list of hosts, the following
4058 hosts (not currently being processed) were also being discarded. This no
4059 longer happens. I'm not sure if this situation could ever has previously
4062 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4065 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4066 header. This is something that is recommended in a new Internet Draft, and
4067 is something that is documented as being done by Sendmail. There are two
4068 possible values. For messages generated by the autoreply transport, Exim
4071 Auto-Submitted: auto-replied
4073 whereas for all other generated messages (e.g. bounces) it adds
4075 Auto-Submitted: auto-generated
4077 63. The "personal" condition in filters now includes a test for the
4078 Auto-Submitted: header. If it contains the string "auto-" the message it
4079 not considered personal.
4081 64. Added rcpt_include_affixes as a generic transport option.
4083 65. Added queue_only_override (default true).
4085 66. Added the syslog_duplication option.
4087 67. If what should have been the first header line of a message consisted of
4088 a space followed by a colon, Exim was mis-interpreting it as a header line.
4089 It isn't of course - it is syntactically invalid and should therefore be
4090 treated as the start of the message body. The misbehaviour could have
4091 caused a number of strange effects, including loss of data in subsequent
4092 header lines, and spool format errors.
4094 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4095 client host had authenticated. This control can now be exercised by an ACL
4096 for more flexibility.
4098 69. By default, callouts do not happen when testing with -bh. There is now a
4099 variant, -bhc, which does actually run the callout code, including
4100 consulting and updating the callout cache.
4102 70. Added support for saslauthd authentication, courtesy of Alexander
4105 71. If statvfs() failed on the spool or log directories while checking their
4106 size for availability, Exim confusingly gave the error "space shortage".
4107 Furthermore, in debugging mode it crashed with a floating point exception.
4108 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4109 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4110 serious problem, Exim now writes to the main and panic logs when this
4111 happens, with details of the failure. It then refuses to accept the
4112 incoming message, giving the message "spool directory problem" or "log
4113 directory problem" with a 421 code for SMTP messages.
4115 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4116 0, 1, and 2 exist, because some OS complain for execs without them (see
4117 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4118 descriptors. However, the code omitted to check that the open succeeded,
4119 causing mysterious errors if for some reason the permissions on /dev/null
4120 got screwed. Now Exim writes a message to the main and panic logs, and
4121 bombs out if it can't open /dev/null.
4123 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4124 interact so that it is all more flexible. It is supposed to remain
4125 backwards compatible. Also added extra_local_interfaces.
4127 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4128 to bomb out with an assertion failure - to the client this appears as a
4129 connection drop. This problem occurs in the part of the code that was taken
4130 from the Samba project. Fortunately, the assertion is in a very simple
4131 function, so I have fixed this by reproducing the function inline in the
4132 one place where it is called, and arranging for authentication to fail
4133 instead of killing the process with assert().
4135 75. The SPA client code was not working when the server requested OEM rather
4136 than Unicode encoding.
4138 76. Added code to make require_files with a specific uid setting more usable in
4139 the case where statting the file as root fails - usually a non-root-mounted
4140 NFS file system. When this happens and the failure is EACCES, Exim now
4141 forks a subprocess and does the per-uid checking as the relevant uid.
4143 77. Added process_log_path.
4145 78. If log_file_path was not explicitly set, a setting of check_log_space or
4146 check_log_inodes was ignored.
4148 79. If a space check for the spool or log partitions fails, the incident is now
4149 logged. Of course, in the latter case the data may get lost...
4151 80. Added the %p formatting code to string_format() so that it can be used to
4152 print addresses in debug_print(). Adjusted all the address printing in the
4153 debugging in store.c to use %p rather than %d.
4155 81. There was a concern that a line of code in smtp_in.c could overflow a
4156 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4157 initially expressed, the concern was not well-founded, because trailing
4158 spaces are removed early. However, if the trailing spaces were followed by
4159 a NULL, they did not get removed, so the overflow was possible. Two fixes
4162 (a) I re-wrote the offending code in a cleaner fashion.
4163 (b) If an incoming SMTP command contains a NULL character, it is rejected
4166 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4167 runs initgroups(), so that if the Exim user is in any additional groups,
4168 they will be used during message reception.
4174 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob