1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.396 2006/09/22 08:41:59 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 TF/01 In the add_headers option to the mail command in an Exim filter,
10 there was a bug that Exim would claim a syntax error in any
11 header after the first one which had an odd number of characters
14 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
15 callout verification, Exim cached a "reject" for the entire domain. This
16 is correct for most verifications, but it is not correct for a recipient
17 verification with use_sender or use_postmaster set, because in that case
18 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
19 case of MAIL FROM:<> rejection from other early rejections (e.g.
20 rejection of HELO). When verifying a recipient using a non-null MAIL
21 address, the cache is ignored if it shows MAIL FROM:<> rejection.
22 Whatever the result of the callout, the value of the domain cache is
23 left unchanged (for any other kind of callout, getting as far as trying
24 RCPT means that the domain itself is ok).
26 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
29 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
30 manifest itself as EPIPE rather than ECONNECT. When tidying away a
31 session, the daemon ignores ECONNECT errors and logs others; it now
32 ignores EPIPE as well.
34 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
35 (quoted-printable decoding).
37 PH/05 Applied Nico Efrurth's refactoring patch to tidy up spool_mbox.c, and
38 later the small subsequent patch to fix an introduced bug.
40 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
42 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
44 PH/08 An error is now given if message_size_limit is specified negative.
46 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
47 to be given (somewhat) arbitrary names.
49 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
52 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
53 miscellaneous code fixes
59 SC/01 Use a glob alias rather than an array ref in eximstats generated
60 parser. This improves both readability and performance.
62 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
63 Don't display local sender or destination tables in eximstats unless
64 there is data to show.
65 Added average volumes into the eximstats top table text output.
67 SC/03 Collect data on the number of addresses (recipients) as well
68 as the number of messages in eximstats.
70 TF/01 Correct an error in the documentation for the redirect router. Exim
71 does (usually) call initgroups() when daemonizing.
73 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
74 with consistent privilege compared to when running as a daemon.
76 TF/03 Note in the spec that $authenticated_id is not set for local
77 submissions from trusted users.
79 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
80 Thanks to Dean Brooks <dean@iglou.com> for the patch.
82 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
83 by adding some example configuration directives to the default
84 configuration file. A little bit of work is required to uncomment the
85 directives and define how usernames and passwords are checked, but
86 there is now a framework to start from.
88 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
89 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
90 without this. I don't know how relevant this is to other LDAP libraries.
92 PH/02 Add the verb name to the "unknown ACL verb" error.
94 PH/03 Magnus Holmgren's patch for filter_prepend_home.
96 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
98 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
99 directory not expanded when it should be if an expanded home directory
100 was set for the address (which is overridden by the transport).
102 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
105 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
106 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
107 because it is too late at that time, and has no effect.
109 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
110 security issue with \' (bugzilla #107). I could not use the
111 PQescapeStringConn() function, because it needs a PGconn value as one of
114 PH/08 When testing addresses using -bt, indicate those final addresses that
115 are duplicates that would not cause an additional delivery. At least one
116 person was confused, thinking that -bt output corresponded to deliveries.
117 (Suppressing duplicates isn't a good idea as you lose the information
118 about possibly different redirections that led to the duplicates.)
120 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
121 systems where poll() doesn't work, in particular OS X.
123 PH/10 Added more information to debugging output for retry time not reached.
125 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
126 operations in malware.c.
128 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
131 PH/13 If write_rejectlog was set false when logging was sent to syslog with
132 syslog_duplication set false, log lines that would normally be written
133 both the the main log and to the reject log were not written to syslog at
136 PH/14 In the default configuration, change the use of "message" in ACL warn
137 statements to "add_header".
139 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
140 not followed by a command (e.g. "seen endif").
142 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
143 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
146 PH/17 Added extra conditions to the default value of delay_warning_condition
150 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
151 { match{$h_precedence:}{(?i)bulk|list|junk} } \
152 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
155 The Auto-Submitted: and various List- headers are standardised, whereas I
156 don't think Precedence: ever was.
158 PH/18 Refactored debugging code in route_finduser() to show more information,
159 in particular, the error code if getpwnam() issues one.
161 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
162 This is apparently needed in addition to the PH/07 change above to avoid
163 any possible encoding problems.
165 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
166 but not after initializing Perl.
168 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
169 output them only if debugging. By default they are written stderr,
170 apparently, which is not desirable.
172 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
175 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
178 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
180 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
181 authenticated or an ident call has been made. Suppress the default
182 values for $authenticated_id and $authenticated_sender (but permit -oMai
183 and -oMas) when testing with -bh.
185 PH/24 Re-jigged the order of the tests in the default configuration so that the
186 tests for valid domains and recipients precede the DNS black list and CSA
187 tests, on the grounds that those ones are more expensive.
189 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
190 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
191 This bug exists in every version of Exim that I still have, right back to
194 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
195 However, an attempt to turn on case-sensitivity in a regex key by
196 including (?-i) didn't work because the subject string was already
197 lowercased, and the effects were non-intuitive. It turns out that a
198 one-line patch can be used to allow (?-i) to work as expected.
204 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
205 other effects) broke the use of negated acl sub-conditions.
207 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
210 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
211 "Deny" causes Exim to reject the incoming connection with a 554 error.
212 Unfortunately, if there is a major crisis, such as a disk failure,
213 tcp-wrappers gives "deny", whereas what one would like would be some
214 kind of temporary error. A kludge has been added to help with this.
215 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
216 554 error is used if errno is still zero or contains ENOENT (which occurs
217 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
220 PH/03 Add -lutil to the default FreeBSD LIBS setting.
222 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
223 errors. Otherwise a message that provokes a temporary error (when other
224 messages do not) can cause a whole host to time out.
226 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
227 addresses were routed directly to files or pipes from a redirect router.
228 File deliveries just didn't batch; pipe deliveries might have suffered
231 PH/06 A failure to get a lock for a hints database would erroneously always say
232 "Failed to get write lock", even when it was really a read lock.
234 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
235 of 0600. This has been changed to use the value of the lockfile_mode
236 option (which defaults to 0600).
238 PH/08 Applied small patch from the Sieve maintainer.
240 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
241 folder from quota calculations, a direct delivery into this folder messed
242 up the contents of the maildirsize file. This was because the regex was
243 used only to exclude .Trash (or whatever) when the size of the mailbox
244 was calculated. There was no check that a delivery was happening into an
245 excluded directory. This bug has been fixed by ignoring all quota
246 processing for deliveries into excluded directories.
248 PH/10 Added the maildirfolder_create_regex option to appendfile.
254 PH/01 The code for finding all the local interface addresses on a FreeBSD
255 system running IPv6 was broken. This may well have applied to all BSD
256 systems, as well as to others that have similar system calls. The broken
257 code found IPv4 interfaces correctly, but gave incorrect values for the
258 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
259 that it would not match correctly against @[] and not recognize the IPv6
262 PH/02 The ipliteral router was not recognizing addresses of the form user@
263 [ipv6:....] because it didn't know about the "ipv6:" prefix.
265 PH/03 Added disable_ipv6.
267 PH/04 Changed $reply_address to use the raw form of the headers instead of the
268 decoded form, because it is most often used to construct To: headers
269 lines in autoreplies, and the decoded form may well be syntactically
270 invalid. However, $reply_address has leading white space removed, and all
271 newlines turned into spaces so that the autoreply transport does not
274 PH/05 If group was specified without a user on a router, and no group or user
275 was specified on a transport, the group from the router was ignored.
277 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
278 for visible compile-time settings that can be used to change these
279 numbers, for those that want even more. Backwards compatibility with old
280 spool files has been maintained. However, going back to a previous Exim
281 release will lost any variables that are in spool files.
283 PH/07 Two small changes when running in the test harness: increase delay when
284 passing a TCP/IP connection to a new process, in case the original
285 process has to generate a bounce, and remove special handling of
286 127.0.0.2 (sic), which is no longer necessary.
288 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
289 be the same on different OS.
291 PH/09 Moved a debug statement in filter processing to avoid a race problem when
294 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
295 whether --show-vars was specified or not
297 JJ/02 exipick: Added support for new ACL variable spool format introduced
300 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
301 syntactically invalid From: or Reply-to: line, and a filter used this to
302 generate an autoreply, and therefore failed to obtain an address for the
303 autoreply, Exim could try to deliver to a non-existent relative file
304 name, causing unrelated and misleading errors. What now happens is that
305 it logs this as a hard delivery error, but does not attempt to create a
308 PH/11 The exinext utility has a -C option for testing purposes, but although
309 the given file was scanned by exinext itself; it wasn't being passed on
310 when Exim was called.
312 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
313 an end-of-file indication when reading a command response.
315 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
316 compiled. In many other places in Exim, IPv6 addresses are always
317 recognized, so I have changed this. It also means that IPv4 domain
318 literals of the form [IPV4:n.n.n.n] are now always recognized.
320 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
321 used if the router is not running as root, for example, when verifying at
322 ACL time, or when using -bh. The debugging output from this situation was
323 non-existent - all you got was a failure to exec. I have made two
326 (a) Failures to set uid/gid, the current directory, or a process leader
327 in a subprocess such as that created by queryprogram now generate
328 suitable debugging ouput when -d is set.
330 (b) The queryprogram router detects when it is not running as root,
331 outputs suitable debugging information if -d is set, and then runs
332 the subprocess without attempting to change uid/gid.
334 PH/15 Minor change to Makefile for building test_host (undocumented testing
337 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
338 additional section of a DNS packet that returns MX or SRV records.
339 Instead, it always explicitly searches for A/AAAA records. This avoids
340 major problems that occur when a DNS server includes only records of one
341 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
342 fixed another bug: if SRV records were looked up and the corresponding
343 address records were *not* found in the additional section, the port
344 values from the SRV records were lost.
346 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
347 using the correct key (the original address) when searching the retry
348 rules in order to find which one to use for generating the retry hint.
350 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
351 adding the default one. Similarly, if it contains a Reply-To: header, the
352 errors_reply_to option, if set, is not used.
354 PH/19 When calculating a retry time, Exim used to measure the "time since
355 failure" by looking at the "first failed" field in the retry record. Now
356 it does not use this if it is later than than the arrival time of the
357 message. Instead it uses the arrival time. This makes for better
358 behaviour in cases where some deliveries succeed, thus re-setting the
359 "first failed" field. An example is a quota failure for a huge message
360 when small messages continue to be delivered. Without this change, the
361 "time since failure" will always be short, possible causing more frequent
362 delivery attempts for the huge message than are intended.
363 [Note: This change was subsequently modified - see PH/04 for 4.62.]
365 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
366 $1, $2, $3) because the numerical variables can be reset during some
367 expansion items (e.g. "match"), thereby losing the authentication data.
369 PH/21 Make -bV show the size of off_t variables so that the test suite can
370 decide whether to run tests for quotas > 2G.
372 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
373 mailbox_size, and mailbox_filecount in the appendfile transport. If a
374 filecount value is greater than 2G or if a quota value is greater than 2G
375 on a system where the size of off_t is not greater than 4, a panic error
378 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
379 never match. The debug and -bh output now contains an explicit error
380 message indicating a malformed IPv4 address or mask.
382 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
383 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
386 PH/25 Do not write to syslog when running in the test harness. The only
387 occasion when this arises is a failure to open the main or panic logs
388 (for which there is an explicit test).
390 PH/26 Added the /no_tell option to "control=freeze".
392 PH/27 If a host name lookup failed very early in a connection, for example, if
393 the IP address matched host_lookup and the reverse lookup yielded a name
394 that did not have a forward lookup, an error message of the form "no IP
395 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
396 could be logged. Now it outputs the IP address instead of "NULL".
398 PH/28 An enabling patch from MH: add new function child_open_exim2() which
399 allows the sender and the authenticated sender to be set when
400 submitting a message from within Exim. Since child_open_exim() is
401 documented for local_scan(), the new function should be too.
403 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
404 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
405 results in an empty string is now treated as unset.
407 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
409 PH/31 Added sender_verify_fail logging option.
411 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
412 needed by RFC 822 but not by RFC 2822 was commented out. I have now
413 tidied the source and removed it altogether.
415 PH/33 When a queue run was abandoned because the load average was too high, a
416 log line was always written; now it is written only if the queue_run log
417 selector is set. In addition, the log line for abandonment now contains
418 information about the queue run such as the pid. This is always present
419 in "start" and "stop" lines but was omitted from the "abandon" line.
421 PH/34 Omit spaces between a header name and the colon in the error message that
422 is given when verify = headers_syntax fails (if there are lots of them,
423 the message gets confusing).
425 PH/35 Change the default for dns_check_names_pattern to allow slashes within
426 names, as there are now some PTR records that contain slashes. This check
427 is only to protect against broken name servers that fall over on strange
428 characters, so the fact that it applies to all lookups doesn't matter.
430 PH/36 Now that the new test suite is complete, we can remove some of the
431 special code in Exim that was needed for the old test suite. For example,
432 sorting DNS records because real resolvers return them in an arbitrary
433 order. The new test suite's fake resolver always returns records in the
436 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
437 bounces) except when queue_only is set, to avoid logging races between
438 the different processes.
440 PH/38 Panic-die if .include specifies a non-absolute path.
442 PH/39 A tweak to the "H" retry rule from its user.
444 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
445 a label. They prevented compilation on older perls.
447 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
448 a warning to be raised on newish perls.
450 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
451 on queue. Changes to match documented behaviour of showing count of
452 messages matching specified criteria.
454 PH/40 Changed the default ident timeout from 30s to 5s.
456 PH/41 Added support for the use of login_cap features, on those BSD systems
457 that have them, for controlling the resources used by pipe deliveries.
459 PH/42 The content-scanning code uses fopen() to create files in which to put
460 message data. Previously it was not paying any attention to the mode of
461 the files. Exim runs with umask(0) because the rest of the code creates
462 files with open(), and sets the required mode explicitly. Thus, these
463 files were ending up world-writeable. This was not a big issue, because,
464 being within the spool directory, they were not world-accessible. I have
465 created a function called modefopen, which takes an additional mode
466 argument. It sets umask(777), creates the file, chmods it to the required
467 mode, then resets the umask. All the relevant calls to fopen() in the
468 content scanning code have been changed to use this function.
470 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
471 to 24 hours. This avoids potential overflow problems when processing G
472 and H retry rules. I suspect nobody ever tinkers with this value.
474 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
476 PH/45 When the plaintext authenticator is running as a client, the server's
477 challenges are checked to ensure they are valid base64 strings. By
478 default, the authentication attempt is cancelled if an invalid string is
479 received. Setting client_ignore_invalid_base64 true ignores these errors.
480 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
481 they are received. Thus, the responses can be made to depend on the
482 challenges. If an invalid string is ignored, an empty string is placed in
485 PH/46 Messages that are created by the autoreply transport now contains a
486 References: header, in accordance with RFCs 2822 and 3834.
488 PH/47 Added authenticated_sender_force to the smtp transport.
490 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
492 PH/49 Installed latest patch from the Sieve maintainer.
494 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
495 was also set, the appendfile transport was unnecessarily scanning a
496 directory of message files (e.g. for maildir delivery) to find the count
497 of files (along with the size), even though it did not need this
498 information. It now does the scan only if it needs to find either the
499 size of the count of files.
501 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
503 PH/52 Two bugs concerned with error handling when the smtp transport is
506 (i) Exim was not creating retry information for temporary errors given
507 for individual recipients after the DATA command when the smtp transport
508 was used in LMTP mode. This meant that they could be retried too
509 frequently, and not timed out correctly.
511 (ii) Exim was setting the flag that allows error details to be returned
512 for LMTP errors on RCPT commands, but not for LMTP errors for individual
513 recipients that were returned after the DATA command.
515 PH/53 This is related to PH/52, but is more general: for any failing address,
516 when detailed error information was permitted to be returned to the
517 sender, but the error was temporary, then after the final timeout, only
518 "retry timeout exceeded" was returned. Now it returns the full error as
519 well as "retry timeout exceeded".
521 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
522 do this, and (what is worse) MTAs that accept it.
524 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
525 will now be deprecated.
527 PH/56 New os.c-cygwin from the Cygwin maintainer.
529 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
530 formats (previously only available in exim formats via -bpr, -bpru,
531 and -bpra. Now also available in native and exiqgrep formats)
533 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
534 with very large, slow to parse queues
536 JJ/08 exipick: added ! as generic prefix to negate any criteria format
538 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
540 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
541 responses to authentication challenges, though it was showing the
542 challenges; (ii) I've removed the CR characters from the debug output for
545 PH/58 Allow for the insertion of a newline as well as a space when a string
546 is turned into more than one encoded-word during RFC 2047 encoding. The
547 Sieve code now uses this.
549 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
550 data_4xx, lost_connection, tls_required.
552 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
553 message was being sent as an SMTP response.
555 PH/61 Add -l and -k options to exicyclog.
557 PH/62 When verifying, if an address was redirected to one new address, so that
558 verification continued, and the new address failed or deferred after
559 having set something in $address_data, the value of $address_data was not
560 passed back to the ACL. This was different to the case when no
561 redirection occurred. The value is now passed back in both cases.
563 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
564 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
565 use login_cap.h, so on its own it isn't the distinguishing feature. The
566 new name refers directly to the setclassresources() function.
568 PH/65 Added configuration files for NetBSD3.
570 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
572 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
573 is preferred over IPv4.
575 PH/68 The bounce_return_message and bounce_return_body options were not being
576 honoured for bounces generated during the reception of non-SMTP messages.
577 In particular, this applied to messages rejected by the ACL. This bug has
578 been fixed. However, if bounce_return_message is true and bounce_return_
579 body is false, the headers that are returned for a non-SMTP message
580 include only those that have been read before the error was detected.
581 (In the case of an ACL rejection, they have all been read.)
583 PH/69 The HTML version of the specification is now built in a directory called
584 spec_html instead of spec.html, because the latter looks like a path with
585 a MIME-type, and this confuses some software.
587 PH/70 Catch two compiler warnings in sieve.c.
589 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
590 function verify_get_ident() calls ip_connect() to connect a socket, but
591 if the "connect()" function timed out, ip_connect() used to close the
592 socket. However, verify_get_ident() also closes the socket later, and in
593 between Exim writes to the log, which may get opened at this point. When
594 the socket was closed in ip_connect(), the log could get the same file
595 descriptor number as the socket. This naturally causes chaos. The fix is
596 not to close the socket in ip_connect(); the socket should be closed by
597 the function that creates it. There was only one place in the code where
598 this was missing, in the iplookup router, which I don't think anybody now
599 uses, but I've fixed it anyway.
601 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
602 well as to direct DNS lookups. Otherwise the handling of names in host
603 lists is inconsistent and therefore confusing.
609 PH/01 Two changes to the default runtime configuration:
611 (1) Move the checks for relay_from_hosts and authenticated clients from
612 after to before the (commented out) DNS black list checks.
614 (2) Add control=submission to the relay_from_hosts and authenticated
615 clients checks, on the grounds that messages accepted by these
616 statements are most likely to be submissions.
618 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
620 (1) Generate an error if the third argument for the ${prvs expansion is
623 (2) Treat a missing third argument of ${prvscheck as if it were an empty
626 (3) Reset the variables that are obtained from the first argument of
627 ${prvscheck and used in the second argument before leaving the code,
628 because their memory is reclaimed, so using them afterwards may do
631 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
632 one (it's much easier than Tom thought :-).
634 (5) Because of (4), we can now allow for the use of $prvscheck_result
635 inside the third argument.
637 PH/03 For some reason, the default setting of PATH when running a command from
638 a pipe transport was just "/usr/bin". I have changed it to
641 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
642 anything to be listed in the output from -bV.
644 PH/05 When a filter generated an autoreply, the entire To: header line was
645 quoted in the delivery log line, like this:
647 => >A.N.Other <ano@some.domain> <original@ddress> ...
649 This has been changed so that it extracts the operative address. There
650 may be more than one such address. If so, they are comma-separated, like
653 => >ano@some.domain,ona@other.domain <original@ddress> ...
655 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
656 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
657 not being looked up in the rDNS to get a host name, Exim was showing the
658 IP address twice in Received: lines, even though the IP addresses were
659 identical. For example:
661 Received: from [1.2.3.4] (helo=[1.2.3.4])
663 However, if the real host name was known, it was omitting the HELO data
664 if it matched the actual IP address. This has been tidied up so that it
665 doesn't show the same IP address twice.
667 PH/07 When both +timestamp and +memory debugging was on, the value given by
668 $tod_xxx expansions could be wrong, because the tod_stamp() function was
669 called by the debug printing, thereby overwriting the timestamp buffer.
670 Debugging no longer uses the tod_stamp() function when +timestamp is set.
672 PH/08 When the original message was included in an autoreply transport, it
673 always said "this is a copy of the message, including all the headers",
674 even if body_only or headers_only was set. It now gives an appropriate
677 PH/09 Applied a patch from the Sieve maintainer which:
679 o fixes some comments
680 o adds the (disabled) notify extension core
681 o adds some debug output for the result of if/elsif tests
682 o points to the current vacation draft in the documentation
683 and documents the missing references header update
687 o fixes a bug in processing the envelope test (when testing
688 multiple envelope elements, the last element determinted the
691 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
692 Electronic Mail") by including:
694 Auto-submitted: auto-generated
696 in the messages that it generates (bounce messages and others, such as
697 warnings). In the case of bounce messages for non-SMTP mesages, there was
698 also a typo: it was using "Auto_submitted" (underscore instead of
699 hyphen). Since every message generated by Exim is necessarily in response
700 to another message, thes have all been changed to:
702 Auto-Submitted: auto-replied
704 in accordance with these statements in the RFC:
706 The auto-replied keyword:
708 - SHOULD be used on messages sent in direct response to another
709 message by an automatic process,
711 - MUST NOT be used on manually-generated messages,
713 - MAY be used on Delivery Status Notifications (DSNs) and Message
714 Disposition Notifications (MDNs),
716 - MUST NOT be used on messages generated by automatic or periodic
717 processes, except for messages which are automatic responses to
720 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
721 to the default Received: header definition.
723 PH/12 Added log selector acl_warn_skipped (default on).
725 PH/13 After a successful wildlsearch lookup, discard the values of numeric
726 variables because (a) they are in the wrong storage pool and (b) even if
727 they were copied, it wouldn't work properly because of the caching.
729 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
730 checking when decoding. Apparently there are clients that generate
731 overlong encoded strings. Why am I not surprised?
733 PH/15 If the first argument of "${if match_address" was not empty, but did not
734 contain an "@" character, Exim crashed. Now it writes a panic log message
735 and treats the condition as false.
737 PH/16 In autoreply, treat an empty string for "once" the same as unset.
739 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
740 extension "envelope-auth". The code is finished and in agreement with
741 other implementations, but there is no documentation so far and in fact,
742 nobody wrote the draft yet. This extension is currently #undef'ed, thus
743 not changing the active code.
745 Print executed "if" and "elsif" statements when debugging is used. This
746 helps a great deal to understand what a filter does.
748 Document more things not specified clearly in RFC3028. I had all this
749 sorted out, when out of a sudden new issues came to my mind. Oops."
751 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
754 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
755 canonical form (as documented). However, after a host name lookup from
756 the IP address, check_host() was doing a simple string comparison with
757 addresses acquired from the DNS when checking that the found name did
758 have the original IP as one of its addresses. Since any found IPv6
759 addresses are likely to be in abbreviated form, the comparison could
760 fail. Luckily, there already exists a function for doing the comparison
761 by converting both addresses to binary, so now that is used instead of
764 PH/20 There was another similar case to PH/19, when a complete host name was
765 given in a host list; looking up its IP address could give an abbreviated
766 form, whereas the current host's name might or might not be abbreviated.
767 The same fix has been applied.
773 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
774 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
777 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
778 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
780 PH/03 Typo: missing ".o" in src/pcre/Makefile.
782 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
783 header line, restrict the check to what is listed in RFCs 2369 and 2929.
784 Also, for "Auto-Submitted", treat anything other than "no" as
785 non-personal, in accordance with RFC 3834. (Previously it treated
786 anything starting "auto-" as non-personal.)
788 TF/01 The control=submission/name=... option had a problem with syntax
789 errors if the name included a slash character. The /name= option
790 now slurps the rest of the string, so it can include any characters
791 but it must come last in the list of options (after /sender_retain
794 PH/05 Some modifications to the interface to the fake nameserver for the new
802 TK/01 Added the "success_on_redirect" address verification option. See
803 NewStuff for rationale and an example.
805 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
807 PH/02 Patch to exigrep to allow it to work on syslog lines.
809 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
810 fread() to skip over the body file's header line, because in Cygwin the
811 header line is locked and is inaccessible.
813 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
814 co-exist for some time) to make it clear that it is the Exim ID that is
815 referenced, not the Message-ID: header line.
817 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
818 string_format() function, because snprintf() does not exist on all
821 PH/06 The use of forbid_filter_existstest now also locks out the use of the
822 ${stat: expansion item.
824 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
825 protocol synchronization error", to keep the pedants happy.
827 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
828 well as for IRIX systems, when gcc is being used. See the host.c source
831 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
833 PH/10 Named domain lists were not working if used in a queue_smtp_domains
836 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
837 transport and to the smtp transport in LMTP mode.
839 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
841 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
842 run a filter in a subprocess. This could lead to confusion in subsequent
843 lookups in the parent process. There should also be a search_tidyup() at
844 the end of the subprocess.
846 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
847 only if the host matched helo_try_verify_hosts, which caused the
848 verification to occur when the EHLO/HELO command was issued. The ACL just
849 tested the remembered result. Now, if a previous verification attempt has
850 not happened, "verify = helo" does it there and then.
852 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
854 TK/03 Fix log output including CR from clamd.
856 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
857 exist provoked a memory error which could cause a segfault.
859 PH/15 Installed PCRE 6.2
861 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
863 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
864 of the problem. Specifically, suggested +O2 rather than +O1 for the
867 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
869 PH/20 If a delivery was routed to a non-standard port by means of an SRV
870 record, the port was not correctly logged when the outgoing_port log
871 selector was set (it logged the transort's default port).
873 PH/21 Added support for host-specific ports to manualroute, queryprogram,
874 fallback_hosts, and "hosts" in the smtp transport.
876 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
877 host errors such as "Connection refused".
879 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
880 authentication with radiusclient 0.4.9:
882 - Error returned from rc_read_config was caught wrongly
883 - Username/password not passed on to radius server due to wrong length.
885 The presumption is that some radiusclient API changes for 4.51/PH/17
886 were not taken care of correctly. The code is still untested by me (my
887 Linux distribution still has 0.3.2 of radiusclient), but it was
888 contributed by a Radius user.
890 PH/24 When doing a callout, the value of $domain wasn't set correctly when
891 expanding the "port" option of the smtp transport.
893 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
894 while reading a MIME header. Thanks to Tom Hughes for a patch.
896 PH/24 Include config.h inside local_scan.h so that configuration settings are
899 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
900 that in an ACL for RCPT (for example), you can examine exactly what was
903 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
904 commands, but it was not correctly comparing the address with the actual
905 client host address. Thus, it would show the EHLO address in Received:
906 header lines when this was not necessary.
908 PH/27 Added the % operator to ${eval:}.
910 PH/28 Exim tries to create and chdir to its spool directory when it starts;
911 it should be ignoring failures (because with -C, for example, it has lost
912 privilege). It wasn't ignoring creation failures other than "already
915 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
918 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
919 because an input line was too long, either on its own, or by virtue of
920 too many continuations, the temporary file was not being removed, and the
921 return code was incorrect.
923 PH/31 Missing "BOOL" in function definition in filtertest.c.
925 PH/32 Applied Sieve patches from the maintainer.
927 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
929 PH/33 Added "verify = not_blind".
931 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
932 Local/Makefile (with some defaults set). These are used in built scripts
933 such as exicyclog, but they have never been used in the exim_install
934 script (though there are many overriding facilities there). I have
935 arranged that the exim_install script now takes note of these two
938 PH/35 Installed configuration files for Dragonfly.
940 PH/36 When a locally submitted message by a trusted user did not contain a
941 From: header, and the sender address was obtained from -f or from an SMTP
942 MAIL command, and the trusted user did not use -F to supply a sender
943 name, $originator_name was incorrectly used when constructing a From:
944 header. Furthermore, $originator_name was used for submission mode
945 messages from external hosts without From: headers in a similar way,
946 which is clearly wrong.
948 PH/37 Added control=suppress_local_fixups.
950 PH/38 When log_selector = +received_sender was set, and the addition of the
951 sender made the log line's construction buffer exactly full, or one byte
952 less than full, an overflow happened when the terminating "\n" was
955 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
956 when the result of a list match is failure because a DNS lookup failed.
958 PH/40 RM_COMMAND is now used in the building process.
960 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
961 the "build-* directories that it finds.
963 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
964 address in a domain literal was a prefix of an interface address.
966 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
967 when verifying a sender address, unless rewrite_headers is false.
969 PH/44 Wrote a long comment about why errors_to addresses are verified as
970 recipients, not senders.
972 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
973 the ratelimit ACL was added.
975 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
977 PH/46 Added extra information about PostgreSQL errors to the error string.
979 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
980 avoiding the need to install special zones in a real server. This is
981 backwards compatible; if it can't find the fake resolver, it drops back.
982 Thus, both old and new test suites can be run.
984 TF/02 Added util/ratelimit.pl
986 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
987 clock is set back in time.
989 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
990 Candler <B.Candler@pobox.com>.
992 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
993 OK for addresses that are the result of redirections.
995 PH/48 A number of further additions for the benefit of the new test suite,
996 including a fake gethostbyname() that interfaces to the fake DNS resolver
999 TF/06 The fix for widen_domains has also been applied to qualify_single and
1000 search_parents which are the other dnslookup options that can cause
1003 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1006 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1008 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1009 service" if it encountered an MX record with an empty target hostname.
1010 The message is now "an MX or SRV record indicated no SMTP service".
1012 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1013 if the DNS of the sending site is misconfigured. This is quite a
1014 common situation. This change restores the behaviour of treating a
1015 helo verification defer as a failure.
1017 PH/51 If self=fail was set on a router, the bounce message did not include the
1018 actual error message.
1024 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1026 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1027 command itself ended in error, the underlying message about the transport
1028 filter timeout was being overwritten with the pipe command error. Now the
1029 underlying error message should be appended to the second error message.
1031 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1033 PH/02 Reduce the amount of output that "make" produces by default. Full output
1034 can still be requested.
1036 PH/03 The warning log line about a condition test deferring for a "warn" verb
1037 was being output only once per connection, rather than after each
1038 occurrence (because it was using the same function as for successful
1039 "warn" verbs). This seems wrong, so I have changed it.
1041 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1042 it should not have, which might have caused a crash in the right
1043 circumstances, but probably never did.
1045 PH/04 Installed a modified version of Tony Finch's patch to make submission
1046 mode fix the return path as well as the Sender: header line, and to
1047 add a /name= option so that you can make the user's friendly name appear
1050 TF/03 Added the control = fakedefer ACL modifier.
1052 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1053 Mark Lowes for thorough testing.
1055 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1057 TK/03 Merged latest SRS patch from Miles Wilton.
1059 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1060 with the definition in sysexits.h (which is #included earlier).
1061 Fortunately, Exim does not actually use EX_OK. The code used to try to
1062 preserve the sysexits.h value, by assumimg that macro definitions were
1063 scanned for macro replacements. I have been disabused of this notion,
1064 so now the code just undefines EX_OK before #including unistd.h.
1066 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1067 in the smtp transport. When a block could not be written in a single
1068 write() function, the timeout was being re-applied to each part-write.
1069 This seems wrong - if the receiver was accepting one byte at a time it
1070 would take for ever. The timeout is now adjusted when this happens. It
1071 doesn't have to be particularly precise.
1073 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1074 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1076 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1077 without a domain if the check to <postmaster@domain> fails.
1079 SC/01 Eximstats: added -xls and the ability to specify output files
1080 (patch written by Frank Heydlauf).
1082 SC/02 Eximstats: use FileHandles for outputing results.
1084 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1086 SC/04 Eximstats: fixed display of large numbers with -nvr option
1088 SC/05 Eximstats: fixed merging of reports with empty tables.
1090 SC/06 Eximstats: added the -include_original_destination flag
1092 SC/07 Eximstats: removed tabs and trailing whitespace.
1094 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1096 TK/06 MBOX spool code: Add real "From " MBOX separator line
1097 so the .eml file is really in mbox format (even though
1098 most programs do not really care). Patch from Alex Miller.
1100 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1101 The latter is generated from $received_to and is only set if the
1102 message has one envelope recipient. SA can use these headers,
1103 obviously out-of-the-box. Patch from Alex Miller.
1105 PH/08 The ${def test on a variable was returning false if the variable's
1106 value was "0", contrary to what the specification has always said!
1107 The result should be true unless the variable is empty.
1109 PH/09 The syntax error of a character other than { following "${if
1110 def:variable_name" (after optional whitespace) was not being diagnosed.
1111 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1112 accidental colon was present, for example, could give incorrect results.
1114 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1115 result is used (not including appendfile, where other changes are about
1118 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1119 This involved changing a lot of size variables from int to off_t. It
1120 should work with maildirs and everything.
1122 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1123 spamd dying while we are connected to it.
1125 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1128 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1129 function for BDB 4.3.
1131 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1133 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1134 the sources for PCRE have been re-arranged and more files are now
1137 PH/15 The code I had for printing potentially long long variables in PH/11
1138 above was not the best (it lost precision). The length of off_t variables
1139 is now inspected at build time, and an appropriate printing format (%ld
1140 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1141 to be "long long int" or "long int". This is needed for the internal
1142 formatting function string_vformat().
1144 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1145 the configuration file to be ":syslog", then the script "guesses" where
1146 the logs files are, rather than using the compiled in default. In our
1147 case the guess is not the same as the compiled default, so the script
1148 suddenly stopped working when I started to use syslog. The patch checks
1149 to see if log_file_path is "". If so, it attempts to read it from exim
1150 with no configuration file to get the compiled in version, before it
1151 falls back to the previous guessing code."
1153 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1154 implementing BATV in an Exim configuration. See NewStuff for the gory
1157 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1158 Makefile that are specific to HP-UX.
1160 PH/18 If the "use_postmaster" option was set for a recipient callout together
1161 with the "random" option, the postmaster address was used as the MAIL
1162 FROM address for the random test, but not for the subsequent recipient
1163 test. It is now used for both.
1165 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1166 patch removes a few documentation additions to RFC 3028, because the
1167 latest draft now contains them. It adds the new en;ascii-case comparator
1168 and a new error check for 8bit text in MIME parts. Comparator and
1169 require names are now matched exactly. I enabled the subaddress
1170 extension, but it is not well tested yet (read: it works for me)."
1172 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1173 rework some of the code of TK/09 above to avoid the hardwired use of
1174 "%lld" and "long long". Replaced the call to snprintf() with a call to
1177 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1178 records point to non-existent hosts", "retry timeout exceeded", and
1179 "retry time not reached for any host after a long failure period".
1181 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1182 experimental DomainKeys support:
1184 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1185 (2) On an error such as an illegally used "control", the wrong name for
1186 the control was given.
1188 These problems did NOT occur unless DomainKeys support was compiled.
1190 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1192 PH/24 Added ${if match_ip condition.
1194 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1195 databases so that it will be absolutely obvious if a crash occurs in the
1196 DB library. This is a regular occurrence (often caused by mis-matched
1199 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1200 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1201 picked up on a user's system that detects such things. There doesn't seem
1202 to be a gcc warning option for this - only an attribute that has to be
1203 put on the function's prototype. It seems that in Fedora Core 4 they have
1204 set this on a number of new functions. No doubt there will be more in due
1207 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1208 specify a transport. However, if an address that was verified by such a
1209 router was the subject of a callout, Exim crashed because it tried to
1210 read the rcpt_include_affixes from the non-existent transport. Now it
1211 just assumes that the setting of that option is false. This bug was
1212 introduced by 4.51/PH/31.
1214 PH/28 Changed -d+all to exclude +memory, because that information is very
1215 rarely of interest, but it makes the output a lot bigger. People tend to
1216 do -d+all out of habit.
1218 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1219 code in os-type was giving problems when libc.so lives in lib64, like on
1222 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1223 aren't the modern standard, and it seems that some systems' include files
1224 don't always have them. Exim was already checking for some of the newer
1225 ones like T_AAAA, and defining it itself. I've added checks for all the
1226 record types that Exim uses.
1228 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1229 not automatically generating a new one, as it is supposed to. This
1230 prevented TLS from working. If the file did exist, but contained invalid
1231 data, a new version was generated, as expected. It was only the case of a
1232 non-existent file that was broken.
1234 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1235 with a change in libdomainkeys > 0.64.
1237 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1238 from DNS. If the selector record carries the flag, it now has
1239 precedence over the domain-wide flag.
1241 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1243 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1244 the use of a port name that isn't defined in /etc/services) occurred, the
1245 message was deferred as in a normal delivery, and thus remained on the
1246 spool, instead of being failed because of the mua_wrapper setting. This
1247 is now fixed, and I tidied up some of the mua_wrapper messages at the
1250 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1251 the messages in a hash of arrays rather than using individual hashes.
1252 This is a bit cleaner and results in dramatic memory savings, albeit
1253 at a slight CPU cost.
1255 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1256 as requested by Marc Sherman.
1258 SC/10 Eximstats: added histograms for user specified patterns as requested
1261 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1263 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1264 fopen() in the content-scanning modules that did not already have it.
1270 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1271 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1273 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1275 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1277 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1278 to test Sieve filters that use "vacation".
1280 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1281 that changes the way the GnuTLS parameters are stored in the cache file.
1282 The new format can be generated externally. For backward compatibility,
1283 if the data in the cache doesn't make sense, Exim assumes it has read an
1284 old-format file, and it generates new data and writes a new file. This
1285 means that you can't go back to an older release without removing the
1288 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1289 work if there are any delivery delays because "one_time" forces the
1290 parent to be marked "delivered", so its unseen clone is never tried
1291 again. For this reason, Exim now forbids the simultaneous setting of
1294 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1295 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1296 entry if you want to know the details. The fix, however, neglected to
1297 consider the case when local delivery batching is involved. The test for
1298 "previously delivered" was not happening when checking to see if an
1299 address could be batched with a previous (undelivered) one; under
1300 certain circumstances this could lead to multiple deliveries to the same
1303 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1304 in its include files, and this causes problems building Exim.
1306 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1307 header_syntax) but Exim was just ignoring anything given after a slash.
1308 In particular, this caused confusion with an attempt to use "verify =
1309 reverse_host_lookup/defer_ok". An error is now given when options are
1310 supplied for verify items that do not have them. (Maybe reverse_host_
1311 lookup should have a defer_ok option, but that's a different point.)
1313 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1314 defined by RFC 821) to 2048, because there were problems with some AUTH
1315 commands, and RFC 1869 says the size should be increased for extended
1316 SMTP commands that take arguments.
1318 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1321 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1322 "unknown" error; now it says that the functionality isn't in the binary.
1324 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1325 an address' error message when a string expansion fails (syntax or
1326 whatever). Otherwise the password may appear in the log. Following change
1327 PH/42 below, there is no longer a chance of it appearing in a bounce
1330 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1332 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1333 of its addresses was ever tried. (Bugzilla bug #2.)
1335 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1336 the result incorrectly in the debug output. (It correctly added a newline
1337 to what was transported.)
1339 TF/01 Added $received_time.
1341 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1342 commented out examples of how to interface to a virus scanner and to
1343 SpamAssassin. Also added commented examples of av_scanner and
1344 spamd_address settings.
1346 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1347 and controls are allowed in which ACLs. There were a couple of minor
1348 errors. Some of the entries in the conditions table (which is a table of
1349 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1350 negation of where the condition IS allowed.
1352 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1354 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1355 header file does not have a version number, so I've had to invent a new
1356 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1357 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1358 radiusclient), but it was contributed by a Radius user.
1360 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1361 files or directories, for OpenSSL.
1363 PH/19 When an Exim process that is running as root has to create an Exim log
1364 file, it does so in a subprocess that runs as exim:exim so as to get the
1365 ownership right at creation (otherwise, other Exim processes might see
1366 the file with the wrong ownership). There was no test for failure of this
1367 fork() call, which would lead to the process getting stuck as it waited
1368 for a non-existent subprocess. Forks do occasionally fail when resources
1369 run out. I reviewed all the other calls to fork(); they all seem to check
1372 PH/20 When checking for unexpected SMTP input at connect time (before writing
1373 the banner), Exim was not dealing correctly with a non-positive return
1374 from the read() function. If the client had disconnected by this time,
1375 the result was a log entry for a synchronization error with an empty
1376 string after "input=" when read() returned zero. If read() returned -1
1377 (an event I could not check), uninitialized data bytes were printed.
1378 There were reports of junk text (parts of files, etc) appearing after
1381 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1383 PH/22 Added support for macro redefinition, and (re)definition in between
1384 driver and ACL definitions.
1386 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1387 forgetting to use the resulting value; it was using the unexpanded value.
1389 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1390 hadn't been configured. The fix is from Juergen Kreileder, who
1391 understands it better than I do:
1393 "Here's what I see happening with three configured cyrus_sasl
1394 authenticators configured (plain, login, cram-md5):
1396 On startup auth_cyrus_sasl_init() gets called for each of these.
1397 This means three calls to sasl_listmech() without a specified mech_list.
1398 => SASL tests which mechs of all available mechs actually work
1399 => three warnings about OTP not working
1400 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1402 With the patch, sasl_listmech() also gets called three times. But now
1403 SASL's mech_list option is set to the server_mech specified in the the
1404 authenticator. Or in other words, the answer from sasl_listmech()
1405 gets limited to just the mech you're testing for (which is different
1407 => the return list contains just 'plain' or 'login', 'cram-md5' or
1408 nothing depending on the value of ob->server_mech.
1410 I've just tested the patch: Authentication still works fine,
1411 unavailable mechs specified in the exim configuration are still
1412 caught, and the auth.log warnings about OTP are gone."
1414 PH/25 When debugging is enabled, the contents of the command line are added
1415 to the debugging output, even when log_selector=+arguments is not
1418 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1419 answer is "GNU", and only if the return is "GNU/something" is the answer
1422 PH/27 $acl_verify_message is now set immediately after the failure of a
1423 verification in an ACL, and so is available in subsequent modifiers. In
1424 particular, the message can be preserved by coding like this:
1426 warn !verify = sender
1427 set acl_m0 = $acl_verify_message
1429 Previously, $acl_verify_message was set only while expanding "message"
1430 and "log_message" when a very denied access.
1432 PH/28 Modified OS/os.c-Linux with
1434 -#ifndef OS_LOAD_AVERAGE
1435 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1437 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1438 nomenclature these days.)
1440 PH/29 Installed patch from the Sieve maintainer that adds the options
1441 sieve_useraddress and sieve_subaddress to the redirect router.
1443 PH/30 In these circumstances:
1444 . Two addresses routed to the same list of hosts;
1445 . First host does not offer TLS;
1446 . First host accepts first address;
1447 . First host gives temporary error to second address;
1448 . Second host offers TLS and a TLS session is established;
1449 . Second host accepts second address.
1450 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1451 and peerdn, if requested) that were in fact used only for the second
1454 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1455 attention to any local part prefix or suffix that was matched by the
1456 router that accepted the address. It now behaves in the same way as it
1457 does for delivery: the affixes are removed from the local part unless
1458 rcpt_include_affixes is set on the transport.
1460 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1461 timeout during the DATA phase of an incoming message.
1463 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1464 applied a patch sanctioned by the Sieve maintainer.
1466 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1467 the uid or gid is negative. A case of a negative gid caused this to be
1468 noticed. The fix allows for either to be negative.
1470 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1471 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1474 PH/36 The change PH/12 above was broken. Fixed it.
1476 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1477 the grounds that routing the same address twice would always produce the
1478 same answer. This might have been true once, but it is certainly no
1479 longer true now. Routing a child address may depend on the previous
1480 routing that produced that child. Some complicated redirection strategies
1481 went wrong when messages had multiple recipients, and made Exim's
1482 behaviour dependent on the order in which the addresses were given.
1484 I have moved the duplicate checking until after the routing is complete.
1485 Exim scans the addresses that are assigned to local and remote
1486 transports, and removes any duplicates. This means that more work will be
1487 done, as duplicates will always all be routed, but duplicates are
1488 presumably rare, so I don't expect this is of any significance.
1490 For deliveries to pipes, files, and autoreplies, the duplicate checking
1491 still happens during the routing process, since they are not going to be
1494 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1495 It corrects a timeout issue with spamd. This is Ian's comment: "The
1496 background is that sometimes spamd either never reads data from a
1497 connection it has accepted, or it never writes response data. The exiscan
1498 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1499 blindly assumes that writes won't block so it may never time out."
1501 PH/39 Allow G after quota size as well as K and M.
1503 PH/40 The value set for $authenticated_id in an authenticator may not contain
1504 binary zeroes or newlines because the value is written to log lines and
1505 to spool files. There was no check on this. Now the value is run through
1506 the string_printing() function so that such characters are converted to
1507 printable escape sequences.
1509 PH/41 $message_linecount is a new variable that contains the total number of
1510 lines in the message. Compare $body_linecount, which is the count for the
1513 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1514 bounce and delay warning messages, except in certain special cases, which
1517 (a) An SMTP error message from a remote host;
1518 (b) A message specified in a :fail: redirection;
1519 (c) A message specified in a "fail" command in a system filter;
1520 (d) A message specified in a FAIL return from the queryprogram router;
1521 (e) A message specified by the cannot_route_message router option.
1523 In these cases only, Exim does include the error details in bounce and
1524 warning messages. There are also a few cases where bland messages such
1525 as "unrouteable address" or "local delivery error" are given.
1527 PH/43 $value is now also set for the "else" part of a ${run expansion.
1529 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1530 being worked on, but at least Exim now implements the latest version to
1533 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1534 process to complete was treated as a delivery failure, a timeout while
1535 writing the message to the pipe was logged, but erroneously treated as a
1536 successful delivery. Such timeouts include transport filter timeouts. For
1537 consistency with the overall process timeout, these timeouts are now
1538 treated as errors, giving rise to delivery failures by default. However,
1539 there is now a new Boolean option for the pipe transport called
1540 timeout_defer, which, if set TRUE, converts the failures into defers for
1541 both kinds of timeout. A transport filter timeout is now identified in
1544 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1545 systems where "make" and "gmake" are different, calling "gmake" at top
1546 level broke things. I've arranged for the value of $(MAKE) to be passed
1547 from the Makefile to this script so that it can call the same version of
1551 A note about Exim versions 4.44 and 4.50
1552 ----------------------------------------
1554 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
1555 changes of various kinds. As a consequence, a big documentation update was
1556 needed. This delayed the release for rather longer than seemed good, especially
1557 in the light of a couple of (minor) security issues. Therefore, the changes
1558 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
1559 release. So 4.44 and 4.50 are in effect two different branches that both start
1562 I have left the 4.50 change log unchanged; it contains all the changes since
1563 4.43. The change log for 4.44 is below; many of its items are identical to
1564 those for 4.50. This seems to be the most sensible way to preserve the
1565 historical information.
1571 1. Minor wording change to the doc/README.SIEVE file.
1573 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
1574 computation of the current number of files was incorrect.
1576 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1577 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1580 4. Give more explanation in the error message when the command for a transport
1581 filter fails to execute.
1583 5. There are several places where Exim runs a non-Exim command in a
1584 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1585 was being done only for the command run by the queryprogram router. It is
1586 now done for all such subprocesses. The other cases are: ${run, transport
1587 filters, and the commands run by the lmtp and pipe transports.
1589 6. Added CONFIGURE_GROUP build-time option.
1591 7. Some older OS have a limit of 256 on the maximum number of file
1592 descriptors. Exim was using setrlimit() to set 1000 as a large value
1593 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1594 systems. I've change it so that if it can't get 1000, it tries for 256.
1596 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1597 was an oversight, and furthermore, ever since the addition of extra
1598 controls (e.g. 4.43/32), the checks on when to allow different forms of
1599 "control" were broken. There should now be diagnostics for all cases when a
1600 control that does not make sense is encountered.
1602 9. Added the /retain_sender option to "control=submission".
1604 10. $recipients is now available in the predata ACL (oversight).
1606 11. Tidy the search cache before the fork to do a delivery from a message
1607 received from the command line. Otherwise the child will trigger a lookup
1608 failure and thereby defer the delivery if it tries to use (for example) a
1609 cached ldap connection that the parent has called unbind on.
1611 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1612 of $address_data from the recipient verification was clobbered by the
1613 sender verification.
1615 13. The value of address_data from a sender verification is now available in
1616 $sender_address_data in subsequent conditions in the ACL statement.
1618 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
1620 15. Added a new option "connect=<time>" to callout options, to set a different
1623 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1624 was its contents. (It was OK if the option was not defined at all.)
1626 17. A "Completed" log line is now written for messages that are removed from
1627 the spool by the -Mrm option.
1629 18. New variables $sender_verify_failure and $recipient_verify_failure contain
1630 information about exactly what failed.
1632 19. Added -dd to debug only the daemon process.
1634 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
1635 handles timeouts, both on the server side and network timeouts. Renamed the
1636 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
1638 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
1639 It is now set to "smtps".
1641 22. $host_address is now set to the target address during the checking of
1642 ignore_target_hosts.
1644 23. When checking ignore_target_hosts for an ipliteral router, no host name was
1645 being passed; this would have caused $sender_host_name to have been used if
1646 matching the list had actually called for a host name (not very likely,
1647 since this list is usually IP addresses). A host name is now passed as
1650 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
1651 code that specifies a non-restarting handler (typically sigaction() in
1652 modern systems) in an attempt to fix a rare and obscure crash bug.
1654 25. Narrowed the window for a race in the daemon that could cause it to ignore
1655 SIGCHLD signals. This is not a major problem, because they are used only to
1656 wake it up if nothing else does.
1658 26. A malformed maildirsize file could cause Exim to calculate negative values
1659 for the mailbox size or file count. Odd effects could occur as a result.
1660 The maildirsize information is now recalculated if the size or filecount
1663 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
1664 support for a long time. Removed HAVE_SYS_VFS_H.
1666 28. Installed the latest version of exipick from John Jetmore.
1668 29. In an address list, if the pattern was not a regular expression, an empty
1669 subject address (from a bounce message) matched only if the pattern was an
1670 empty string. Non-empty patterns were not even tested. This was the wrong
1671 because it is perfectly reasonable to use an empty address as part of a
1672 database query. An empty address is now tested by patterns that are
1673 lookups. However, all the other forms of pattern expect the subject to
1674 contain a local part and a domain, and therefore, for them, an empty
1675 address still always fails if the pattern is not itself empty.
1677 30. Exim went into a mad DNS loop when attempting to do a callout where the
1678 host was specified on an smtp transport, and looking it up yielded more
1679 than one IP address.
1681 31. Re-factored the code for checking spool and log partition space into a
1682 function that finds that data and another that does the check. The former
1683 is then used to implement four new variables: $spool_space, $log_space,
1684 $spool_inodes, and $log_inodes.
1686 32. The RFC2047 encoding function was originally intended for short strings
1687 such as real names; it was not keeping to the 75-character limit for
1688 encoded words that the RFC imposes. It now respects the limit, and
1689 generates multiple encoded words if necessary. To be on the safe side, I
1690 have increased the buffer size for the ${rfc2047: expansion operator from
1693 33. It is now permitted to omit both strings after an "if" condition; if the
1694 condition is true, the result is "true". As before, when the second string
1695 is omitted, a false condition yields an empty string. This makes it less
1696 cumbersome to write custom ACL and router conditions.
1698 34. Failure to deliver a bounce message always caused it to be frozen, even if
1699 there was an errors_to setting on the router. The errors_to setting is now
1702 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
1703 canonical form (fully expanded) before being placed in
1704 $sender_host_address.
1706 36. The table in the code that translates DNS record types into text (T_A to
1707 "A" for instance) was missing entries for NS and CNAME. It is just possible
1708 that this could have caused confusion if both these types were looked up
1709 for the same domain, because the text type is used as part of Exim's
1710 per-process caching. But the chance of anyone hitting this buglet seems
1713 37. The dnsdb lookup has been extended in a number of ways.
1715 (1) There is a new type, "zns", which walks up the domain tree until it
1716 finds some nameserver records. It should be used with care.
1718 (2) There is a new type, "mxh", which is like "mx" except that it returns
1719 just the host names, not the priorities.
1721 (3) It is now possible to give a list of domains (or IP addresses) to be
1722 looked up. The behaviour when one of the lookups defers can be
1723 controlled by a keyword.
1725 (4) It is now possible to specify the separator character for use when
1726 multiple records are returned.
1728 38. The dnslists ACL condition has been extended: it is now possible to supply
1729 a list of IP addresses and/or domains to be looked up in a particular DNS
1732 39. Added log_selector=+queue_time_overall.
1734 40. When running the queue in the test harness, wait just a tad after forking a
1735 delivery process, to get repeatability of debugging output.
1737 41. Include certificate and key file names in error message when GnuTLS fails
1738 to set them up, because the GnuTLS error message doesn't include the name
1739 of the failing file when there is a problem reading it.
1741 42. Allow both -bf and -bF in the same test run.
1743 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
1745 44. The "Exiscan patch" is now merged into the mainline Exim source.
1747 45. Sometimes the final signoff response after QUIT could fail to get
1748 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
1749 < 0 before doing a fflush(). This bug looks as though it goes back to the
1750 introduction of TLS in release 3.20, but "sometimes" must have been rare
1751 because the tests only now provoked it.
1753 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
1754 (this can affect the format of dates).
1756 47. exim_tidydb, when checking for the continued existence of a message for
1757 which it has found a message-specific retry record, was not finding
1758 messages that were in split spool directories. Consequently, it was
1759 deleting retry records that should have stayed in existence.
1761 48. Steve fixed some bugs in eximstats.
1763 49. The SPA authentication driver was not abandoning authentication and moving
1764 on to the next authenticator when an expansion was forced to fail,
1765 contradicting the general specification for all authenticators. Instead it
1766 was generating a temporary error. It now behaves as specified.
1768 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
1769 (the order specifies the preference for clients). The order is now AES256,
1770 AES128, 3DES, ARCFOUR128.
1772 51. Small patch to Sieve code - explicitly set From: when generating an
1775 52. Exim crashed if a remote delivery caused a very long error message to be
1776 recorded - for instance if somebody sent an entire SpamAssassin report back
1777 as a large number of 550 error lines. This bug was coincidentally fixed by
1778 increasing the size of one of Exim's internal buffers (big_buffer) that
1779 happened as part of the Exiscan merge. However, to be on the safe side, I
1780 have made the code more robust (and fixed the comments that describe what
1783 53. Now that there can be additional text after "Completed" in log lines (if
1784 the queue_time_overall log selector is set), a one-byte patch to exigrep
1785 was needed to allow it to recognize "Completed" as not the last thing in
1788 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
1789 patch that reportedly fixes this has been added. I am not expert enough to
1790 create a test for it. This is what the patch creator wrote:
1792 "I found a little strange behaviour of ldap code when working with
1793 Windows 2003 AD Domain, where users was placed in more than one
1794 Organization Units. When I tried to give exim partial DN, the exit code
1795 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
1796 But simultaneously result of request was absolutely normal ldap result,
1797 so I produce this patch..."
1799 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
1800 REFERENCE, so I have modified the code to exclude the patch when that macro
1803 55. Some experimental protocols are using DNS PTR records for new purposes. The
1804 keys for these records are domain names, not reversed IP addresses. The
1805 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
1806 leaves it alone. Component reversal etc. now happens only for IP addresses.
1809 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
1811 57. Double the size of the debug message buffer (to 2048) so that more of very
1812 long debug lines gets shown.
1814 58. The exicyclog utility now does better if the number of log files to keep
1815 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
1817 59. Two changes related to the smtp_active_hostname option:
1819 (1) $smtp_active_hostname is now available as a variable.
1820 (2) The default for smtp_banner uses $smtp_active_hostname instead
1821 of $primary_hostname.
1823 60. The host_aton() function is supposed to be passed a string that is known
1824 to be a valid IP address. However, in the case of IPv6 addresses, it was
1825 not checking this. This is a hostage to fortune. Exim now panics and dies
1826 if the condition is not met. A case was found where this could be provoked
1827 from a dnsdb PTR lookup with an IPv6 address that had more than 8
1828 components; fortuitously, this particular loophole had already been fixed
1829 by change 4.50/55 above.
1831 If there are any other similar loopholes, the new check in host_aton()
1832 itself should stop them being exploited. The report I received stated that
1833 data on the command line could provoke the exploit when Exim was running as
1834 exim, but did not say which command line option was involved. All I could
1835 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
1836 running as the user.
1839 61. There was a buffer overflow vulnerability in the SPA authentication code
1840 (which came originally from the Samba project). I have added a test to the
1841 spa_base64_to_bits() function which I hope fixes it.
1844 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
1845 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
1847 63. The daemon start-up calls getloadavg() while still root for those OS that
1848 need the first call to be done as root, but it missed one case: when
1849 deliver_queue_load_max is set with deliver_drop_privilege. This is
1850 necessary for the benefit of the queue runner, because there is no re-exec
1851 when deliver_drop_privilege is set.
1853 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
1854 This has been fixed.
1856 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
1857 was in use, was not putting the data itself into the right store pool;
1858 consequently, it could be overwritten for a subsequent message in the same
1859 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
1862 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
1864 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
1865 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
1866 the function were treating the return as a boolean value, which happened to
1867 work because 0=false and not-0=true, but is not correct code.
1869 68. The host_aton() function was not handling scoped IPv6 addresses (those
1870 with, for example, "%eth0" on the end) correctly.
1872 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
1873 negated items (that is, ~something) in unsigned ints. Some compilers
1874 apparently mutter when there is no cast.
1876 70. If an address verification called from an ACL failed, and did not produce a
1877 user-specific message (i.e. there was only a "system" message), nothing was
1878 put in $acl_verify_message. In this situation, it now puts the system
1881 71. Change 4.23/11 added synchronization checking at the start of an SMTP
1882 session; change 4.31/43 added the unwanted input to the log line - except
1883 that it did not do this in the start of session case. It now does.
1885 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
1886 This is wrong and can cause the other end to generate a synchronization
1887 error if it is another Exim or anything else that does the synchronization
1888 check. A QUIT command is no longer sent after a timeout.
1890 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
1891 during host lookups.
1893 74. The defer_ok option of callout verification was not working if it was used
1894 when verifying addresses in header lines, that is, for this case:
1896 verify = header_sender/callout=defer_ok
1898 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
1899 those file descriptors could be used for SMTP connections. If anything
1900 wrote to stderr (the example that came up was "warn" in embedded Perl), it
1901 could be sent to the SMTP client, causing chaos. The daemon now opens
1902 stdin, stdout, and stderr to /dev/null when it puts itself into the
1905 76. Arrange for output from Perl's "warn" command to be written to Exim's main
1906 log by default. The user can override this with suitable Perl magic.
1908 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
1909 the log message when discard triggers, was not working for the DATA ACL or
1910 for the non-SMTP ACL.
1912 78. Error message wording change in sieve.c.
1914 79. If smtp_accept_max_per_host was set, the number of connections could be
1915 restricted to fewer than expected, because the daemon was trying to set up
1916 a new connection before checking whether the processes handling previous
1917 connections had finished. The check for completed processes is now done
1918 earlier. On busy systems, this bug wouldn't be noticed because something
1919 else would have woken the daemon, and it would have reaped the completed
1922 80. If a message was submitted locally by a user whose login name contained one
1923 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
1924 It caused a spool format error. I have fixed the spool reading code. A
1925 related problem was that the "from" clause in the Received: line became
1926 illegal because of the space(s). It is now covered by ${quote_local_part.
1928 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
1931 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
1933 83. Patch from Sieve maintainer to fix unterminated string problem in
1934 "vacation" handling.
1936 84. Some minor changes to the Linux configuration files to help with other
1937 OS variants using glibc.
1939 85. One more patch for Sieve to update vacation handling to latest spec.
1942 ----------------------------------------------------
1943 See the note above about the 4.44 and 4.50 releases.
1944 ----------------------------------------------------
1950 1. Change 4.43/35 introduced a bug that caused file counts to be
1951 incorrectly computed when quota_filecount was set in an appendfile
1954 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
1955 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
1958 3. Give more explanation in the error message when the command for a transport
1959 filter fails to execute.
1961 4. There are several places where Exim runs a non-Exim command in a
1962 subprocess. The SIGUSR1 signal should be disabled for these processes. This
1963 was being done only for the command run by the queryprogram router. It is
1964 now done for all such subprocesses. The other cases are: ${run, transport
1965 filters, and the commands run by the lmtp and pipe transports.
1967 5. Some older OS have a limit of 256 on the maximum number of file
1968 descriptors. Exim was using setrlimit() to set 1000 as a large value
1969 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
1970 systems. I've change it so that if it can't get 1000, it tries for 256.
1972 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
1973 was an oversight, and furthermore, ever since the addition of extra
1974 controls (e.g. 4.43/32), the checks on when to allow different forms of
1975 "control" were broken. There should now be diagnostics for all cases when a
1976 control that does not make sense is encountered.
1978 7. $recipients is now available in the predata ACL (oversight).
1980 8. Tidy the search cache before the fork to do a delivery from a message
1981 received from the command line. Otherwise the child will trigger a lookup
1982 failure and thereby defer the delivery if it tries to use (for example) a
1983 cached ldap connection that the parent has called unbind on.
1985 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
1986 of $address_data from the recipient verification was clobbered by the
1987 sender verification.
1989 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
1990 was its contents. (It was OK if the option was not defined at all.)
1992 11. A "Completed" log line is now written for messages that are removed from
1993 the spool by the -Mrm option.
1995 12. $host_address is now set to the target address during the checking of
1996 ignore_target_hosts.
1998 13. When checking ignore_target_hosts for an ipliteral router, no host name was
1999 being passed; this would have caused $sender_host_name to have been used if
2000 matching the list had actually called for a host name (not very likely,
2001 since this list is usually IP addresses). A host name is now passed as
2004 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2005 code that specifies a non-restarting handler (typically sigaction() in
2006 modern systems) in an attempt to fix a rare and obscure crash bug.
2008 15. Narrowed the window for a race in the daemon that could cause it to ignore
2009 SIGCHLD signals. This is not a major problem, because they are used only to
2010 wake it up if nothing else does.
2012 16. A malformed maildirsize file could cause Exim to calculate negative values
2013 for the mailbox size or file count. Odd effects could occur as a result.
2014 The maildirsize information is now recalculated if the size or filecount
2017 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2018 support for a long time. Removed HAVE_SYS_VFS_H.
2020 18. Updated exipick to current release from John Jetmore.
2022 19. Allow an empty sender to be matched against a lookup in an address list.
2023 Previously the only cases considered were a regular expression, or an
2026 20. Exim went into a mad DNS lookup loop when doing a callout where the
2027 host was specified on the transport, if the DNS lookup yielded more than
2030 21. The RFC2047 encoding function was originally intended for short strings
2031 such as real names; it was not keeping to the 75-character limit for
2032 encoded words that the RFC imposes. It now respects the limit, and
2033 generates multiple encoded words if necessary. To be on the safe side, I
2034 have increased the buffer size for the ${rfc2047: expansion operator from
2037 22. Failure to deliver a bounce message always caused it to be frozen, even if
2038 there was an errors_to setting on the router. The errors_to setting is now
2041 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2042 canonical form (fully expanded) before being placed in
2043 $sender_host_address.
2045 24. Updated eximstats to version 1.33
2047 25. Include certificate and key file names in error message when GnuTLS fails
2048 to set them up, because the GnuTLS error message doesn't include the name
2049 of the failing file when there is a problem reading it.
2051 26. Expand error message when OpenSSL has problems setting up cert/key files.
2054 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2055 (this can affect the format of dates).
2057 28. exim_tidydb, when checking for the continued existence of a message for
2058 which it has found a message-specific retry record, was not finding
2059 messages that were in split spool directories. Consequently, it was
2060 deleting retry records that should have stayed in existence.
2062 29. eximstats updated to version 1.35
2063 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2064 1.35 - bugfix such that pie charts by volume are generated correctly
2066 30. The SPA authentication driver was not abandoning authentication and moving
2067 on to the next authenticator when an expansion was forced to fail,
2068 contradicting the general specification for all authenticators. Instead it
2069 was generating a temporary error. It now behaves as specified.
2071 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2072 (the order specifies the preference for clients). The order is now AES256,
2073 AES128, 3DES, ARCFOUR128.
2075 31. Small patch to Sieve code - explicitly set From: when generating an
2078 32. Exim crashed if a remote delivery caused a very long error message to be
2079 recorded - for instance if somebody sent an entire SpamAssassin report back
2080 as a large number of 550 error lines. This bug was coincidentally fixed by
2081 increasing the size of one of Exim's internal buffers (big_buffer) that
2082 happened as part of the Exiscan merge. However, to be on the safe side, I
2083 have made the code more robust (and fixed the comments that describe what
2086 33. Some experimental protocols are using DNS PTR records for new purposes. The
2087 keys for these records are domain names, not reversed IP addresses. The
2088 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2089 leaves it alone. Component reversal etc. now happens only for IP addresses.
2092 34. The host_aton() function is supposed to be passed a string that is known
2093 to be a valid IP address. However, in the case of IPv6 addresses, it was
2094 not checking this. This is a hostage to fortune. Exim now panics and dies
2095 if the condition is not met. A case was found where this could be provoked
2096 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2097 components; fortuitously, this particular loophole had already been fixed
2098 by change 4.50/55 or 4.44/33 above.
2100 If there are any other similar loopholes, the new check in host_aton()
2101 itself should stop them being exploited. The report I received stated that
2102 data on the command line could provoke the exploit when Exim was running as
2103 exim, but did not say which command line option was involved. All I could
2104 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2105 running as the user.
2108 35. There was a buffer overflow vulnerability in the SPA authentication code
2109 (which came originally from the Samba project). I have added a test to the
2110 spa_base64_to_bits() function which I hope fixes it.
2113 36. The daemon start-up calls getloadavg() while still root for those OS that
2114 need the first call to be done as root, but it missed one case: when
2115 deliver_queue_load_max is set with deliver_drop_privilege. This is
2116 necessary for the benefit of the queue runner, because there is no re-exec
2117 when deliver_drop_privilege is set.
2119 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2120 was in use, was not putting the data itself into the right store pool;
2121 consequently, it could be overwritten for a subsequent message in the same
2122 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2125 38. Sometimes the final signoff response after QUIT could fail to get
2126 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2127 < 0 before doing a fflush(). This bug looks as though it goes back to the
2128 introduction of TLS in release 3.20, but "sometimes" must have been rare
2129 because the tests only now provoked it.
2135 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2136 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2137 Now it returns an integer. A number of calls were still expecting a T/F
2138 return. Fortuitously, in all cases, the tests worked in OK situations,
2139 which is the norm. However, things would have gone wrong on any write
2140 failures on the smtp file descriptor. This function is used when sending
2141 messages over SMTP and also when doing verify callouts.
2143 2. When Exim is called to do synchronous delivery of a locally submitted
2144 message (the -odf or -odi options), it no longer closes stderr before doing
2147 3. Implemented the mua_wrapper option.
2149 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2151 5. Implemented the functions header_remove(), header_testname(),
2152 header_add_at_position(), and receive_remove_recipient(), and exported them
2155 6. If an ACL "warn" statement specified the addition of headers, Exim already
2156 inserted X-ACL-Warn: at the start if there was no header name. However, it
2157 was not making this test for the second and subsequent header lines if
2158 there were newlines in the string. This meant that an invalid header could
2159 be inserted if Exim was badly configured.
2161 7. Allow an ACL "warn" statement to add header lines at the start or after all
2162 the Received: headers, as well as at the end.
2164 8. Added the rcpt_4xx retry error code.
2166 9. Added postmaster_mailfrom=xxx to callout verification option.
2168 10. Added mailfrom=xxxx to the callout verification option, for verify=
2171 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2172 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2174 12. Inserted some casts to stop certain compilers warning when using pointer
2175 differences as field lengths or precisions in printf-type calls (mostly
2176 affecting debugging statements).
2178 13. Added optional readline() support for -be (dynamically loaded).
2180 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2181 same clock tick as a message's arrival, so that its received time was the
2182 same as the "first fail" time on the retry record, and that message
2183 remained on the queue past the ultimate address timeout, every queue runner
2184 would try a delivery (because it was past the ultimate address timeout) but
2185 after another failure, the ultimate address timeout, which should have then
2186 bounced the address, did not kick in. This was a "< instead of <=" error;
2187 in most cases the first failure would have been in the next clock tick
2188 after the received time, and all would be well.
2190 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2191 being recognized when the domain list was tested by the match_domain
2192 condition in an expansion string.
2194 16. Added the ${str2b64: operator.
2196 17. Exim was always calling setrlimit() to set a large limit for the number of
2197 processes, without checking whether the existing limit was already
2198 adequate. (It did check for the limit on file descriptors.) Furthermore,
2199 errors from getrlimit() and setrlimit() were being ignored. Now they are
2200 logged to the main and panic logs, but Exim does carry on, to try to do its
2201 job under whatever limits there are.
2203 18. Imported PCRE 5.0.
2205 19. Trivial typo in log message " temporarily refused connection" (the leading
2208 20. If the log selector return_path_on_delivery was set and an address was
2209 redirected to /dev/null, the delivery process crashed because it assumed
2210 that a return path would always be set for a "successful" delivery. In this
2211 case, the whole delivery is bypassed as an optimization, and therefore no
2214 21. Internal re-arrangement: the function for sending a challenge and reading
2215 a response while authentication was assuming a zero-terminated challenge
2216 string. It's now changed to take a pointer and a length, to allow for
2217 binary data in such strings.
2219 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2221 23. Exim was not respecting finduser_retries when seeking the login of the
2222 uid under which it was called; it was always trying 10 times. (The default
2223 setting of finduser_retries is zero.) Also, it was sleeping after the final
2224 failure, which is pointless.
2226 24. Implemented tls_on_connect_ports.
2228 25. Implemented acl_smtp_predata.
2230 26. If the domain in control=submission is set empty, Exim assumes that the
2231 authenticated id is a complete email address when it generates From: or
2232 Sender: header lines.
2234 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2235 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2236 chown and chgrp in /bin and hostname in /usr/bin.
2238 28. Exim was keeping the "process log" file open after each use, just as it
2239 does for the main log. This opens the possibility of it remaining open for
2240 long periods when the USR1 signal hits a daemon. Occasional processlog
2241 errors were reported, that could have been caused by this. Anyway, it seems
2242 much more sensible not to leave this file open at all, so that is what now
2245 29. The long-running daemon process does not normally write to the log once it
2246 has entered its main loop, and it closes the log before doing so. This is
2247 so that log files can straightforwardly be renamed and moved. However,
2248 there are a couple of unusual error situations where the daemon does write
2249 log entries, and I had neglected to close the log afterwards.
2251 30. The text of an SMTP error response that was received during a remote
2252 delivery was being truncated at 512 bytes. This is too short for some of
2253 the long messages that one sometimes sees. I've increased the limit to
2256 31. It is now possible to make retry rules that apply only when a message has a
2257 specific sender, in particular, an empty sender.
2259 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2260 it possible to be selective about when SMTP synchronization is enforced.
2262 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2264 32. Implemented hosts_connection_nolog.
2266 33. Added an ACL for QUIT.
2268 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2271 35. Added mailbox_size and mailbox_filecount to appendfile.
2273 36. Added control = no_multiline_responses to ACLs.
2275 37. There was a bug in the logic of the code that waits for the clock to tick
2276 in the case where the clock went backwards by a substantial amount such
2277 that the microsecond fraction of "now" was more than the microsecond
2278 fraction of "then" (but the whole seconds number was less).
2280 38. Added support for the libradius Radius client library this is found on
2281 FreeBSD (previously only the radiusclient library was supported).
2287 1. When certain lookups returned multiple values in the form name=value, the
2288 quoting of the values was not always being done properly. Specifically:
2289 (a) If the value started with a double quote, but contained no whitespace,
2291 (b) If the value contained whitespace other than a space character (i.e.
2292 tabs or newlines or carriage returns) it was not quoted.
2293 This fix has been applied to the mysql and pgsql lookups by writing a
2294 separate quoting function and calling it from the lookup code. The fix
2295 should probably also be applied to nisplus, ibase and oracle lookups, but
2296 since I cannot test any of those, I have not disturbed their existing code.
2298 2. A hit in the callout cache for a specific address caused a log line with no
2299 reason for rejecting RCPT. Now it says "Previous (cached) callout
2300 verification failure".
2302 3. There was an off-by-one bug in the queryprogram router. An over-long
2303 return line was truncated at 256 instead of 255 characters, thereby
2304 overflowing its buffer with the terminating zero. As well as fixing this, I
2305 have increased the buffer size to 1024 (and made a note to document this).
2307 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2308 when Exim is waiting for an SMTP response from a remote server, Exim
2309 restarts its select() call on the socket, thereby resetting its timeout.
2310 This is not a problem when such interrupts are rare. Somebody set up a cron
2311 job to run exiwhat every 2 minutes, which is less than the normal select()
2312 timeout (5 or 10 minutes). This meant that the select() timeout never
2313 kicked in because it was always reset. I have fixed this by comparing the
2314 time when an interrupt arrives with the time at the start of the first call
2315 to select(). If more time than the timeout has elapsed, the interrupt is
2316 treated as a timeout.
2318 5. Some internal re-factoring in preparation for the addition of Sieve
2319 extensions (by MH). In particular, the "personal" test is moved to a
2320 separate function, and given an option for scanning Cc: and Bcc: (which is
2321 not set for Exim filters).
2323 6. When Exim created an email address using the login of the caller as the
2324 local part (e.g. when creating a From: or Sender: header line), it was not
2325 quoting the local part when it contained special characters such as @.
2327 7. Installed new OpenBSD configuration files.
2329 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2330 try to make them clearer.
2332 9. Callout options, other than the timeout value, were being ignored when
2333 verifying sender addresses in header lines. For example, when using
2335 verify = header_sender/callout=no_cache
2337 the cache was (incorrectly) being used.
2339 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2340 only the Cygwin environment.
2342 11. When return_path_on_delivery was set as a log selector, if different remote
2343 addresses in the same message used different return paths and parallel
2344 remote delivery occurred, the wrong values would sometimes be logged.
2345 (Whenever a remote delivery process finished, the return path value from
2346 the most recently started remote delivery process was logged.)
2348 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2349 lines when AUTH and/or TLS are in use. This is the "received protocol"
2350 field. Exim used to use "asmtp" for authenticated SMTP, without any
2351 indication (in the protocol name) for TLS use. Now it follows the RFC and
2352 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2353 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2354 names appear in log lines as well as in Received: header lines.
2356 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2357 extensions, and comparison tests, and to fix some bugs.
2359 14. Changes to the "personal" filter test:
2361 (1) The test was buggy in that it was just doing the equivalent of
2362 "contains" tests on header lines. For example, if a user's address was
2363 anne@some.where, the "personal" test would incorrectly be true for
2365 To: susanne@some.where
2367 This test is now done by extracting each address from the header in turn,
2368 and checking the entire address. Other tests that are part of "personal"
2369 are now done using regular expressions (for example, to check local parts
2370 of addresses in From: header lines).
2372 (2) The list of non-personal local parts in From: addresses has been
2373 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2374 taken from the Sieve specification recommendations.
2376 (3) If the message contains any header line starting with "List-" it is
2377 treated as non-personal.
2379 (4) The test for "circular" in the Subject: header line has been removed
2380 because it now seems ill-conceived.
2382 15. Minor typos in src/EDITME comments corrected.
2384 16. Installed latest exipick from John Jetmore.
2386 17. If headers_add on a router specified a text string that was too long for
2387 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2388 of string_sprintf() is now avoided.
2390 18. $message_body_size was not set (it was always zero) when running the DATA
2391 ACL and the local_scan() function.
2393 19. For the "mail" command in an Exim filter, no default was being set for
2394 the once_repeat time, causing a random time value to be used if "once" was
2395 specified. (If the value happened to be <= 0, no repeat happened.) The
2396 default is now 0s, meaning "never repeat". The "vacation" command was OK
2397 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2398 (I found it when inspecting the code).
2400 20. There is now an overall timeout for performing a callout verification. It
2401 defaults to 4 times the callout timeout, which applies to individual SMTP
2402 commands during the callout. The overall timeout applies when there is more
2403 than one host that can be tried. The timeout is checked before trying the
2404 next host. This prevents very long delays if there are a large number of
2405 hosts and all are timing out (e.g. when the network connections are timing
2406 out). The value of the overall timeout can be changed by specifying an
2407 additional sub-option for "callout", called "maxwait". For example:
2409 verify = sender/callout=5s,maxwait=20s
2411 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2412 to the end before writing, but this should make it even safer).
2414 22. Exim was forgetting that it had advertised PIPELINING for the second and
2415 subsequent messages on an SMTP connection. It was also not resetting its
2416 memory on STARTTLS and an internal HELO.
2418 23. When Exim logs an SMTP synchronization error within a session, it now
2419 records whether PIPELINING has been advertised or not.
2421 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2422 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2423 rather than long int.
2425 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2427 26. Added the never_mail option to autoreply.
2433 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2434 crash if the getsockname() call failed; this can happen if a connection is
2435 closed very soon after it is established. The problem was simply in the
2436 order in which certain operations were done, causing Exim to try to write
2437 to the SMTP stream before it had set up the file descriptor. The bug has
2438 been fixed by making things happen in the correct order.
2444 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2445 before the connection was closed, thus losing the rejection response.
2447 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2448 some early Solaris releases, but causes trouble in current releases where
2449 socklen_t is defined.
2451 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2454 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2457 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2458 that releasing the top of it at the end releases what was used for sub-
2459 expansions (unless the block got too big). However, discard this block if
2460 the first thing is a variable or header, so that we can use its block when
2461 it is dynamic (useful for very large $message_headers, for example).
2463 6. Lookups now cache *every* query, not just the most recent. A new, separate
2464 store pool is used for this. It can be recovered when all lookup caches are
2465 flushed. Lookups now release memory at the end of their result strings.
2466 This has involved some general refactoring of the lookup sources.
2468 7. Some code has been added to the store_xxx() functions to reduce the amount
2469 of flapping under certain conditions.
2471 8. log_incoming_interface used to affect only the <= reception log lines. Now
2472 it causes the local interface and port to be added to several more SMTP log
2473 lines, for example "SMTP connection from", and rejection lines.
2475 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2477 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2479 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2480 could be overwritten at the end of the current message (or the start of a
2481 new message if it was set in a HELO ACL). The value is now preserved for
2482 the duration of the SMTP connection.
2484 12. If a transport had a headers_rewrite setting, and a matching header line
2485 contained an unqualified address, that address was qualified, even if it
2486 did not match any rewriting rules. The underlying bug was that the values
2487 of the flags that permit the existence of unqualified sender and recipient
2488 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2489 non-local messages, and by -bnq for local messages) were not being
2490 preserved with the message after it was received.
2492 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2493 "next input=" as part of the text comprising the host identity instead of
2494 the correct text. The code was using the same buffer for two different
2495 strings. However, depending on which order the printing function evaluated
2496 its arguments, the bug did not always show up. Under Linux, for example, my
2497 test suite worked just fine.
2499 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2500 doesn't work with some older versions of Perl. It has been changed to "my",
2501 which in any case is probably the better facility to use.
2503 15. A really picky compiler found some instances of statements for creating
2504 error messages that either had too many or two few arguments for the format
2507 16. The size of the buffer for calls to the DNS resolver has been increased
2508 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2509 for addresses that have a lot of PTR records. This alleviates a problem; it
2510 does not fully solve it.
2512 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2513 buffer now truncates the list and logs the incident, which is the same
2514 action as happens when Exim is looking up a host name and its aliases.
2515 Previously in this situation something unpredictable would happen;
2516 sometimes it was "internal error: store_reset failed".
2518 18. If a server dropped the connection unexpectedly when an Exim client was
2519 using GnuTLS and trying to read a response, the client delivery process
2520 crashed while trying to generate an error log message.
2522 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2523 string, for example:
2525 warn message = H1: something\nH2: something
2527 the text was added as a single header line from Exim's point of view
2528 though it ended up OK in the delivered message. However, searching for the
2529 second and subsequent header lines using $h_h2: did not work. This has been
2530 fixed. Similarly, if a system filter added multiple headers in this way,
2531 the routers could not see them.
2533 20. Expanded the error message when iplsearch is called with an invalid key to
2534 suggest using net-iplsearch in a host list.
2536 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2537 ACLs are no longer actually imposed (and a message to that effect is
2540 22. If a "gecos" field in a passwd entry contained escaped characters, in
2541 particular, if it contained a \" sequence, Exim got it wrong when building
2542 a From: or a Sender: header from that name. A second bug also caused
2543 incorrect handling when an unquoted " was present following a character
2544 that needed quoting.
2546 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2547 was not being matched caselessly.
2549 24. Arranged for all hyphens in the exim.8 source to be escaped with
2552 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2553 properly. Recipient callout cache records were still being keyed to include
2554 the sender, even when use_sender was set false. This led to far more
2555 callouts that were necessary. The sender is no longer included in the key
2556 when use_sender is false.
2558 26. Added "control = submission" modifier to ACLs.
2560 27. Added the ${base62d: operator to decode base 62 numbers.
2562 28. dnsdb lookups can now access SRV records.
2564 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
2565 the configuration file.
2567 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
2568 (-v) mode. This makes the output for a verbose queue run more intelligible.
2570 31. Added a use_postmaster feature to recipient callouts.
2572 32. Added the $body_zerocount variable, containing the number of binary zero
2573 bytes in the message body.
2575 33. The time of last modification of the "new" subdirectory is now used as the
2576 "mailbox time last read" when there is a quota error for a maildir
2579 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
2581 35. Added +ignore_unknown as a special item in host lists.
2583 36. Code for decoding IPv6 addresses in host lists is now included, even if
2584 IPv6 support is not being compiled. This fixes a bug in which an IPv6
2585 address was recognized as an IP address, but was then not correctly decoded
2586 into binary, causing unexpected and incorrect effects when compared with
2593 1. Very minor rewording of debugging text in manualroute to say "list of
2594 hosts" instead of "hostlist".
2596 2. If verify=header_syntax was set, and a header line with an unqualified
2597 address (no domain) and a large number of spaces between the end of the
2598 name and the colon was received, the reception process suffered a buffer
2599 overflow, and (when I tested it) crashed. This was caused by some obsolete
2600 code that should have been removed. The fix is to remove it!
2602 3. When running in the test harness, delay a bit after writing a bounce
2603 message to get a bit more predictability in the log output.
2605 4. Added a call to search_tidyup() just before forking a reception process. In
2606 theory, someone could use a lookup in the expansion of smtp_accept_max_
2607 per_host which, without the tidyup, could leave open a database connection.
2609 5. Added the variables $recipient_data and $sender_data which get set from a
2610 lookup success in an ACL "recipients" or "senders" condition, or a router
2611 "senders" option, similar to $domain_data and $local_part_data.
2613 6. Moved the writing of debug_print from before to after the "senders" test
2616 7. Change 4.31/66 (moving the time when the Received: is generated) caused
2617 problems for message scanning, either using a data ACL, or using
2618 local_scan() because the Received: header was not generated till after they
2619 were called (in order to set the time as the time of reception completion).
2620 I have revised the way this works. The header is now generated after the
2621 body is received, but before the ACL or local_scan() are called. After they
2622 are run, the timestamp in the header is updated.
2628 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
2629 before starting a queue runner without re-exec. This happened only when
2630 deliver_drop_privilege was set or when the Exim user was set to root. The
2631 effect of the bug was that timeouts during subsequent deliveries caused
2632 crashes instead of being properly handled. The handler is now left at its
2633 default (and expected) setting.
2635 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
2636 message, again when deliver_drop_privilege is set or Exim is run as root.
2637 The bug described in (1) was not present in this case, but the tidying up
2638 of the other signals was missing. I have made the two cases consistent.
2640 3. The ignore_target_hosts setting on a manualroute router was being ignored
2641 for hosts that were looked up using the /MX notation.
2643 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
2646 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
2647 operated on the sender address. After changing the $sender_address to <>
2648 for the sender address verify, Exim was re-instated it as the original
2649 (before rewriting) address, but remembering that it had rewritten it, so it
2650 wasn't rewriting it again. This bug also had the effect of breaking the
2651 sender address verification caching when the sender address was rewritten.
2653 6. The ignore_target_hosts option was being ignored by the ipliteral router.
2654 This has been changed so that if the ip literal address matches
2655 ignore_target_hosts, the router declines.
2657 7. Added expansion conditions match_domain, match_address, and match_local_
2658 part (NOT match_host).
2660 8. The placeholder for the Received: header didn't have a length field set.
2662 9. Added code to Exim itself and to exim_lock to test for a specific race
2663 condition that could lead to file corruption when using MBX delivery. The
2664 issue is with the lockfile that is created in /tmp. If this file is removed
2665 after a process has opened it but before that process has acquired a lock,
2666 there is the potential for a second process to recreate the file and also
2667 acquire a lock. This could lead to two Exim processes writing to the file
2668 at the same time. The added code performs the same test as UW imapd; it
2669 checks after acquiring the lock that its file descriptor still refers to
2670 the same named file.
2672 10. The buffer for building added header lines was of fixed size, 8192 bytes.
2673 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
2676 11. Added the smtp_active_hostname option. If used, this will typically be made
2677 to depend on the incoming interface address. Because $interface_address is
2678 not set up until the daemon has forked a reception process, error responses
2679 that can happen earlier (such as "too many connections") no longer contain
2682 12. If an expansion in a condition on a "warn" statement fails because a lookup
2683 defers, the "warn" statement is abandoned, and the next ACL statement is
2684 processed. Previously this caused the whole ACL to be aborted.
2686 13. Added the iplsearch lookup type.
2688 14. Added ident_timeout as a log selector.
2690 15. Added tls_certificate_verified as a log selector.
2692 16. Added a global option tls_require_ciphers (compare the smtp transport
2693 option of the same name). This controls incoming TLS connections.
2695 17. I finally figured out how to make tls_require_ciphers do a similar thing
2696 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
2697 before starting the TLS session.
2699 18. Tabs are now shown as \t in -bP output.
2701 19. If the log selector return_path_on_delivery was set, Exim crashed when
2702 bouncing a message because it had too many Received: header lines.
2704 20. If two routers both had headers_remove settings, and the first one included
2705 a superfluous trailing colon, the final name in the first list and the
2706 first name in the second list were incorrectly joined into one item (with a
2707 colon in the middle).
2713 1. Added -C and -D options to the exinext utility, mainly to make it easier
2714 to include in the automated testing, but these could be helpful when
2715 multiple configurations are in use.
2717 2. The exinext utility was not formatting the output nicely when there was
2718 an alternate port involved in the retry record key, nor when there was a
2719 message id as well (for retries that were specific to a specific message
2720 and a specific host). It was also confused by IPv6 addresses, because of
2721 the additional colons they contain. I have fixed the IPv4 problem, and
2722 patched it up to do a reasonable job for IPv6.
2724 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
2725 delivery, the log line now contains "pipelined" if PIPELINING was used.
2727 4. An SMTP transport process used to panic and die if the bind() call to set
2728 an explicit outgoing interface failed. This has been changed; it is now
2729 treated in the same way as a connect() failure.
2731 5. A reference to $sender_host_name in the part of a conditional expansion
2732 that was being skipped was still causing a DNS lookup. This no longer
2735 6. The def: expansion condition was not recognizing references to header lines
2736 that used bh_ and bheader_.
2738 7. Added the _cache feature to named lists.
2740 8. The code for checking quota_filecount in the appendfile transport was
2741 allowing one more file than it should have been.
2743 9. For compatibility with Sendmail, the command line option
2751 and sets the incoming protocol and host name (for trusted callers). The
2752 host name and its colon can be omitted when only the protocol is to be set.
2753 Note the Exim already has two private options, -pd and -ps, that refer to
2754 embedded Perl. It is therefore impossible to set a protocol value of "d" or
2755 "s", but I don't think that's a major issue.
2757 10. A number of refactoring changes to the code, none of which should affect
2760 (a) The number of logging options was getting close to filling up the
2761 32-bit word that was used as a bit map. I have split them into two classes:
2762 those that are passed in the argument to log_write(), and those that are
2763 only ever tested independently outside of that function. These are now in
2764 separate 32-bit words, so there is plenty of room for expansion again.
2765 There is no change in the user interface or the logging behaviour.
2767 (b) When building, for example, log lines, the code previously used a
2768 macro that called string_cat() twice, in order to add two strings. This is
2769 not really sufficiently general. Furthermore, there was one instance where
2770 it was actually wrong because one of the argument was used twice, and in
2771 one call a function was used. (As it happened, calling the function twice
2772 did not affect the overall behaviour.) The macro has been replaced by a
2773 function that can join an arbitrary number of extra strings onto a growing
2776 (c) The code for expansion conditions now uses a table and a binary chop
2777 instead of a serial search (which was left over from when there were very
2778 few conditions). Also, it now recognizes conditions like "pam" even when
2779 the relevant support is not compiled in: a suitably worded error message is
2780 given if an attempt is made to use such a condition.
2782 11. Added ${time_interval:xxxxx}.
2784 12. A bug was causing one of the ddress fields not to be passed back correctly
2785 from remote delivery subprocesses. The field in question was not being
2786 subsequently used, so this caused to problems in practice.
2788 13. Added new log selectors queue_time and deliver_time.
2790 14. Might have fixed a bug in maildirsizefile handling that threw up
2791 "unexpected character" debug warnings, and recalculated the data
2792 unnecessarily. In any case, I expanded the warning message to give more
2795 15. Added the message "Restricted characters in address" to the statements in
2796 the default ACL that block characters like @ and % in local parts.
2798 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
2799 Three changes have been made:
2801 (a) There was a serious bug; a negative response to MAIL caused the whole
2802 recipient domain to be cached as invalid, thereby blocking all messages
2803 to all local parts at the same domain, from all senders. This bug has
2804 been fixed. The domain is no longer cached after a negative response to
2805 MAIL if the sender used is not empty.
2807 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
2810 (c) A new callout option, "use_sender" has been added for people who want
2811 the modified behaviour.
2817 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
2820 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
2821 indeed breaks things for older releases.
2823 3. Added additional logging to the case where there is a problem reading data
2824 from a filter that is running in a subprocess using a pipe, in order to
2825 try to track down a specific problem.
2827 4. Testing facility fudge: when running in the test harness and attempting
2828 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
2829 getting "No route to host". Convert this to a timeout.
2831 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
2834 6. Some OS don't have socklen_t but use size_t instead. This affects the
2835 fifth argument of getsockopt() amongst other things. This is now
2836 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
2837 can be set for individual OS. I have set it for SunOS5, OSF1, and
2838 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
2839 some earlier ones do not.
2841 7. Change 4.30/15 was not doing the test caselessly.
2843 8. The standard form for an IPv6 address literal was being rejected by address
2844 parsing in, for example, MAIL and RCPT commands. An example of this kind of
2845 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
2846 this, as well as the form without the "IPv6" on the front (but only when
2847 address literals are enabled, of course).
2849 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
2851 10. Exim crashed if a message with an empty sender address specified by -f
2852 encountered a router with an errors_to setting. This could be provoked only
2853 by a command such as
2857 where an empty string was supplied; "<>" did not hit this bug.
2859 11. Installed PCRE release 4.5.
2861 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
2862 remained set. It is now erased.
2864 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
2865 times from message ids (which are base 36 rather than the normal 62).
2867 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
2868 were being counted as actual protocol errors, and logged if the log
2869 selector +smtp_protocol_error was set. One cannot be perfect in this test,
2870 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
2871 and DATA following a set of rejected RCPTs do not count as protocol errors.
2872 In other words, Exim assumes they were pipelined, though this may not
2873 actually be the case. Of course, in all cases the client gets an
2874 appropriate error code.
2876 15. If a lookup fails in an ACL condition, a message about the failure may
2877 be available; it is used if testing the ACL cannot continue, because most
2878 such messages specify what the cause of the deferral is. However, some
2879 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
2880 that caused an old message to be retained and used if a later statement
2881 caused a defer, replacing the real cause of the deferral.
2883 16. If an IP address had so many PTR records that the DNS lookup buffer
2884 was not large enough to hold them, Exim could crash while trying to process
2885 the truncated data. It now detects and logs this case.
2887 17. Further to 4.21/58, another change has been made: if (and only if) the
2888 first line of a message (the first header line) ends with CRLF, a bare LF
2889 in a subsequent header line has a space inserted after it, so as not to
2890 terminate the header.
2892 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
2893 unnecessarily, used atoi() instead of strtol(), and didn't check the
2894 termination when getting file sizes from file names by regex.
2896 19. Completely re-implemented the support for maildirsize files, in the light
2897 of a number of problems with the previous contributed implementation
2898 (4.30/29). In particular:
2900 . If the quota is zero, the maildirsize file is maintained, but no quota is
2903 . If the maildir directory does not exist, it is created before any attempt
2904 to write a maildirsize file.
2906 . The quota value in the file is just a cache; if the quota is changed in
2907 the transport, the new value overrides.
2909 . A regular expression is available for excluding directories from the
2912 20. The autoreply transport checks the characters in options that define the
2913 message's headers; it allows continued headers, but it was checking with
2914 isspace() after an embedded newline instead of explicitly looking for a
2917 21. If all the "regular" hosts to which an address was routed had passed their
2918 expiry times, and had not reached their retry times, the address was
2919 bounced, even if fallback hosts were defined. Now Exim should go on to try
2922 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
2923 equivalent code in the SMTP transport. Some hosts send humungous responses
2924 to HELO/EHLO, more than 1024 it seems.
2926 23. Refactoring: code in filter.c used (void *) for "any old type" but this
2927 gives compiler warnings in some environments. I've now done it "properly",
2930 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
2931 (because of problems with the built-in one) was declared to return uschar *
2932 instead of char *, causing compiler failure.
2934 25. Fixed a file descriptor leak when processing alias/forward files.
2936 26. Fixed a minor format string issue in dbfn.c.
2938 27. Typo in exim.c: ("dmbnz" for "dbmnz").
2940 28. If a filter file refered to $h_xxx or $message_headers, and the headers
2941 contained RFC 2047 "words", Exim's memory could, under certain conditions,
2944 29. When a sender address is verified, it is cached, to save repeating the test
2945 when there is more than one recipient in a message. However, when the
2946 verification involves a callout, it is possible for different callout
2947 options to be set for different recipients. It is too complicated to keep
2948 track of this in the cache, so now Exim always runs a verification when a
2949 callout is required, relying on the callout cache for the optimization.
2950 The overhead is duplication of the address routing, but this should not be
2953 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
2954 to be verified with callout=postmaster, and the main callout worked but the
2955 postmaster check failed, the verification correctly failed. However, if a
2956 subsequent RCPT command asked for sender verification *without* the
2957 postmaster check, incorrect caching caused this verification also to fail,
2960 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
2961 it was not caching the DNS options (qualify_single, search_parents) that
2962 were used when the lookup failed. A subsequent lookup with different
2963 options therefore always gave the same answer, though there were cases
2964 where it should not have. (Example: a "domains = !$mx_any" option on a
2965 dnslookup router: the "domains" option is always processed without any
2966 widening, but the router might have qualify_single set.) Now Exim uses the
2967 cached value only when the same options are set.
2969 32. Added John Jetmore's "exipick" utility to the distribution.
2971 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
2972 than a timeout (e.g. a certificate is required, and is not provided), an
2973 Exim server now closes the connection immediately. Previously it waited for
2974 the client to close - but if the client is SSL, it seems that they each
2975 wait for each other, leading to a delay before one of them times out.
2977 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
2978 maintained 0.8.x compatibility because I don't think many are using it, and
2979 it is clearly obsolete.
2981 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
2984 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
2985 client certificate was expired. A simple patch fixes this, though I don't
2986 understand the full logic of why the verify callback is called multiple
2989 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
2990 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
2991 which causes problems with some clients (such as the Certicom SSL Plus
2992 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
2993 disables the coutermeasure allowing Eudora to connect."
2995 38. Exim was not checking that a write() to a log file succeeded. This could
2996 lead to Bad Things if a log got too big, in particular if it hit a file
2997 size limit. Exim now panics and dies if it cannot write to a log file, just
2998 as it does if it cannot open a log file.
3000 39. Modified OS/Makefile-Linux so that it now contains
3002 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3004 The two -D definitions ensure that Exim is compiled with large file
3005 support, which makes it possible to handle log files that are bigger than
3008 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3009 instance) a domain was checked against a named list that involved a lookup,
3010 causing $domain_data to be set, then another domain was checked against the
3011 same list, then the first domain was re-checked, the value of $domain_data
3012 after the final check could be wrong. In particular, if the second check
3013 failed, it could be set empty. This bug probably also applied to
3016 41. The strip_trailing_dot option was not being applied to the address given
3017 with the -f command-line option.
3019 42. The code for reading a message's header from the spool was incrementing
3020 $received_count, but never initializing it. This meant that the value was
3021 incorrect (doubled) while delivering a message in the same process in which
3022 it was received. In the most common configuration of Exim, this never
3023 happens - a fresh exec is done - but it can happen when
3024 deliver_drop_privilege is set.
3026 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3027 it now includes up to 150 characters of the unexpected data in the log
3030 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3031 and building data strings. The size of both of these buffers was 10 000
3032 bytes - far larger than anybody would *ever* want, thought I. Needless to
3033 say, somebody hit the limit. I have increased the maximum line length to
3034 20 000 and the maximum data length of concatenated lines to 100 000. I have
3035 also fixed two bugs, because there was no checking on these buffers. Tsk,
3036 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3037 buffer is too small.
3039 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3040 lsearch lookups. Now it does.
3042 46. When parsing a route_list item in a manualroute router, a fixed-length
3043 buffer was used for the list of hosts. I made this 1024 bytes long,
3044 thinking that nobody would ever have a list of hosts that long. Wrong.
3045 Somebody had a whole pile of complicated expansion conditions, and the
3046 string was silently truncated, leading to an expansion error. It turns out
3047 that it is easier to change to an unlimited length (owing to other changes
3048 that have happened since this code was originally written) than to build
3049 structure for giving a limitation error. The length of the item that
3050 expands into the list of hosts is now unlimited.
3052 47. The lsearch lookup could not handle data where the length of text line was
3053 more than 4095 characters. Such lines were truncated, leading to shortened
3054 data being returned. It should now handle lines of any length.
3056 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3057 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3060 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3061 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3062 no longer happens in comment lines. A list of replacements is now placed
3063 at the head of all of the source files, except those whose only change is
3064 to replace PERL_COMMAND in the very first #! line.
3066 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3067 messages on the queue, with a bottom-up merge sort, using code contributed
3068 by Michael Haardt. This should make operations like -bp somewhat faster on
3069 large queues. It won't affect queue runners, except when queue_run_in_order
3072 51. Installed eximstats 1.31 in the distribution.
3074 52. Added support for SRV lookups to the dnslookup router.
3076 53. If an ACL referred to $message_body or $message_body_end, the value was not
3077 reset for any messages that followed in the same SMTP session.
3079 54. The store-handling optimization for building very long strings was not
3080 differentiating between the different store pools. I don't think this
3081 actually made any difference in practice, but I've tidied it.
3083 55. While running the routers to verify a sender address, $sender_address
3084 was still set to the sender address. This is wrong, because when routing to
3085 send a bounce to the sender, it would be empty. Therefore, I have changed
3086 it so that, while verifying a sender address, $sender_address is set to <>.
3087 (There is no change to what happens when verifying a recipient address.)
3089 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3090 target A or AAAA records (if not already returned) without resetting the
3091 qualify_single or search_parents options of the DNS resolver. These are
3092 inappropriate in this case because the targets of MX and SRV records must
3093 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3094 target that, when qualified, matched something in the local domain. These
3095 two options are now turned off when doing these lookups.
3097 57. It seems that at least some releases of Reiserfs (which does not have the
3098 concept of a fixed number of inodes) returns zero and not -1 for the
3099 number of available inodes. This interacted badly with check_spool_inodes,
3100 which assumed that -1 was the "no such thing" setting. What I have done is
3101 to check that the total number of inodes is greater than zero before doing
3102 the test of how many are available.
3104 58. When a "warn" ACL statement has a log_message modifier, the message is
3105 remembered, and not repeated. This is to avoid a lot of repetition when a
3106 message has many recipients that cause the same warning to be written.
3107 Howewer, Exim was preserving the list of already written lines for an
3108 entire SMTP session, which doesn't seem right. The memory is now reset if a
3109 new message is started.
3111 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3112 debugging output unless log_rewrite was also set.
3114 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3115 of (int)(handle) when we know that handle contains (void *)(-1).
3117 61. The Exim daemon panic-logs an error return when it closes the incoming
3118 connection. However "connection reset by peer" seems to be common, and
3119 isn't really an error worthy of noting specially, so that particular error
3122 62. When Exim is trying to find all the local interfaces, it used to panic and
3123 die if the ioctl to get the interface flags failed. However, it seems that
3124 on at least one OS (Solaris 9) it is possible to have an interface that is
3125 included in the list of interfaces, but for which you get a failure error
3126 for this call. This happens when the interface is not "plumbed" into a
3127 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3128 failure of the "get flags" call assumes that the interface is down.
3130 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3131 makes life easier for people who are doing arithmetic on fields extracted
3132 from dates, where you often get leading zeros that should not be
3133 interpreted as octal.
3135 64. Added qualify_domain to the redirect router, to override the global
3138 65. If a pathologically long header line contained very many addresses (the
3139 report of this problem mentioned 10 000) and each of them was rewritten,
3140 Exim could use up a very large amount of memory. (It kept on making new
3141 copies of the header line as it rewrote, and never released the old ones.)
3142 At the expense of a bit more processing, the header rewriting function has
3143 been changed so that it no longer eats memory in this way.
3145 66. The generation of the Received: header has been moved from the time that a
3146 message starts to be received, to the time that it finishes. The timestamp
3147 in the Received: header should now be very close to that of the <= log
3148 line. There are two side-effects of this change:
3150 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3151 logged header lines no longer include the local Received: line, because
3152 it has not yet been created. The same applies to a copy of the message
3153 that is returned to a non-SMTP sender when a message is rejected.
3155 (b) When a filter file is tested using -bf, no additional Received: header
3156 is added to the test message. After some thought, I decided that this
3159 This change does not affect the value of $received_for. It is still set
3160 after address rewriting, but before local_scan() is called.
3162 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3164 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3165 gave an unhelpful panic error message, and a defer error. I have managed to
3166 change this behaviour so that it now rejects any supplied certificate,
3167 which seems right, as the list of acceptable certificates is empty.
3169 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3170 gave an unhelpful defer error. I have not managed to make this reject any
3171 supplied certificates, but the error message it gives is "no certificate
3172 supplied", which is not helpful.
3174 70. exigrep's output now also includes lines that are not associated with any
3175 message, but which match the given pattern. Implemented by a patch from
3176 Martin Sluka, which also tidied up the Perl a bit.
3178 71. Recipient callout verification, like sender verification, was using <> in
3179 the MAIL FROM command. This isn't really the right thing, since the actual
3180 sender may affect whether the remote host accepts the recipient or not. I
3181 have changed it to use the actual sender in the callout; this means that
3182 the cache record is now keyed on a recipient/sender pair, not just the
3183 recipient address. There doesn't seem to be a real danger of callout loops,
3184 since a callout by the remote host to check the sender would use <>.
3185 [SEE ABOVE: changed after hitting problems.]
3187 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3188 temporary errors. However, in the case of such a code being given after
3189 the end of a data transmission (i.e. after ".") Exim was failing to write
3190 a retry record for the message. (Yes, there was some broken host that was
3191 actually sending 8xx at this point.)
3193 73. An unknown lookup type in a host list could cause Exim to panic-die when
3194 the list was checked. (An example that provoked this was putting <; in the
3195 middle of a list instead of at the start.) If this happened during a DATA
3196 ACL check, a -D file could be left lying around. This kind of configuration
3197 error no longer causes Exim to die; instead it causes a defer errror. The
3198 incident is still logged to the main and panic logs.
3200 74. Buglet left over from Exim 3 conversion. The message "too many messages
3201 in one connection" was written to the rejectlog but not the mainlog, except
3202 when address rewriting (yes!) was being logged.
3204 75. Added write_rejectlog option.
3206 76. When a system filter was run not as root (that is, when system_filter_user
3207 was set), the values of the $n variables were not being returned to the
3208 main process; thus, they were not subsequently available in the $sn
3211 77. Added +return_path_on_delivery log selector.
3213 78. A connection timeout was being treated differently from recipients deferred
3214 when testing hosts_max_try with a message that was older than the host's
3215 retry timeout. (The host should not be counted, thus allowing all hosts to
3216 be tried at least once before bouncing.) This may have been the cause of an
3217 occasionally reported bug whereby a message would remain on the queue
3218 longer than the retry timeout, but would be bounced if a delivery was
3219 forced. I say "may" because I never totally pinned down the problem;
3220 setting up timeout/retry tests is difficult. See also the next item.
3222 79. The ultimate address timeout was not being applied to errors that involved
3223 a combination of host plus message (for example, a timeout on a MAIL
3224 command). When an address resolved to a number of possible hosts, and they
3225 were not all tried for each delivery (e.g. because of hosts_max_try), a
3226 message could remain on the queue longer than the retry timeout.
3228 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3231 81. Fixed an obscure SMTP outgoing bug which required at least the following
3232 conditions: (a) there was another message waiting for the same server;
3233 (b) the server returned 5xx to all RCPT commands in the first message so
3234 that the message was not completed; (c) the server dropped the connection
3235 or gave a negative response to the RSET that Exim sends to abort the
3236 transaction. The observed case was a dropped connection after DATA that had
3237 been sent in pipelining mode. That is, the server had advertised PIPELINING
3238 but was not implementing it correctly. The effect of the bug was incorrect
3239 behaviour, such as trying another host, and this could lead to a crash.
3245 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3246 and daemon.c were passed as pointers to ints; they should have been
3247 pointers to socklen_t variables (which are typically unsigned ints).
3249 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3252 3. Fixed a really odd bug that affected only the testing scheme; patching a
3253 certain fixed string in the binary changed the value of another string that
3254 happened to be identical to the end of the original first string.
3256 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3257 name", it returns that address as the IP address. On some operating
3258 systems (e.g. Solaris), it also passes back the IP address string as the
3259 "host name". However, on others (e.g. Linux), it passes back an empty
3260 string. Exim wasn't checking for this, and was changing the host name to an
3261 empty string, assuming it had been canonicized.
3263 5. Although rare, it is permitted to have more than one PTR record for a given
3264 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3265 all the names associated with an address, because they do in Solaris.
3266 However, it seems that they do not in Linux for data that comes from the
3267 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3268 I found this out when I moved to a new Linux workstation and tried to run
3269 the Exim test suite.
3271 To get round this problem I have changed the code so that it now does its
3272 own call to the DNS to look up PTR records when searching for a host name.
3273 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3274 addresses that are only in /etc/hosts are still found.
3276 This behaviour is, however, controlled by an option called host_lookup_
3277 order, which defaults to "bydns:byaddr". If people want to use the other
3278 order, or indeed, just use one or the other means of lookup, they can
3279 specify it in this variable.
3281 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3282 some operating systems, this comes back from gethostbyaddr() as an empty
3283 string, and this is what Exim used to test for. However, it seems that in
3284 other systems, "." is yielded. Exim now tests for this case too.
3286 7. The values of check_spool_space and check_log_space are now held internally
3287 as a number of kilobytes instead of an absolute number of bytes. If a
3288 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3289 kilobyte. This means that much larger values can be stored.
3291 8. Exim monitor: an attempt to get the action menu when not actually pointing
3292 at a message produces an empty menu entitled "No message selected". This
3293 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3294 no entries in it ("Shell widget menu has zero width and/or height"). So I
3295 have added a single, blank menu entry in this case.
3297 9. Added ${quote_local_part.
3299 10. MIME decoding is now applied to the contents of Subject: header lines when
3302 11. Now that a reference to $sender_host_address automatically causes a reverse
3303 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3304 host lookup before query-style lookups in lists that might use this
3305 variable. This has therefore been abolished, and the "net-" prefix is no
3306 longer necessary for query-style lookups.
3308 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3309 have been a typo for LFLAGS, so it has been changed.
3311 13. The install script calls Exim with "-C /dev/null" in order to find the
3312 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3313 to be output. Howeve, since Exim outputs its version number before the
3314 error, it didn't break the script. It just looked ugly. I fixed this by
3315 always allowing "-C /dev/null" if the caller is root.
3317 14. Ignore overlarge ACL variable number when reading spool file - insurance
3318 against a later release with more variables having written the file.
3320 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3321 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3322 this, as well as the form without the "IPv6" on the front.
3324 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3325 OS/Makefile-Darwin file.
3327 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3328 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3330 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3331 variable. However, some OS use ints for inodes. I've added cast to long int
3332 to get rid of the compiler warning.
3334 19. I had forgotten to lock out "/../" in configuration file names when
3335 ALT_CONFIG_PREFIX was set.
3337 20. Routers used for verification do not need to specify transports. However,
3338 if such a router generated a host list, and callout was configured, Exim
3339 crashed, because it could not find a port number from the (non-existent)
3340 transport. It now assumes port 25 in this circumstance.
3342 21. Added the -t option to exigrep.
3344 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3345 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3346 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3347 them is set, LOOKUP_LSEARCH is forced.
3349 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3351 24. Added sender and host information to the "rejected by local_scan()" log
3352 line; previously there was no indication of these.
3354 25. Added .include_if_exists.
3356 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3357 Linux. It turns out that not all file systems support this. Apparently some
3358 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3359 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3360 not supported on the file descriptor, is now ignored when Exim is trying to
3361 sync a directory. This applies only to Linux.
3363 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3365 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3366 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3369 29. Added support for maildirsize files from supplied patch (modified a bit).
3371 30. The use of :fail: followed by an empty string could lead Exim to respond to
3372 sender verification failures with (e.g.):
3374 550 Verification failed for <xxx>
3375 550 Sender verify failed
3377 where the first response line was missing the '-' that indicates it is not
3378 the final line of the response.
3380 31. The loop for finding the name of the user that called Exim had a hardwired
3381 limit of 10; it now uses the value of finduser_retries, which is used for
3382 all other user lookups.
3384 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3387 33. Exim was neglecting to zero errno before one call of strtol() when
3388 expanding a string and expecting an integer value. On some systems this
3389 resulted in spurious "integer overflow" errors. Also, it was casting the
3390 result into an int without checking.
3392 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3393 did not work. The code looks as if it has *never* worked, though it appears
3394 to have been documented since at least releast 1.62. I have made it work.
3396 35. The "timeout_DNS" error in retry rules, also documented since at least
3397 1.62, also never worked. As it isn't clear exactly what this means, and
3398 clearly it isn't a major issue, I have abolished the feature by treating it
3399 as "timeout", and writing a warning to the main and panic logs.
3401 36. The display of retry rules for -brt wasn't always showing the error code
3404 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3405 timeout_connect_A, timeout_connect_MX.
3407 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3408 to the empty sender.
3410 39. The daemon was not analysing the content of -oX till after it had closed
3411 stderr and disconnected from the controlling terminal. This meant that any
3412 syntax errors were only noted on the panic log, and the return code from
3413 the command was 0. By re-arranging the code a little, I've made the
3414 decoding happen first, so such errors now appear on stderr, and the return
3415 code is 1. However, the actual setting up of the sockets still happens in
3416 the disconnected process, so errors there are still only recorded on the
3419 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3420 connections (as happens on some IP stacks) was logged at start up time as
3421 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3422 it from "IPv6 and IPv4", which means that two separate sockets are being
3425 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3426 says whether AF_INET or AF_INET6 was passed as an argument.
3428 42. Exiwhat output was messed up when time zones were included in log
3431 43. Exiwhat now gives more information about the daemon's listening ports,
3432 and whether -tls-on-connect was used.
3434 44. The "port" option of the smtp transport is now expanded.
3436 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3437 silently ignored. Now an error message is written to the main and panic
3440 46. There's a new ACL modifier called "logwrite" which writes to a log file
3441 as soon as it is encountered.
3443 47. Added $local_user_uid and $local_user_gid at routing time.
3445 48. Exim crashed when trying to verify a sender address that was being
3448 49. Exim was recognizing only a space character after ".include". It now also
3449 recognizes a tab character.
3451 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3452 extracting the relevant information from the specification. The man page no
3453 longer contains scrambled data for the -d option, and I've added a section
3454 at the front about calling Exim under different names.
3456 51. Added "extra_headers" argument to the "mail" command in filter files.
3458 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3461 53. Installed eximstats 1.29.
3463 54. Added transport_filter_timeout as a generic transport option.
3465 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3466 Cc: header lines. This was required by RFC 822, but it not required by RFC
3469 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3470 incoming messages that did not have them. Now it does so only if the
3471 message originates locally, that is, if there is no associated remote host
3472 address. When Resent- header lines are present, this applies to the Resent-
3473 lines rather than the non-Resent- lines.
3475 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3476 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3478 58. Messages for configuration errors now include the name of the main
3479 configuration file - useful now that there may be more than one file in a
3480 list (.included file names were always shown).
3482 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3483 for those rare installations that do not start the daemon as root or run it
3484 setuid root. I've cut out the call to initgroups() if the daemon is not
3487 60. The Exim user and group can now be bound into the binary as text strings
3488 that are looked up at the start of Exim's processing.
3490 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3492 62. Added $mailstore_basename variable.
3494 63. Installed patch to sieve.c from Michael Haardt.
3496 64. When Exim failed to open the panic log after failing to open the main log,
3497 the original message it was trying to log was written to stderr and debug
3498 output, but if they were not available (the usual case in production), it
3499 was lost. Now it is written to syslog before the two lines that record the
3500 failures to open the logs.
3502 65. Users' Exim filters run in subprocesses under the user's uid. It is
3503 possible for a "deliver" command or an alias in a "personal" command to
3504 provoke an address rewrite. If logging of address rewriting is configured,
3505 this fails because the process is not running as root or exim. There may be
3506 a better way of dealing with this, but for the moment (because 4.30 needs
3507 to be released), I have disabled address rewrite logging when running a
3508 filter in a non-root, non-exim process.
3514 1. The buildconfig auxiliary program wasn't quoting the value set for
3515 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3516 not defined. This bug was masked in 4.22 by the effect that was fixed in
3519 2. Some messages that were rejected after a message id was allocated were
3520 shown as "incomplete" by exigrep. It no longer does this for messages that
3521 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3523 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3524 have allow_domain_literals set, the ID did not get logged in the <= line.
3525 Domain literals are now always recognized in Message-ID: header lines.
3527 4. The first argument for a ${extract expansion item is the key name or field
3528 number. Leading and trailing spaces in this item were not being ignored,
3529 causing some misleading effects.
3531 5. When deliver_drop_privilege was set, single queue runner processes started
3532 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3533 same command in the process it spins off) were not dropping privilege.
3535 6. When the daemon running as "exim" started a queue runner, it always
3536 re-executed Exim in the spun-off process. This is a waste of effort when
3537 deliver_drop_privilege is set. The new process now just calls the
3538 queue-runner function directly.
3544 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3547 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3548 ignored. Though the use of -oP was forcing the writing of a pid file, it
3549 was always written to the default place.
3551 3. If the message "no IP address found for host xxxx" is generated during
3552 incoming verification, it is now followed by identification of the incoming
3553 connection (so you can more easily find what provoked it).
3555 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
3557 5. Added some features to "harden" Exim a bit more against certain attacks:
3559 (a) There is now a build-time option called FIXED_NEVER_USERS that can
3560 be put in Local/Makefile. This is like the never_users runtime option,
3561 but it cannot be overridden. The default setting is "root".
3563 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
3564 prefix string with which any file named in a -C command line option
3567 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
3568 is retained for -C and -D only if the caller of Exim is root. Without
3569 it, the exim user may also use -C and -D and retain privilege.
3571 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
3572 command line option is disabled.
3574 6. Macro names set by the -D option must start with an upper case letter, just
3575 like macro names defined in the configuration file.
3577 7. Added "dereference=" facility to LDAP.
3579 8. Two instances of the typo "uknown" in the source files are fixed.
3581 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
3582 the Configure-Makefile script screwed up while processing it.
3584 10. Incorporated PCRE 4.4.
3586 11. The SMTP synchronization check was not operating right at the start of an
3587 SMTP session. For example, it could not catch a HELO sent before the client
3588 waited for the greeting. There is now a check for outstanding input at the
3589 point when the greeting is written. Because of the duplex, asynchronous
3590 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
3591 way, but not yet received, when the check is performed.
3593 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
3594 on TCP/IP sockets, because this apparently causes some broken clients to
3597 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
3598 unchanged) from the Cygwin maintainer.
3600 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
3601 was supported instead of testing for mbx. Effectively a typo.
3603 15. The spa authenticator server code was not checking that the input it
3604 received was valid base64.
3606 16. The debug output line for the "set" modifier in ACLs was not showing the
3607 name of the variable that was being set.
3609 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
3611 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
3612 lookup on its own. Something else was needed to trigger the lookup. For
3613 example, a match in host_lookup or the need for a host name in a host list.
3614 Now, if $sender_host_name is referenced and the host name has not yet been
3615 looked up, a lookup is performed. If the lookup fails, the variable remains
3616 empty, and $host_lookup_failed is set to "1".
3618 19. Added "eqi" as a case-independent comparison operator.
3620 20. The saslauthd authentication condition could segfault if neither service
3621 nor realm was specified.
3623 21. If an overflowing value such as "2048M" was set for message_size_limit, the
3624 error message that was logged was misleading, and incoming SMTP
3625 connections were dropped. The message is now more accurate, and temporary
3626 errors are given to SMTP connections.
3628 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
3629 (except RSET) with a 421 error, until QUIT is received. However, it was
3630 failing to send a response to QUIT.
3632 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
3633 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
3634 ACL is now run after the helo_try_verify_hosts code.
3636 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
3637 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
3638 case-independent, so other case variants are also recognized). Apparently
3639 some systems use these upper case variants.
3641 25. If more than two messages were waiting for the same host, and a transport
3642 filter was specified for the transport, Exim sent two messages over the
3643 same TCP/IP connection, and then failed with "socket operation on non-
3644 socket" when it tried to send the third.
3646 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
3648 27. The extern definition of crypt16() in expand.c was not being excluded when
3649 the OS had its own crypt16() function.
3651 28. Added bounce_return_body as a new option, and bounce_return_size_limit
3652 as a preferred synonym for return_size_limit, both as an option and as an
3655 29. Added LIBS=-liconv to OS/Makefile-OSF1.
3657 30. Changed the default configuration ACL to relax the local part checking rule
3658 for addresses that are not in any local domains. For these addresses,
3659 slashes and pipe symbols are allowed within local parts, but the sequence
3660 /../ is explicitly forbidden.
3662 31. SPA server authentication was not clearing the challenge buffer before
3665 32. log_message in a "warn" ACL statement was writing to the reject log as
3666 well as to the main log, which contradicts the documentation and doesn't
3667 seem right (because no rejection is happening). So I have stopped it.
3669 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
3670 However, I am unable to do any testing of this.
3672 34. Fixed an infelicity in the appendfile transport. When checking directories
3673 for a mailbox, to see if any needed to be created, it was accidentally
3674 using path names with one or more superfluous leading slashes; tracing
3675 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
3677 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
3678 contents are added to the log line that is written for every discarded
3679 recipient. (Previously a log_message setting was ignored.)
3681 36. The ${quote: operator now quotes the string if it is empty.
3683 37. The install script runs exim in order to find its version number. If for
3684 some reason other than non-existence or emptiness, which it checks, it
3685 could not run './exim', it was installing it with an empty version number,
3686 i.e. as "exim-". This error state is now caught, and the installation is
3689 38. An argument was missing from the function that creates an error message
3690 when Exim fails to connect to the socket for saslauthd authentication.
3691 This could cause Exim to crash, or give a corrupted message.
3693 39. Added isip, isip4, and isip6 to ${if conditions.
3695 40. The ACL variables $acl_xx are now saved with the message, and can be
3696 accessed later in routers, transports, and filters.
3698 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
3699 strings in the file are not string-expanded.
3701 42. If a MAIL command specified a SIZE value that was too large to fit into an
3702 int variable, the check against message_size_limit failed. Such values are
3703 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
3704 day this will have to be increased, but I don't think I want to be around
3705 when emails are that large.
3712 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
3713 iconv() is not standard in FreeBSD.
3715 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
3716 IPv6 enabled. The observed symptom was a segmentation fault on return from
3717 the function os_common_find_running_interfaces() in src/os.c.
3719 3. In the check_special_case() function in daemon.c I had used "errno" as an
3720 argument name, which causes warnings on some systems. This was basically a
3721 typo, since it was named "eno" in the comments!
3723 4. The code that waits for the clock to tick (at a resolution of some fraction
3724 of a second) so as to ensure message-id uniqueness was always waiting for
3725 at least one whole tick, when it could have waited for less. [This is
3726 almost certainly not relevant at current processor speeds, where it is
3727 unlikely to ever wait at all. But we try to future-proof.]
3729 5. The function that sleeps for a time interval that includes fractions of a
3730 second contained a race. It did not block SIGALRM between setting the
3731 timer, and suspending (a couple of lines later). If the interval was short
3732 and the sigsuspend() was delayed until after it had expired, the suspension
3733 never ended. On busy systems this could lead to processes getting stuck for
3736 6. Some uncommon configurations may cause a lookup to happen in a queue runner
3737 process, before it forks any delivery processes. The open lookup caching
3738 mechanism meant that the open file or database connection was passed into
3739 the delivery process. The problem was that delivery processes always tidy
3740 up cached lookup data. This could cause a problem for the next delivery
3741 process started by the queue runner, because the external queue runner
3742 process does not know about the closure. So the next delivery process
3743 still has data in the lookup cache. In the case of a file lookup, there was
3744 no problem because closing a file descriptor in a subprocess doesn't affect
3745 the parent. However, if the lookup was caching a connection to a database,
3746 the connection was closed, and the second delivery process was likely to
3747 see errors such as "PGSQL: query failed: server closed the connection
3748 unexpectedly". The problem has been fixed by closing all cached lookups
3749 in a queue runner before running a delivery process.
3751 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
3752 seem to have the "const" qualifier which it has on other OS. I've
3755 8. Change 4.21/2 was too strict. It is only if there are two authenticators
3756 *of the same type* (client or server) with the same public name that an
3757 error should be diagnosed.
3759 9. When Exim looked up a host name for an IP address, but failed to find the
3760 original IP address when looking up the host name (a safety check), it
3761 output the message "<ip address> does not match any IP for NULL", which was
3762 confusing, to say the least. The bug was that the host name should have
3763 appeared instead of "NULL".
3765 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
3766 user that is built into the binary, and the -C or -D options is used, root
3767 privilege is dropped before the configuration file is read. In addition,
3768 logging is switched to stderr instead of the normal log files. If the
3769 configuration then re-defines the Exim user, the unprivileged environment
3770 is probably not what is expected, so Exim logs a panic warning message (but
3773 However, if deliver_drop_privilege is set, the unprivileged state may well
3774 be exactly what is intended, so the warning has been cut out in that case,
3775 and Exim is allowed to try to write to its normal log files.
3781 1. smtp_return_error_details was not giving details for temporary sender
3782 or receiver verification errors.
3784 2. Diagnose a configuration error if two authenticators have the same public
3787 3. Exim used not to create the message log file for a message until the first
3788 delivery attempt. This could be confusing when incoming messages were held
3789 for policy or load reasons. The message log file is now created at the time
3790 the message is received, and an initial "Received" line is written to it.
3792 4. The automatically generated man page for command line options had a minor
3793 bug that caused no ill effects; however, a more serious problem was that
3794 the procedure for building the man page automatically didn't always
3795 operate. Consequently, release 4.20 contains an out-of-date version. This
3796 shouldn't happen again.
3798 5. When building Exim with embedded Perl support, the script that builds the
3799 Makefile was calling 'perl' to find its compile-time parameters, ignoring
3800 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
3802 6. The freeze_tell option was not being used for messages that were frozen on
3803 arrival, either by an ACL or by local_scan().
3805 7. Added the smtp_incomplete_transaction log selector.
3807 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
3808 was accepting AUTH without a new EHLO.
3810 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
3811 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
3814 10. Logging of TCP/IP connections (when configured) now happens in the main
3815 daemon process instead of the child process, so that the TCP/IP connection
3816 count is more accurate (but it can never be perfect).
3818 11. The use of "drop" in a nested ACL was not being handled correctly in the
3819 outer ACL. Now, if condition failure induced by the nested "drop" causes
3820 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
3821 or "require"), the connection is dropped.
3823 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
3824 that yield "discard" can now be used with an "accept" or a "discard" verb,
3825 but an error is generated for any others (because I can't see a useful way
3826 to define what should happen).
3828 13. When an ACL is read dynamically from a file (or anywhere else), the lines
3829 are now processed in the same way as lines in the Exim configuration file.
3830 In particular, continuation lines are supported.
3832 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
3834 15. Added -ti meaning -t -i.
3836 16. Check for letters, digits, hyphens, and dots in the names of dnslist
3837 domains, and warn by logging if others are found.
3839 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
3840 returned from GIFCONF when Exim is trying to find the list of interfaces on
3841 a host. The code in os.c has been modified to copy each ifreq to an aligned
3842 structure in all cases.
3844 Also, in some cases, the returned ifreq's were being copied to a 'struct
3845 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
3846 means the last couple of bytes of an IPv6 address could be chopped if the
3847 ifreq contained only a normal sockaddr (14 bytes storage).
3849 18. Named domain lists were not supported in the hosts_treat_as_local option.
3850 An entry such as +xxxx was not recognized, and was treated as a literal
3853 19. Ensure that header lines added by a DATA ACL are included in the reject log
3854 if the ACL subsequently rejects the message.
3856 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
3857 MD5 (which is deprecated).
3859 21. When testing a filter file using -bf, Exim was writing a message when it
3860 took the sender from a "From " line in the message, but it was not doing so
3861 when it took $return_path from a Return-Path: header line. It now does.
3863 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
3864 with a valid header line field name (a series of printing characters
3865 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
3867 23. Changed "disc" in the source to "disk" to conform to the documentation and
3868 the book and for uniformity.
3870 24. Ignore Sendmail's -Ooption=value command line item.
3872 25. When execve() failed while trying to run a command in a pipe transport,
3873 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
3874 could be confused with a return value of 69 from the command itself. This
3875 has been changed to 127, the value the shell returns if it is asked to run
3876 a non-existent command. The wording for the related log line suggests a
3877 non-existent command as the problem.
3879 26. If received_header_text expands to an empty string, do not add a Received:
3880 header line to the message. (Well, it adds a token one on the spool, but
3881 marks it "old" so that it doesn't get used or transmitted.)
3883 27. Installed eximstats 1.28 (addition of -nt option).
3885 28. There was no check for failure on the call to getsockname() in the daemon
3886 code. This can fail if there is a shortage of resources on the system, with
3887 ENOMEM, for example. A temporary error is now given on failure.
3889 29. Contrary to the C standard, it seems that in some environments, the
3890 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
3891 program. Exim now does this explicitly; it affects the formatting of
3892 timestamps using strftime().
3894 30. If exiqsumm was given junk data, it threw up some uninitialized variable
3895 complaints. I've now initialized all the variables, to avoid this.
3897 32. Header lines added by a system filter were not being "seen" during
3898 transport-time rewrites.
3900 33. The info_callback() function passed to OpenSSL is set up with type void
3901 (*)(SSL *, int, int), as described somewhere. However, when calling the
3902 function (actually a macro) that sets it up, the type void(*)() is
3903 expected. I've put in a cast to prevent warnings from picky compilers.
3905 34. If a DNS black list lookup found a CNAME record, but there were no A
3906 records associated with the domain it pointed at, Exim crashed.
3908 35. If a DNS black list lookup returned more than one A record, Exim ignored
3909 all but the first. It now scans all returned addresses if a particular IP
3910 value is being sought. In this situation, the contents of the
3911 $dnslist_value variable are a list of all the addresses, separated by a
3914 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
3915 to accept a host name and all its aliases if the forward lookup for any of
3916 them yielded the IP address of the incoming connection. Now it accepts only
3917 those names whose forward lookup yields the correct IP address. Any other
3918 names are discarded. This closes a loophole whereby a rogue DNS
3919 administrator could create reverse DNS records to break through a
3920 wildcarded host restriction in an ACL.
3922 37. If a user filter or a system filter that ran in a subprocess used any of
3923 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
3924 the wrong values were passed to the pipe command ($thisaddress had the
3925 value of $0, $0 had the value of $1, etc). This bug was introduced by
3926 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
3928 38. Improved the line breaking for long SMTP error messages from ACLs.
3929 Previously, if there was no break point between 40 and 75 characters, Exim
3930 left the rest of the message alone. Two changes have been made: (a) I've
3931 reduced the minimum length to 35 characters; (b) if it can't find a break
3932 point between 35 and 75 characters, it looks ahead and uses the first one
3933 that it finds. This may give the occasional overlong line, but at least the
3934 remaining text gets split now.
3936 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
3937 file descriptors might be low, and that setting 1000 would always raise it.
3938 It turns out that in some environments, the limit is already over 1000 and
3939 that lowering it causes trouble. So now Exim takes care not to decrease it.
3941 40. When delivering a message, the value of $return_path is set to $sender_
3942 address at the start of routing (routers may change the value). By an
3943 oversight, this default was not being set up when an address was tested by
3944 -bt or -bv, which affected the outcome if any router or filter referred to
3947 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
3948 log only when "message" or "log_message" are set. However, if one of the
3949 conditions was an address verification, or a call to a nested ACL, the
3950 messages generated by the underlying test were being passed through. This
3951 no longer happens. The underlying message is available in $acl_verify_
3952 message for both "message" and "log_message" expansions, so it can be
3953 passed through if needed.
3955 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
3956 new expansion $bh_ to give the encoded byte string without charset
3957 translation. Translation happens only if iconv() is available; HAVE_ICONV
3958 indicates this at build time. HEADERS_CHARSET gives the charset to
3959 translate to; headers_charset can change it in the configuration, and
3960 "headers charset" can change it in an individual filter file.
3962 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
3963 that creates RFC 2047 encoded "words" labels them as that charset instead
3964 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
3965 expansion operator; (ii) when Exim creates a From: line for a local
3966 message; (iii) when a header line is rewritten to include a "phrase" part.
3968 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
3969 buggy, causing it to skip the first lines of messages whose message ID
3970 ended in 'D'. This would not have bitten before Exim release 4.14, because
3971 message IDs were unlikely to end in 'D' before then. The effect was to have
3972 incorrect size information for certain domains.
3974 45. #include "config.h" was missing at the start of the crypt16.c module. This
3975 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
3978 46. If there was a timeout during a "random" callout check, Exim treated it as
3979 a failure of the random address, and carried on sending RSET and the real
3980 address. If the delay was just some slowness somewhere, the response to the
3981 original RCPT would be taken as a response to RSET and so on, causing
3982 mayhem of various kinds.
3984 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
3985 when I implemented it. It didn't allow for the fact that some option values
3986 may legitimatetly be negative (e.g. size_addition), and it didn't even do
3987 the right test for positive values.
3989 48. Domain names in DNS records are case-independent. Exim always looks them up
3990 in lower case. Some resolvers return domain names in exactly the case they
3991 appear in the zone file, that is, they may contain uppercase letters. Not
3992 all resolvers do this - some return always lower case. Exim was treating a
3993 change of case by a resolver as a change of domain, similar to a widening
3994 of a domain abbreviation. This triggered its re-routing code and so it was
3995 trying to route what was effectively the same domain again. This normally
3996 caused routing to fail (because the router wouldn't handle the domain
3997 twice). Now Exim checks for this case specially, and just changes the
3998 casing of the domain that it ultimately uses when it transmits the message
4001 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4004 50. If a filter generated a file delivery with a non-absolute name (possible if
4005 no home directory exists for the router), the forbid_file option was not
4008 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4009 the existing equality matching.
4011 52. Exim was using ints instead of ino_t variables in some places where it was
4012 dealing with inode numbers.
4014 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4015 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4016 called TMPDIR, and if it finds it is different, it changes its value.
4018 54. The smtp_printf() function is now made available to local_scan() so
4019 additional output lines can be written before returning. There is also an
4020 smtp_fflush() function to enable the detection of a dropped connection.
4021 The variables smtp_input and smtp_batched_input are exported to
4024 55. Changed the default runtime configuration: the message "Unknown user"
4025 has been removed from the ACL, and instead placed on the localuser router,
4026 using the cannot_route_message feature. This means that any verification
4027 failures that generate their own messages won't get overridden. Similarly,
4028 the "Unrouteable address" message that was in the ACL for unverifiable
4029 relay addresses has also been removed.
4031 56. Added hosts_avoid_esmtp to the smtp transport.
4033 57. The exicyclog script was not checking for the esoteric option
4034 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4035 will work only if exicyclog is run under the appropriate euid.
4037 58. Following a discussion on the list, the rules by which Exim recognises line
4038 endings on incoming messages have been changed. The -dropcr and drop_cr
4039 options are now no-ops, retained only for backwards compatibility. The
4040 following line terminators are recognized: LF CRLF CR. However, special
4041 processing applies to CR:
4043 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4044 nor a local message in the state where . is a terminator.
4046 (ii) If a bare CR is encountered in a header line, an extra space is added
4047 after the line terminator so as not to end the header. The reasoning
4048 behind this is that bare CRs in header lines are most likely either
4049 to be mistakes, or people trying to play silly games.
4051 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4052 was being incorrectly given as 18 bytes larger than it should have been.
4053 This is a VOB (very old bug).
4055 60. This may never have affected anything current, but just in case it has:
4056 When the local host is found other than at the start of a list of hosts,
4057 the local host, those with the same MX, and any that follow, are discarded.
4058 When the list in question was part of a longer list of hosts, the following
4059 hosts (not currently being processed) were also being discarded. This no
4060 longer happens. I'm not sure if this situation could ever has previously
4063 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4066 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4067 header. This is something that is recommended in a new Internet Draft, and
4068 is something that is documented as being done by Sendmail. There are two
4069 possible values. For messages generated by the autoreply transport, Exim
4072 Auto-Submitted: auto-replied
4074 whereas for all other generated messages (e.g. bounces) it adds
4076 Auto-Submitted: auto-generated
4078 63. The "personal" condition in filters now includes a test for the
4079 Auto-Submitted: header. If it contains the string "auto-" the message it
4080 not considered personal.
4082 64. Added rcpt_include_affixes as a generic transport option.
4084 65. Added queue_only_override (default true).
4086 66. Added the syslog_duplication option.
4088 67. If what should have been the first header line of a message consisted of
4089 a space followed by a colon, Exim was mis-interpreting it as a header line.
4090 It isn't of course - it is syntactically invalid and should therefore be
4091 treated as the start of the message body. The misbehaviour could have
4092 caused a number of strange effects, including loss of data in subsequent
4093 header lines, and spool format errors.
4095 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4096 client host had authenticated. This control can now be exercised by an ACL
4097 for more flexibility.
4099 69. By default, callouts do not happen when testing with -bh. There is now a
4100 variant, -bhc, which does actually run the callout code, including
4101 consulting and updating the callout cache.
4103 70. Added support for saslauthd authentication, courtesy of Alexander
4106 71. If statvfs() failed on the spool or log directories while checking their
4107 size for availability, Exim confusingly gave the error "space shortage".
4108 Furthermore, in debugging mode it crashed with a floating point exception.
4109 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4110 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4111 serious problem, Exim now writes to the main and panic logs when this
4112 happens, with details of the failure. It then refuses to accept the
4113 incoming message, giving the message "spool directory problem" or "log
4114 directory problem" with a 421 code for SMTP messages.
4116 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4117 0, 1, and 2 exist, because some OS complain for execs without them (see
4118 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4119 descriptors. However, the code omitted to check that the open succeeded,
4120 causing mysterious errors if for some reason the permissions on /dev/null
4121 got screwed. Now Exim writes a message to the main and panic logs, and
4122 bombs out if it can't open /dev/null.
4124 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4125 interact so that it is all more flexible. It is supposed to remain
4126 backwards compatible. Also added extra_local_interfaces.
4128 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4129 to bomb out with an assertion failure - to the client this appears as a
4130 connection drop. This problem occurs in the part of the code that was taken
4131 from the Samba project. Fortunately, the assertion is in a very simple
4132 function, so I have fixed this by reproducing the function inline in the
4133 one place where it is called, and arranging for authentication to fail
4134 instead of killing the process with assert().
4136 75. The SPA client code was not working when the server requested OEM rather
4137 than Unicode encoding.
4139 76. Added code to make require_files with a specific uid setting more usable in
4140 the case where statting the file as root fails - usually a non-root-mounted
4141 NFS file system. When this happens and the failure is EACCES, Exim now
4142 forks a subprocess and does the per-uid checking as the relevant uid.
4144 77. Added process_log_path.
4146 78. If log_file_path was not explicitly set, a setting of check_log_space or
4147 check_log_inodes was ignored.
4149 79. If a space check for the spool or log partitions fails, the incident is now
4150 logged. Of course, in the latter case the data may get lost...
4152 80. Added the %p formatting code to string_format() so that it can be used to
4153 print addresses in debug_print(). Adjusted all the address printing in the
4154 debugging in store.c to use %p rather than %d.
4156 81. There was a concern that a line of code in smtp_in.c could overflow a
4157 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4158 initially expressed, the concern was not well-founded, because trailing
4159 spaces are removed early. However, if the trailing spaces were followed by
4160 a NULL, they did not get removed, so the overflow was possible. Two fixes
4163 (a) I re-wrote the offending code in a cleaner fashion.
4164 (b) If an incoming SMTP command contains a NULL character, it is rejected
4167 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4168 runs initgroups(), so that if the Exim user is in any additional groups,
4169 they will be used during message reception.
4175 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob